UV LPs,

Welcome to our 2026 Annual LP Update. As a reminder, at Unpopular Ventures we focus on “the best companies, off the beaten path.” We invest in exceptional founders who are building high potential businesses, that are non-consensus in some way.

We invest from both an AngelList Rolling Fund, with typical checks of $25–100k, and we also lead one of the biggest syndicates on AngelList — in which we invite our LPs to invest in specific companies through us, and together we typically invest checks of $100–500k. We usually make ~80–120 investments per year from our fund, and 15–30 from our syndicate.

To help us find great opportunities around the world, we leverage a “Scout Program,” through which we share significant portions of our carried interest with anyone who helps us to identify, evaluate, and diligence companies that we invest in. If you would like to refer an opportunity, please check out our Scout Program Guidelines. By tapping into the network of our community, we are able to invest with a lot more reach and volume than we’d ordinarily be able to with only two investing partners.

This is our annual update, which will be longer than our quarterly updates. In addition to our usual commentary about the market, investment thinking, and an update on our portfolio, you’ll also find an expression of gratitude to our community, an explanation about how we do our portfolio tracking, and links back to some of our other past content.

Thank you all for your continued partnership!

Sincerely,

Peter Livingston and Thibault Reichelt

Unpopular Ventures

GRATITUDE

We have a lot to be thankful for, so we always like to start our annual updates with gratitude. Thank you to:

OUR PORTFOLIO FOUNDERS. Thank you to the Founders and teams of the nearly 600 portfolio companies of ours (past and present) — who allowed us to invest in their companies, and have worked like crazy to build their dreams into realities. Great Founders have endless choices of investors to take money from. Thank you for choosing us, and for allowing us to ride along as a small part of your journeys.

OUR LPS. Thank you to all of our ~5700 Syndicate LPs and 100+ Rolling Fund LPs who have entrusted us with your capital. There are endless places to invest your money: dozens of asset classes, thousands of VC firms, and hundreds of syndicates and funds just on AngelList to choose from. We are deeply grateful to you for choosing us.

OUR SCOUTS, GUEST LEADS, AND VENTURE PARTNERS. As you know, we share carry with anyone in our community (including LPs, Portfolio Founders, and friends) — for their assistance in identifying, assessing, and negotiating one or more of our investments. Thank you to our extended family of contributors who have helped us to do more than we would have been able to on our own. There are hundreds of you — you know who you are — and we are grateful to you.

THE ANGELLIST TEAM, who make the magic happen behind the scenes. Before AngelList existed, it was not possible to operate at the scale we do with such a small team; it is extraordinary what they have unlocked for us. Special thanks in particular to Lauren Delfs and Tarina Touret — our primary managers at AngelList over the last year, who have been wonderful to work with.

MARKET AND INVESTING COMMENTARY

“EMBRACE THE ESTABLISHED, NEGLECT THE NEW”

In case you didn’t know, we are a non-consensus VC firm. We know “non-consensus” is a bit of a joke in our industry. Every VC claims to be non-consensus, but somehow they are all non-consensus in the same ways and chase the same hot deals. We aren’t perfect, but we do strive to live by our name.

If our aim is to make non-consensus investments, it helps to start by thinking about the opposite: what is consensus today? As Charlie Munger famously said, “invert, always invert.” So we thought we’d share some ideas about where consensus lies, and how they might present opportunities for us (and you) to take advantage.

We’ve written repeatedly about the “K-shaped” economy in the past, and how it shows up at all levels of the economy. I.e. a small number of people and companies are doing better than ever, while everyone else is having a hard time. We see it at the individual level — more wealth inequality than ever before, the richest are getting richer, while the poor/middle class are seeing wages and opportunities decline. We see it in the public markets, where the 7–10 biggest companies in the world keep growing in value, while all the rest are flat, if not down. We see it among VC firms, where the biggest are raising an increasingly disproportionate share of the money, while most smaller firms are shrinking or closing. And of course we see it among startups too: a small number of companies are getting pumped with money at insane valuations, while for the rest, it’s the toughest market anyone can remember.

That’s a phenomenon, but that phenomenon is driven by people’s decisions. What decisions? We’d argue that the widespread consensus decision of the moment is: “embrace the established, neglect the new.”

We see it in hiring. New grads are having the hardest time in the labor market:

We see it in asset allocation, with many LPs questioning why they should invest in Venture Capital at all, when the Nasdaq and S&P 500 performed better over the last decade:

Within VC, LPs and GPs have significantly shifted toward late stage/growth investing, shunning early stage. Even Jason Calacanis, who long touted his early stage prowess (did you know he was the third or fourth investor in Uber? 😉), has shifted to late stage:

If you stand back and look at it, it’s kind of a strange phenomenon. Have the established choices always been preferred? Of course not (or at least, not to this extreme). We are old enough to remember the articles from the late 2010s into 2021, celebrating early stage VC as the best asset class ever. Or the endless stories of entry level Computer Science grads with $200k starting salaries.

Will the established choices be preferred to this extreme in the future? We don’t believe so.

Setting aside the irony that everything established was once new, and that if we don’t support the new we will strangle the supply of the established in the future — we also don’t believe it’s rational, nor in market participants’ immediate economic interests, to so overwhelmingly prefer the established.

It’s particularly odd that this is the preference right now, just as we are experiencing the biggest technological revolution since the internet (and possibly bigger) in AI. If this new technology is going to re-orient the entire world, wouldn’t the young people be the best hires? Wouldn’t new startups, built around this new technology from the outset, be the best things to invest in? Shouldn’t the greatest opportunities be at the earliest stages? I.e. why has the majority of the venture market deemed OpenAI and Anthropic the main things worth investing in? And shouldn’t the small, nimble, and hungry VC firms be the best at identifying the best early opportunities? I.e. why have all the LPs crowded into A16Z, Sequoia, General Catalyst, and Thrive?

Maybe it’s fear, or uncertainty amidst chaos. No one really knows yet what AI means, how it affects the world, or how disruptive it will be. With so much change afoot, perhaps it feels safer to reach for the things that have been around for a while, the people you know, and the firms that have longer track records of performance. “No one ever got fired for buying IBM” as they say.

Or to put a finer point on it, people know there is tremendous change happening, but they don’t trust themselves to navigate it. So they say “Google seems to know what they are doing;” or A16Z; or “I’ve worked with Tom for 10 years, he’s always good at figuring it out;” or “everyone says Claude is the best, so I’ll just use that.” They outsource the adoption of the novelty to the established entities they trust.

On the investing side, recent performance is probably part of it too. Many agreed that early stage VC was the best asset class to invest in, leading up to and peaking in 2021; too much money poured in, valuations got crazy, too many competitors got funded, and the returns since then have been terrible. LPs raced for the exits (especially retail), which exacerbated the downturn further. The LPs that have stayed (especially institutional) have invested less and/or concentrated their allocations around the big firms they perceive to be safe. The market always overdoes it, to one extreme or the other.

So if the mass consensus decision right now is to embrace the established, we posit that the non-consensus move right now, and where the alpha lies, is in embracing the new.

If you are a hiring manager, now is probably the best time in a decade to hire new grads. They are desperate, hungry, will work harder for less money, and will be more loyal to you than ever. Most importantly, they will be the best able to implement AI in your organization, in novel ways that your experienced employees won’t think of.

If you are deciding which asset classes to allocate to, yes, the data shows that the public markets have outperformed average VC over the last decade. Does that mean the next decade will be the same? Of course not. In fact, it makes it more likely that the opposite will happen.

Every now and then we see one of these tables that ranks the performance of asset classes by year. They only show publicly traded assets (privates are opaque, hard to measure, and the marks lag). Usually, the intent is to demonstrate how unpredictable the market is, with the best-performing asset classes perpetually changing one year to the next:

What is striking in this current table, however, is how frequently Large Caps have outperformed this decade — 4 times out of the last 7 years, and near the top in most of the other years too. Clearly, the biggest and most established companies have been the place to be.

But just as the best performing asset class changes from one year to the next, it also changes from one decade to the next. The prior decade did not have a single year in which Large Caps were the best:

In fact, the last time Large Caps outperformed, was 1998:

This feels like a surprising insight. Not only did Large Caps outperform back then, but they did so repeatedly, kind of like now. 3 years as number 1, and within the top 5 for 6 years in a row. Perhaps most interesting of all: it coincides with a period of similar technological disruption: the rise of the internet.

It appears that during the last period of significant technological change, the market’s initial reaction was to similarly buy the biggest, most established companies.

One might argue that this time is different, because today the biggest companies are all tech companies, and they weren’t back then. Well, here are the 10 most valuable companies in 1999:

• Microsoft: Led the global market during the height of the tech bubble.
• General Electric: A massive industrial and financial conglomerate.
• Cisco Systems: A high-flying networking hardware giant of the dot-com era.
• ExxonMobil: Formed from the 1999 merger of Exxon and Mobil.
• Walmart: The dominant American multinational retail corporation.
• Intel: The leading semiconductor chip manufacturer of the decade.
• Nippon Telegraph and Telephone (NTT): A major Japanese telecommunications firm.
• Lucent Technologies: The telecommunications equipment spin-off from AT&T.
• Nokia: The world leader in mobile phones at the turn of the millennium.
• IBM: Long-standing hardware and computing leader.

We count at least 7–8 tech companies here among the top 10. We’d argue General Electric was as much a tech company back then, as Apple is today. And even Walmart (which we didn’t count) was arguably as technologically interesting back then as Amazon is today. I.e. which part of Amazon is “tech?” That they sell stuff on the internet? Or that they have AWS, which launched in 2006? When was the last time a lot of these — especially Apple, Amazon, or Meta — launched something truly new?

In fact, are all of the Mag7 really “tech” companies anymore? Yes — they are big, impactful, and they have great businesses for the moment. But it’s hard to argue that they are more technologically innovative today, than the 10 biggest companies in 1999 were.

You might also argue that the P/E ratios of most of the Mag7 today are low, and therefore they remain good investments. Well, the P/Es of IBM and GE in 1999 were ~25x and ~30x — right in line with the P/Es of most of the Mag7 today (excluding Tesla).

And by the way, P/E is only one part of the story. Free cash flow is another:

After the long period of outperformance in the 1990s, Large Caps then had 3 negative years in a row, and it took them 20 years to be the best performing asset class again. So as much as everyone loves the Large Caps right now (especially the Mag7), and it feels like they are destined to go up forever — we won’t be surprised to see the pendulum swing back in the other direction — just like it did after their last run. It’s impossible to time the market precisely, and it could be another year (or longer) before it happens. But no tree grows to the sky. Eventually, the market will turn and Large Caps will have a lost decade (or two) while other asset classes outperform — right when everyone counted the other asset classes out.

Zeroing back in on VC, the point is: VC will not always underperform. Yes, it’s had a rough few years, but it won’t last forever. And within VC itself, if you are choosing between late stage or early stage, late stage is the crowded side right now. 2 companies, OpenAI and Anthropic, absorbed 43% of all VC funding in the first half of this year (source), and 87.5% of all dollars went to rounds of $100M or more (source). In Q2 in particular, out of $137B of VC invested, only $4.9B went to seed or angel rounds — a mere 3.6% (source).

Early stage is empty. We have been surprised at how many early stage companies that would have raised easily in the past, are struggling to raise now — even those with exceptional teams, strong traction, and big market opportunities. It feels like the early stage market has finally come back to earth (for all but a handful of “hot” companies), and because of that we think the current vintage is going to deliver exceptionally good returns.

The “subsidized” mass excess buildout of AI — perhaps comparable to the fiber optic or railway booms of the past — is laying fertile ground for the next generation of startups to create and capture value in new ways. While Anthropic and OpenAI are currently loved by investors, the models from Google/Gemini, SpaceX/Grok, DeepSeek, Xiaomi, Zhipu, MiniMax, Moonshot, and Alibaba are not much worse. Because of this competition, combined with high CapEx, minimal moats, and low margins — we don’t believe these companies will deliver good investment returns. But the competition over performance and price *is* creating cheap and powerful tools for everyone else.

People love to say “in a gold rush, sell the picks and shovels.” But what if there is a picks and shovels rush? If a dozen+ companies are spending trillions of dollars, fighting to make the cheapest shovels that dig 1000x faster… maybe it’s better to be (and invest in) the gold miners.

AI is probably the highest leverage tool that has ever existed. In many ways, having access to LLMs for coding and other purposes is a great equalizer between entrepreneurs and large companies. There has never been a better time for small teams to accomplish huge goals. Thanks to the efficiency gains and automation opportunities brought about by AI, companies being started today can also remain lean as they scale.

We therefore believe that unpopular startups going after billion-dollar “niches” may have a better chance to capture long-term value than the large enterprises we read about in the news which operate in hyper-competitive industries with few moats. Our portfolio companies are building monopolies in domains that will have great profit margins 10 years from now precisely because these domains are currently underestimated by most market participants.

In the context of the current technological paradigm shift, it is worth remembering how Schumpeter articulated that creative destruction is an essential part of capitalism and the fundamental driving force behind its economic progress. Our portfolio is a diversified bet that this creative destruction will continue to propel young companies to the forefront. Especially in times of great upheaval such as the one we are going through at the moment.

WE DON’T SELL EARLY

As we’ve shared in the past, my (Peter’s) Partner Thibault has held a role for many years in which he invested in AngelList syndicates on behalf of a family office. He and that family office have been one of the biggest LPs on AngelList for many years, with over 400 investments on the platform.

Know what his biggest frustration was? GPs selling positions in winners too early.

He invested in several mega winners that should have returned 100x+, including Anthropic, Perplexity, Wintermute, and Groq. He picked so well, and wanted to hold these positions for as long as possible to capture all of the compounding growth over a long period of time. Sadly, the syndicate leads in most of his best investments sold large parts (or all) of the positions out from under him after only a few years, returning a fraction of what he expected.

That’s one of the biggest problems with investing in SPVs (or VC as a whole for that matter). You are not only betting on the individual companies; it’s a bet on the GPs too. And that extra layer tends to add more downside than upside.

So we thought now would be a good time to remind everyone in our LP community (both syndicate and fund): we don’t sell early.

We know that not everyone likes that, which is why we want to be clear about it. We frequently get requests from LPs for early liquidity — so evidently some segment of our LP base doesn’t want to hold for a long time. But we also know that a lot of you (especially the LPs that are more serious and sophisticated) *do* want to hold for as long as possible. That’s the side we are aligned with.

We are big believers in the power of compound interest over a long period of time (or technically, compound growth in this case). While the picking matters, the actual dollar gains mostly come from the holding and waiting.

We often think about the Sequoia partners who invested in Nvidia’s seed round, who still hold their personal shares today. Or John Doerr, who led Google’s first round, and still sits on their board. 99%+ of their dollar gains in those investments have come after the first 10 years of holding. That’s how compounding always works: the biggest gains come in the latest years.

To illustrate that with math: if you are fortunate to invest in a startup that ~doubles in value year after year consistently (has roughly happened many times in history), after 5 years your multiple will be 32x. Many investors would sell then, thinking it was a big win. But if you held for another 5 years, you would have 1024x. By selling at year 5 instead of year 10, you left 97% of the gains on the table. Same thing if the investment grows 42% annually, and you sell in year 10 vs. year 20.

Time will tell if we have an Nvidia or Google in our portfolio, but even if we have lesser successes — I for one will be holding my personal shares in a lot of our investments 30+ years from now. Of course, when our companies IPO, we’ll distribute shares or cash (LPs’ choice). And when companies get acquired, the proceeds will be distributed to our LPs. But the shares I receive from my carry and personal investments in our companies — I will be holding for as long as possible. I simply know that great companies go a lot farther than anyone expects, and that compound interest really kicks in in the latest years. I’m not here to make a quick buck. I invest in startups with the explicit purpose of building generational wealth over a long period of time.

The point of this that is important to you, our LPs: if you want your cash back in a short amount of time, don’t invest with us. There are plenty of other funds and syndicates that will shoot for quick exits.

But if you are an investor that is playing this game to make generational wealth over a long period of time, maximizing your gains through the power of compound interest, then know that we will not sell our investments out from under you. You can trust us to hold for the long term.

VOLATILITY

On that last topic, one thing worth touching on is that when you hold high performing assets for a long time, the price is bound to be volatile. Have you ever wondered what the stock price of a startup would look like, if it was publicly traded from the beginning? You don’t actually have to wonder; this is the price chart of one of the most successful startups of the last decade:

You can argue that Bitcoin is different from other startups, but even so — I am quite certain that anyone who has seen the trajectory of a successful startup up close, would agree that the path was comparably volatile, and that if the stock price was publicly traded, it would not look much different.

In fact, going beyond the startup stage, most of the best performing assets in history have also been the most volatile. High returns and high volatility go hand in hand. Look at the max drawdown of the 7 companies that today are the most valuable in the world:

They all had bumpy roads, even as public companies. Amazon in particular has had 8 drawdowns of 30% or more. Even SpaceX already had a 53% drawdown in its first 6 weeks of trading.

The point is, if you choose to hold assets that will deliver the highest returns over the long term, you also have to accept high volatility along the way. And that is more true than ever in VC and startups.

In times of volatility, I’m often reminded of Warren Buffett’s famous quote: “price is what you pay, value is what you get.” The price of an asset jumps around with the madness of the market; sometimes wildly overpricing the asset, and sometimes pricing it irrationally low. But the price is not the value — and if you don’t intend to sell, the near term price doesn’t really matter. Over time, as long as the business keeps growing in underlying value, the price will eventually follow.

The volatile mismatch of price to value can stay in one direction for many years too. In our early years 2019 through 2021, a lot of money flowed into VC, and prices for many of our leading positions probably rose above their realistic values. In early 2022, AngelList reported an IRR for our aggregate portfolio of over 100%. That was the peak of a multi-year price:value expansion. Since then, the tide has steadily gone out for many years in a row. Although a lot of our leading positions have grown dramatically in underlying value, the price:value ratios have continuously compressed — in many cases reaching prices that I believe are irrationally below the companies’ intrinsic values. Our aggregate IRR has come down with it, and now lies somewhere in the high teens (I don’t have an exact number right now, because AngelList has not yet updated the marks for some of our recent big movers). I have no doubt that eventually, the price:value will expand again; but it is certainly gut wrenching to ride this volatility in the near term.

You’ll read below that we had a bit of volatility in our portfolio this quarter. My intent here is to remind — ourselves more than anyone — that volatility is one of the prices we pay in our journey toward outsized investment returns. Volatility sucks when it’s in the downward direction, but it’s super fun when it’s up. At some point, maybe soon, I am confident we will start to see more volatility in the fun direction again.

PORTFOLIO

SUMMARY

$87M of capital invested has grown into $190M of portfolio value. This is down vs. last quarter ($85M/$208M).

Our numbers dipped this quarter because one of our larger positions repriced at a lower valuation. We’re not concerned; the business is performing better than ever by every operating metric, and we expect the price to recover to its prior high and exceed it within one to two years.

The company didn’t need the capital, it had ample cash on hand, but chose to raise ahead of an inflection point to make sure growth wasn’t capital-constrained. The market is a bit wonky these days, pricing everything except front and center AI at low multiples, and this round was no exception — valuing the company at a surprisingly low multiple of revenue. The round was clean: 1x non-participating preference, no pay-to-play, no recap. The only change for us is a lower marked share price, not a change in our ownership or the deal’s structure.

It’s uncomfortable to see a markdown, but we think the founder made the right call for long-term value creation. This kind of reset, done from strength rather than distress, has preceded some of the best outcomes we’ve seen, and we expect this to be one of them.

We’re not able to name this company publicly until the round is formally announced, but we’ll share details directly with LPs who are invested in it as soon as we have the founder’s go-ahead.

On the positive side, we also had another portfolio company jump dramatically in value, and we now count it as a top position.

COMPANIES

We have almost 600 portfolio companies now, but 4 of them already make up 42% of our aggregate portfolio value. As we’ve discussed in the past — this is the way venture investing goes. Even if you have lots of successful investments (as we do), the nature of the power law is that some go so far that they end up being worth more than everything else combined. Our 4 top positions are:

Novig: a leading sports prediction market. This is a new top portfolio company for us, with a new round underway at a dramatically higher price than the last. It is now live across the US and growing rapidly, having finally overcome the regulatory hurdles that held it back for a few years. Thanks to Thibault for leading our investments in the company, and Cindy Bi’s collaboration on a syndicate we led for them at a $20M valuation in 2023.

This one is particularly exciting because it is now the largest single position in our rolling fund, at $4.9M against a total of $100k invested. We have invested $24M across all of our Rolling Funds to date (since 2021), so this one company is already worth ~20% of what we’ve invested. It does not have to go much further to become an “everything returner” for our RF.

Although we have invested in the below companies from our fund, our original investments in them (at the lowest valuations) happened before our fund existed, and were therefore done via the syndicate only.

Yassir: The leading super app in Francophone Africa. We led their seed round, and they are now a unicorn+. Yassir has an enormous amount of revenue, among the highest in our portfolio. Thanks to Thibault for this one; it was the first company he ever brought to UV.

Jeeves: Stablecoin-native financial operating system for global enterprises. We were their first investor (besides YC), they now serve 3,000+ enterprise customers in 30+ countries, and have grown revenue/volume 3.5x year over year with 80%+ gross margins and best in class NRR north of 130%.

Blissway: Next generation tolling and traffic management infrastructure. We were their first investor, and the company’s tech is now deployed on highways across multiple US states. The company is profitable, growing, and has conducted regular tender offers to buy shares back from employees and investors.

And here are some of our other portfolio companies that are on promising trajectories and might grow to become large positions in the future. The list is by no means exhaustive.

Zepto: Leading Indian quick-commerce startup that delivers groceries and essentials in 10 minutes. For a long time, it was the fastest-growing company in Y Combinator’s history. Thibault invested us in their first round, and they are now a multi-unicorn. The only reason it’s not a top position above, is that we were only able to invest from our fund (and not the syndicate), so our starting ownership was on the smaller side relative to the others.

Yummy: Venezuela’s largest technology “super app” for delivery and ridesharing. We were one of their earliest investors, and recently followed on in their Series B. Hat tip to Steven Coulis for the intro.

Volantis: Photonically integrated computer architectures and motherboards designed specifically to accelerate AI data centers and large-model inference. We were one of their first investors, were followed by Sam Altman, Naval Ravikant, Lachy Groom, and they recently raised a large Series A. H/T: James Graham

Undermind: An AI “co-researcher” whose deep-search agent scours the scientific literature to surface the most relevant papers on a topic far more thoroughly than a keyword search. It’s a striking example of how, in the age of AI, a tiny team can become profitable almost immediately and never need to raise beyond a seed round.

Ultralight (fka Vibrant Health): Software that helps independent clinicians deliver precision, data-driven care, turning hours of patient-data analysis into instant, personalized treatment plans. As more doctors leave traditional medicine to build a better model, Ultralight recently raised $9.3M to power what comes next.

Turion Space: Builds spacecraft for in-orbit servicing, space-domain awareness, and debris removal, giving commercial and defense customers eyes and hands in orbit. It recently closed its Series B and has won U.S. Space Force contracts for its debris-capture technology.

ThisGen 911: A workforce platform for 911 emergency call centers spanning the full employee lifecycle — candidate screening and realistic job previews (PreHire), high-fidelity call and radio simulation training with AI-powered callers, and automated QA that reviews live calls and coaches dispatchers. It now serves 911 centers across 40+ states, and reports cutting training time by 40%. H/T: Tommy Leep for the intro and co-syndicating.

The Swarm: A relationship-intelligence platform that maps the collective networks of a company’s team, investors, and advisors to unlock warm introductions for sales and recruiting. It recently acquired Commsor, consolidating its lead in warm-intro and go-to-market relationship data.

Taiv: Uses AI to take over the TVs in bars and restaurants, replacing generic commercial breaks with targeted, venue-specific advertising and content. The company raised a ~$14M Series A to scale its in-venue ad network across North America, and has been growing extremely fast in recent quarters.

Synova: Maker of the Synova WAVE, a device that processes fat tissue to extract highly viable fat cells in 3 minutes for regenerative-medicine and fat-transfer procedures. It recently obtained its 510(k) clearance from the FDA.

Sygaldry: Developing novel quantum hardware to rethink the infrastructure that powers AI, aiming to make large-scale AI compute dramatically more efficient. The company has already raised ~$139M across two funding rounds. We invested in their first round during YC.

Stepful: Online healthcare training platform that prepares people for entry-level medical jobs in 4–5 months. It offers remote coursework, live online classes, and optional hands-on clinical externships to help students get certified and find jobs. We invested in the first round and they have raised $111M to date.

Sniffspot: “Airbnb for dog parks,” a marketplace where people rent private yards and spaces by the hour so their dogs can play safely off-leash. It has become the go-to marketplace of its kind, giving homeowners a side income and dog owners safe, private space. The company is cash flow positive and growing at a steady clip. H/T: Pascal Levy-Garboua for co-syndicating.

Sky Fortress: A low-cost acoustic detection system — a distributed network of microphones plus AI — that locates and tracks incoming drones cheaply and quickly, first proven at scale in Ukraine. The technology has been deployed by the US across the Middle East. Thanks to Sergii Zhuk, we were among the first investors when Sky Fortress got started.

Rio: A stablecoin platform for Latin America that lets businesses and individuals move, hold, and convert dollars through stablecoins and OTC liquidity. Rio is now processing over $1B per quarter.

Ready: A cloud-based, data-driven software platform that helps ISPs and governments plan, fund, and build broadband networks — the quiet infrastructure behind much of America’s broadband buildout. It is well positioned to benefit from the tens of billions in federal BEAD funding now flowing to close the digital divide. H/T: James Graham

Rampa: Payment infrastructure for enterprise cross-border financial flows, handling multi-currency disbursements, settlement, and compliance for companies operating across LATAM, the US, Europe, and Asia. Rampa grew 10x last year. H/T: Helen Chen

Polymath Robotics: Offers plug-and-play autonomy software that lets manufacturers add self-driving capability to large industrial vehicles — tractors, boats, mining equipment — without building an autonomy stack from scratch. The company is bringing general-purpose autonomy to the heavy industries that big self-driving firms overlook. Nice progress and growth in recent quarters.

Pave: A fintech offering cash-flow-underwriting and credit-risk APIs that help lenders and card issuers make smarter decisions using real-time bank-transaction data. The company is making great progress and recently raised a Series A. H/T: Jeson Patel

Palitronica: Builds hardware-and-software security technology that protects critical infrastructure and defense supply chains from tampering at the electronics level. Recognized as a top security innovator by In-Q-Tel, it has closed a Series A to scale across defense and critical-infrastructure markets.

Orbital Operations: Developing high-thrust space vehicles designed to maneuver, defend, and service satellites in orbit. Excellent progress since we first invested, so we followed on, and the company recently secured several major government contracts. H/T: Sergii Zhuk

nSave: A Swiss fintech that gives people from high-inflation and unstable economies access to safe, hard-currency ($/€/£) accounts abroad. The offshore banking long reserved for the wealthy, opened up to everyone — tackling financial exclusion for hundreds of millions of people.

Marathon Data: A brand-measurement platform that uses machine learning — validated against causal incrementality studies — to quantify how much incremental revenue a company’s brand marketing actually drives across the full funnel. Excellent progress since our investments; growing, and profitable.

MagicMind: A fast-growing consumer wellness brand best known for its two-ounce “mental performance” shots — blends of nootropics and adaptogens like lion’s mane and ashwagandha marketed as a healthier alternative to coffee. It has built a large direct-to-consumer subscription following.

Kashin: A mobile financial app providing microcredit and everyday banking to the underbanked, informal-economy workers of Latin America. The YC-backed company is becoming a preferred financial app for millions the traditional banks ignore. H/T: Ivan Montoya, Ali Jamal

Inversion Space: Building reentry vehicles — autonomous capsules that can store cargo in orbit and deliver it back to any point on Earth within an hour, on demand. It has partnered with Anduril on their Golden Dome bid.

Hint Health: The leading software platform powering direct primary care, giving membership-based, cash-pay clinics the tools to enroll patients, bill memberships, and run their practices. It has become core infrastructure for the fast-growing direct-care movement. We were a major investor in them before they were majority-acquired by Private Equity, and they have grown significantly since.

Hapi: A commission-free investing app that lets people across Latin America buy U.S. stocks and ETFs directly from their phones. It is opening access to U.S. capital markets for a fast-growing base of retail investors in the region. H/T: Tiago del Rio

Hadrius: An AI-native compliance platform that automates SEC and regulatory compliance for investment managers and financial firms, cutting review time by roughly 70%. It raised a $22M Series A from CRV and already protects hundreds of institutions.

GitButler (fka Sturdy): A modern, visual Git client that lets developers work on multiple branches at once and wrangle the messy, fast-moving changes produced by AI coding tools. It raised a $17M Series A in 2026 to reinvent version control for the AI-driven development era. H/T: Declan Kelly

Felix: FinTech platform that lets people in the United States send international money transfers (remittances) to Latin America directly through WhatsApp using an AI-powered chatbot. It eliminates the need to download separate mobile apps or visit physical remittance offices. Big news coming soon here.

Engage Bio: A biotech developing non-viral delivery technology for genetic medicines — a safer, more scalable way to get DNA-based therapies into the body. We originally invested on a $15M cap, and it was recently acquired by Eli Lily in a deal valued at ~$202M. H/T: Sergii Zhuk

Decentro: An API-banking platform that lets companies embed payments, collections, and financial services into their products through a single integration, primarily in India. Recently crossed $10M ARR.

CopilotIQ: Provides in-home remote monitoring and virtual nursing to help seniors manage chronic conditions like diabetes and hypertension, catching problems before they become emergencies. It acquired Biofourmis to build one of the largest in-home chronic-care platforms.

Constrafor: A procurement and finance platform for the construction industry that streamlines subcontractor management, payments, and compliance while offering early-payment financing to ease cash-flow crunches. It has raised over $260M in equity and credit to fund contractors and keep projects moving. H/T: Zach Kruth

Conduit: Deploys AI agents that automate guest communication and back-office operations for hospitality businesses, handling customer service across email, WhatsApp, phone, and social while adding voice agents, chat agents, and internal workflow automation. It already serves 300+ hospitality brands across 120+ countries — including Marriott — with enterprise-grade security (SOC 2 Type II, HIPAA). Excellent progress since we invested in their seed round. H/T: Gabriel Jarrosson

Community Phone: A landline phone company for the modern age — its wireless base plugs in anywhere with cell coverage to deliver reliable home phone service with powerful spam-call blocking, no copper wires required. It has become a fast-growing alternative for customers nationwide as legacy carriers retire traditional landlines. H/T: Rohit Taneja

Code Four: Builds AI software that modernizes the police workflow — automating report writing, field reporting, and investigations so officers spend less time on paperwork. The YC-backed company is equipping more and more police departments.

Chowdeck: Nigeria’s leading on-demand food-delivery platform, connecting restaurants and customers with fast, reliable logistics. Profitable and fast-growing.

Capi Money: A cross-border payments platform that lets African importers pay their international suppliers quickly and cheaply, settling invoices in as little as 48 hours. Its gross transaction volume now exceeds $1B a year.

Brellium: An AI-powered clinical-compliance platform that automatically audits patient charts against payer and regulatory rules, helping healthcare organizations stay compliant and capture every earned dollar. It supports over 200 healthcare organizations across all 50 states.

Baton Market: An online marketplace and toolset for buying and selling small businesses, bringing transparency and modern software to what has long been an opaque, broker-driven process. It has become the best place to buy or sell a small business.

Avails Medical: A diagnostics company whose eQUANT system rapidly determines which antibiotics will beat a bloodstream infection, delivering results directly from positive blood cultures far faster than traditional lab methods. Johns Hopkins has started to use the company’s eQUANT technology.

Amara: A wellness platform that brings fitness classes, personal training, and mental- and social-wellness programming directly to residents of apartment buildings through an app. It is turning residential buildings into full-service wellness communities. Nice growth since we invested. H/T: Tommy Leep for the intro and syndicate co-lead.

99Minutos: A last-mile logistics and delivery company built for Latin American e-commerce, offering fast, sustainable shipping and returns across Mexico, Colombia, Chile, and Peru. It has grown into one of the region’s leading logistics providers. H/T: Ed Roman for the syndicate co-lead.

222: An app that uses machine learning to bring back real-world serendipity, matching strangers into small groups for dinners and activities designed to spark genuine friendships. It has built a passionate following among young people seeking offline connection in major cities. Excellent growth since we invested, and raised a Series A led by Upfront Ventures.

On the earlier side, there are a lot of exciting things going on. To name but a few:

Tornyol: Builds tiny, 40-gram autonomous micro-drones that detect mosquitoes by sound and kill them in mid-air. The YC-backed company recently notched the first air-to-air insect kill and opened its mosquito-hunting drone for preorder.

PAX Markets: Building financial-exchange and high-frequency-trading infrastructure directly onto a single chip (FPGA), collapsing the matching engine into hardware for extreme speed. The YC-backed company has built the lowest latency exchange in the world — literally an exchange on a chip.

Job Machine: An AI-powered job-placement platform that automates sourcing, matching, and placing candidates into roles. It has already passed $1M ARR.

Daivin: Developing tankless dive gear that lets divers breathe underwater without carrying a traditional scuba tank, aiming at breath autonomy at sea, on land, and beyond. The YC-backed company is pursuing a first in human history: compact, tankless underwater breathing.

Lastly, one of our most fun recent investments to watch has been Memelord, an AI-powered meme marketing platform for social media marketers, founders, and creators to generate, schedule, and scale viral memes. Thanks to Jack Plank for the intro. They are doing great, growing fast, and we loved this post:

We regularly post company news to X, so please follow us there for more updates.

PORTFOLIO STATISTICS

Our aggregate numbers, across every investment in UV’s life:

Numbers for only the UV Syndicate:

Annualized numbers for only the UV Rolling Fund:

Statistics about our investing activity:

We rolled over ~$83k of uninvested capital from the Q2 Rolling Fund, and raised another $623k in the Q3 RF.

Thank you for reading, and thank you for your support of Unpopular Ventures!

APPENDIX

PAST CONTENT

Links back to some of our past updates. Quarterly updates from the past year:

• Q2 2026
• Q1 2026
• Q4 2025

All annual LP updates since UV inception:

• 2025
• 2024
• 2023
• 2022
• 2021
• 2020
• 2019

PORTFOLIO TRACKING METHODOLOGY

As we originally explained in the Q4 ’22 update, we present our numbers to be in line with the SEC’s guidance that if fund managers present gross performance numbers, they must also present the numbers net of all fees. This is a bit tricky for us, because we offer numerous fee options to investors in our rolling fund — each of which would require a different calculation. 2 and 20 is the default option, but we also offer 1 and 30, and even 0 and 20 with a 6 year commitment. That last option is our favorite (we prefer long term stable capital) but recognize that 2 and 20 is the market standard. Most LPs prefer a shorter commitment duration — so we have set that as the default choice.

To produce “net” numbers, we standardize them along the 2 and 20 fee structure. To do this, we are starting with the amount that we actually invested into companies, which is “real” — regardless of whether you are signed up with management fees or not. We are then creating an “imaginary” number for capital raised, which is the amount of money we would have had to raise based on how much we invested, if all of our LPs had signed up with 2% annual management fees. But because some of our LPs are signed up without management fees — this imaginary “raised amount” does not actually represent how much we raised. It just serves to facilitate proper calculations net of 2 and 20 fees. If you have a different fee structure than 2 and 20, the net numbers are different for you.

We also separate out our syndicate and fund investing. We kept them together before Q4 ’22, because our intent was to simply measure ourselves — to see if we were doing a good job investing — and we didn’t really care about which type of vehicle the capital came from. But because the fees on these two entities are different, we have to separate them in order to produce numbers that are net of fees.

Marking to SAFE Caps

In addition to incorporating fees, dilution, and carry into our calculations, one thing we do that differs from AngelList is that we mark to SAFE caps — both up and down.

The intent is to produce more-accurate reporting. For example, we have some companies that have raised a down round after an equity round on a lower valuation cap SAFE. AngelList still holds those at the last equity round price, but we value them lower. We do it in the upside direction too.

We approach this with some nuance:

• Must be within reason. If a company raises on an outrageously high cap (or uncapped) or with a huge discount — we don’t count it. It only counts if the discount is zero/minimal, the cap is commensurate with the company’s progress, and reasonably close to what the valuation would be if it had been a priced round.
• If it’s a post money cap, we treat it as the post money valuation. If pre, we treat it as the pre-money valuation.
• We track and incorporate all of the dilution that would occur between the rounds, to the best of our ability. For example, if we invested on a $5 M post money SAFE, and the company then raises $2 M on a $12 M post money SAFE, we would mark that as a 2.0x — the delta between our post money (5) to the next pre money (10). If they then took $3 M more on an $18 M pre money cap SAFE, we would mark it up to a 3.0x → (10/5)*(18/12) = 3.
• If a priced round is signed but not closed yet, we use similar math, but also add extra dilution to account for potential option pool expansion. If there has been no priced round for a long time, we add an extra 10% dilution. If there was a priced round very recently, with the option pool already refreshed, we incorporate less.
• As soon as we receive a pro forma cap table or the exact share price, we revert to the official share price multiplied by the number of shares we have.

There is some debate about whether you can mark to a higher cap SAFE/note, or if you have to wait for a priced round before updating the valuation. AngelList, for example, only marks to priced rounds. However, many other seed stage VCs mark to caps. In fact, the NVCA appears to believe it’s acceptable to use a cap as the effective valuation:

Historically, it made sense to only mark to priced rounds, because most rounds used to be priced rounds. And particularly for AngelList, with thousands of investments and a lean team — it’s most efficient for them to only mark to a share price. However, with the growing use of SAFEs, we are increasingly seeing early stage startups do 2–6 SAFE rounds over 2–4 years before eventually doing a priced round.

We certainly could wait for only priced rounds before updating our marks — but that means there will be no meaningful data to report for years — and the data will lag. We (and probably you too) would like to estimate our investment performance on a more granular level. We think the best way to do that is to mark to reasonable caps as rough markers of valuation, while calculating the effect as if they had been priced rounds.

If another investor is investing on a higher or lower cap SAFE — they are legitimately investing on a higher/lower valuation — and we think it’s reasonable to incorporate that.

Unpopular Ventures: Annual Update 2026 was originally published in Unpopular VC on Medium, where people are continuing the conversation by highlighting and responding to this story.

This post is a segment from our 2026 LP update, written by Peter Livingston and Thibault Reichelt

In case you didn’t know, we are a non-consensus VC firm. We know “non-consensus” is a bit of a joke in our industry. Every VC claims to be non-consensus, but somehow they are all non-consensus in the same ways and chase the same hot deals. We aren’t perfect, but we do strive to live by our name.

If our aim is to make non-consensus investments, it helps to start by thinking about the opposite: what is consensus today? As Charlie Munger famously said, “invert, always invert.” So we thought we’d share some ideas about where consensus lies, and how they might present opportunities for us (and you) to take advantage.

We’ve written repeatedly about the “K-shaped” economy in the past, and how it shows up at all levels of the economy. I.e. a small number of people and companies are doing better than ever, while everyone else is having a hard time. We see it at the individual level — more wealth inequality than ever before, the richest are getting richer, while the poor/middle class are seeing wages and opportunities decline. We see it in the public markets, where the 7–10 biggest companies in the world keep growing in value, while all the rest are flat, if not down. We see it among VC firms, where the biggest are raising an increasingly disproportionate share of the money, while most smaller firms are shrinking or closing. And of course we see it among startups too: a small number of companies are getting pumped with money at insane valuations, while for the rest, it’s the toughest market anyone can remember.

That’s a phenomenon, but that phenomenon is driven by people’s decisions. What decisions? We’d argue that the widespread consensus decision of the moment is: “embrace the established, neglect the new.”

We see it in hiring. New grads are having the hardest time in the labor market:

We see it in asset allocation, with many LPs questioning why they should invest in Venture Capital at all, when the Nasdaq and S&P 500 performed better over the last decade:

Within VC, LPs and GPs have significantly shifted toward late stage/growth investing, shunning early stage. Even Jason Calacanis, who long touted his early stage prowess (did you know he was the third or fourth investor in Uber? 😉), has shifted to late stage:

If you stand back and look at it, it’s kind of a strange phenomenon. Have the established choices always been preferred? Of course not (or at least, not to this extreme). We are old enough to remember the articles from the late 2010s into 2021, celebrating early stage VC as the best asset class ever. Or the endless stories of entry level Computer Science grads with $200k starting salaries.

Will the established choices be preferred to this extreme in the future? We don’t believe so.

Setting aside the irony that everything established was once new, and that if we don’t support the new we will strangle the supply of the established in the future — we also don’t believe it’s rational, nor in market participants’ immediate economic interests, to so overwhelmingly prefer the established.

It’s particularly odd that this is the preference right now, just as we are experiencing the biggest technological revolution since the internet (and possibly bigger) in AI. If this new technology is going to re-orient the entire world, wouldn’t the young people be the best hires? Wouldn’t new startups, built around this new technology from the outset, be the best things to invest in? Shouldn’t the greatest opportunities be at the earliest stages? I.e. why has the majority of the venture market deemed OpenAI and Anthropic the main things worth investing in? And shouldn’t the small, nimble, and hungry VC firms be the best at identifying the best early opportunities? I.e. why have all the LPs crowded into A16Z, Sequoia, General Catalyst, and Thrive?

Maybe it’s fear, or uncertainty amidst chaos. No one really knows yet what AI means, how it affects the world, or how disruptive it will be. With so much change afoot, perhaps it feels safer to reach for the things that have been around for a while, the people you know, and the firms that have longer track records of performance. “No one ever got fired for buying IBM” as they say.

Or to put a finer point on it, people know there is tremendous change happening, but they don’t trust themselves to navigate it. So they say “Google seems to know what they are doing;” or A16Z; or “I’ve worked with Tom for 10 years, he’s always good at figuring it out;” or “everyone says Claude is the best, so I’ll just use that.” They outsource the adoption of the novelty to the established entities they trust.

On the investing side, recent performance is probably part of it too. Many agreed that early stage VC was the best asset class to invest in, leading up to and peaking in 2021; too much money poured in, valuations got crazy, too many competitors got funded, and the returns since then have been terrible. LPs raced for the exits (especially retail), which exacerbated the downturn further. The LPs that have stayed (especially institutional) have invested less and/or concentrated their allocations around the big firms they perceive to be safe. The market always overdoes it, to one extreme or the other.

So if the mass consensus decision right now is to embrace the established, we posit that the non-consensus move right now, and where the alpha lies, is in embracing the new.

If you are a hiring manager, now is probably the best time in a decade to hire new grads. They are desperate, hungry, will work harder for less money, and will be more loyal to you than ever. Most importantly, they will be the best able to implement AI in your organization, in novel ways that your experienced employees won’t think of.

If you are deciding which asset classes to allocate to, yes, the data shows that the public markets have outperformed average VC over the last decade. Does that mean the next decade will be the same? Of course not. In fact, it makes it more likely that the opposite will happen.

Every now and then we see one of these tables that ranks the performance of asset classes by year. They only show publicly traded assets (privates are opaque, hard to measure, and the marks lag). Usually, the intent is to demonstrate how unpredictable the market is, with the best-performing asset classes perpetually changing one year to the next:

What is striking in this current table, however, is how frequently Large Caps have outperformed this decade — at the top 4 times out of the last 7 years, and near the top in most of the other years too. Clearly, the biggest and most established companies have been the place to be.

But just as the best performing asset class changes from one year to the next, it also changes from one decade to the next. The prior decade did not have a single year in which Large Caps were the best:

In fact, the last time Large Caps outperformed, was 1998:

This feels like a surprising insight. Not only did Large Caps outperform back then, but they did so repeatedly, kind of like now. 3 years as number 1, and within the top 5 for 6 years in a row. Perhaps most interesting of all: it coincides with a period of similar technological disruption: the rise of the internet.

It appears that during the last period of significant technological change, the market’s initial reaction was to similarly buy the biggest, most established companies.

One might argue that this time is different, because today the biggest companies are all tech companies, and they weren’t back then. Well, here are the 10 most valuable companies in 1999:

• Microsoft: Led the global market during the height of the tech bubble.
• General Electric: A massive industrial and financial conglomerate.
• Cisco Systems: A high-flying networking hardware giant of the dot-com era.
• ExxonMobil: Formed from the 1999 merger of Exxon and Mobil.
• Walmart: The dominant American multinational retail corporation.
• Intel: The leading semiconductor chip manufacturer of the decade.
• Nippon Telegraph and Telephone (NTT): A major Japanese telecommunications firm.
• Lucent Technologies: The telecommunications equipment spin-off from AT&T.
• Nokia: The world leader in mobile phones at the turn of the millennium.
• IBM: Long-standing hardware and computing leader.

We count at least 7–8 tech companies here among the top 10. We’d argue General Electric was as much a tech company back then, as Apple is today. And even Walmart (which we didn’t count) was arguably as technologically interesting back then as Amazon is today. I.e. which part of Amazon is “tech?” That they sell stuff on the internet? Or that they have AWS, which launched in 2006? When was the last time a lot of these — especially Apple, Amazon, or Meta — launched something truly new?

In fact, are all of the Mag7 really “tech” companies anymore? Yes — they are big, impactful, and they have great businesses for the moment. But it’s hard to argue that they are more technologically innovative today, than the 10 biggest companies in 1999 were.

You might also argue that the P/E ratios of most of the Mag7 today are low, and therefore they remain good investments. Well, the P/Es of IBM and GE in 1999 were ~25x and ~30x — right in line with the P/Es of most of the Mag7 today (excluding Tesla). How have those done?

And by the way, P/E is only one part of the story. Free cash flow is another:

After the long period of outperformance in the 1990s, Large Caps then had 3 negative years in a row, and it took them 20 years to be the best performing asset class again. So as much as everyone loves the Large Caps right now (especially the Mag7), and it feels like they are destined to go up forever — we won’t be surprised to see the pendulum swing back in the other direction — just like it did after their last run. It’s impossible to time the market precisely, and it could be another year (or longer) before it happens. But no tree grows to the sky. Eventually, the market will turn and Large Caps will have a lost decade (or two) while other asset classes outperform — right when everyone counted the other asset classes out.

Zeroing back in on VC, the point is: VC will not always underperform. Yes, it’s had a rough few years, but it won’t last forever. And within VC itself, if you are choosing between late stage or early stage, late stage is the crowded side right now. 2 companies, OpenAI and Anthropic, absorbed 43% of all VC funding in the first half of this year (source), and 87.5% of all dollars went to rounds of $100M or more (source). In Q2 in particular, out of $137B of VC invested, only $4.9B went to seed or angel rounds — a mere 3.6% (source).

Early stage is empty. We have been surprised at how many early stage companies that would have raised easily in the past, are struggling to raise now — even those with exceptional teams, strong traction, and big market opportunities. It feels like the early stage market has finally come back to earth (for all but a handful of “hot” companies), and because of that we think the current vintage is going to deliver exceptionally good returns.

The “subsidized” mass excess buildout of AI — perhaps comparable to the fiber optic or railway booms of the past — is laying fertile ground for the next generation of startups to create and capture value in new ways. While Anthropic and OpenAI are currently loved by investors, the models from Google/Gemini, SpaceX/Grok, DeepSeek, Xiaomi, Zhipu, MiniMax, Moonshot, andAlibaba are not much worse. Because of this competition, combined with high CapEx, minimal moats, and low margins — we don’t believe these companies will deliver good investment returns. But the competition over performance and price *is* creating cheap and powerful tools for everyone else.

People love to say “in a gold rush, sell the picks and shovels.” But what if there is a picks and shovels rush? If a dozen+ companies are spending trillions of dollars, fighting to make the cheapest shovels that dig 1000x faster… maybe it’s better to be (and invest in) the gold miners.

AI is probably the highest leverage tool that has ever existed. In many ways, having access to LLMs for coding and other purposes is a great equalizer between entrepreneurs and large companies. There has never been a better time for small teams to accomplish huge goals. Thanks to the efficiency gains and automation opportunities brought about by AI, companies being started today can also remain lean as they scale.

We therefore believe that unpopular startups going after billion-dollar “niches” may have a better chance to capture long-term value than the large enterprises we read about in the news which operate in hyper-competitive industries with few moats. Our portfolio companies are building monopolies in domains that will have great profit margins 10 years from now precisely because these domains are currently underestimated by most market participants.

In the context of the current technological paradigm shift, it is worth remembering how Schumpeter articulated that creative destruction is an essential part of capitalism and the fundamental driving force behind its economic progress. Our portfolio is a diversified bet that this creative destruction will continue to propel young companies to the forefront. Especially in times of great upheaval such as the one we are going through at the moment.

“Embrace the Established, Neglect the New” was originally published in Unpopular VC on Medium, where people are continuing the conversation by highlighting and responding to this story.

UV LPs,

Welcome to our Q2 2026 LP Update. As a reminder, at Unpopular Ventures we focus on “the best companies, off the beaten path.” We invest in exceptional founders who are building high potential businesses, that are non-consensus in some way.

To help us find great opportunities around the world, we leverage a “Scout Program,” through which we share significant portions of our carried interest with anyone who helps us to identify, evaluate, and diligence companies that we invest in. If you would like to refer an opportunity, please check out our Scout Program Guidelines.

This is our quarterly update, which will be shorter than the annual update we write once a year. If you’d like to read about some of our most promising portfolio companies, how we do our portfolio tracking, or navigate to some of the past updates/content we’ve written, you can find our last Annual Update here.

By the way, we are going to begin posting more about our portfolio companies on X. If you’d like to keep up, please follow us there via this link. Now that a large part of our portfolio is getting more mature, there is starting to be more info that is actually public and worth sharing.

Thank you all for your continued partnership!

Sincerely,

Peter Livingston and Thibault Reichelt

Unpopular Ventures

We usually like to post our updates in the middle of the quarter, and this one is coming late. Sorry for the delay! We’ve had a flurry of investments in the last month that have been keeping us busy; a nice contrast from the lull we wrote about last quarter. For whatever reason, the flow of investable deals always ebbs and flows. When it’s dry it’s dry, and when it rains it pours.

I’ve also been a little busy personally over the last couple weeks, while I’ve been taking a hunting course in Sweden, which required quite a bit of study beforehand to pass the exam they require. I always love these kinds of experiences; it’s a complete departure from the world I usually inhabit and the things I usually think about (i.e. tech, AI, startups, macro trends, etc.). Engaging in this course, meeting people from completely different places and walks of life, and focusing on nature and being outdoors re-opened my perspective. As they say, sometimes it’s good to “touch grass.” As big, everywhere, and all-encompassing AI seems to be, I barely thought of it for those few days. Rest assured: there remains a wide world out there that is untouched by AI. Now that I’m done, I’m enjoying being back in my usual world with a fresh perspective.

Speaking of different worlds, Thibault and I were just discussing how it feels like we inhabit a different world of “Venture Capital” than the one that is in the news. All the mind share is focused on a small number of companies that are the most highly valued private companies in history, consuming extraordinary amounts of capital, along with a small number of “VCs” that are bigger than ever before and capturing a disproportionate share of the LP capital.

This is not the Venture Capital we grew up with. In times like this, it’s almost hard to remember that Kleiner Perkins Fund I was $4M (sure, that’s probably more like $20M inflation-adjusted — still small). Initialized Fund I was $7M. And Lowercase Capital I was $8M. These were among the best-performing VC funds in history, with Lowercase in particular returning something around 250x. That is what comes to mind when I think “Venture Capital.” Whatever this is today, with individual funds 1000 times bigger — I find incomprehensible. This is something different.

And really, it’s a luxury to not be that big. Josh Koppelman recently wrote a thoughtful post about what he termed “The Venture Arrogance Score:”

The gist is that the bigger your fund is, the higher ownership you need in a higher proportion of the most valuable companies created in that era. As he writes about the large funds today:

“So this hypothetical fund would need to capture $58B per year of value, which equates to about 30% of all value created during their investment window. That is hard to do. Very hard to do. And, in my opinion, would earn this hypothetical fund a high “Venture Arrogance Score”. I don’t believe there has been a fund group which has repeatedly captured more than 10% of the total value created by US founders… But maybe this time it’s different ;-)”

That’s not fun. Having to capture such a huge share of the market, and basically needing a corporation to do it… not for me. In my mind, VC is supposed to be fun; you take early risks on bold founders with big ideas, there is a great deal of uncertainty, and you ride a roller coaster of ups and downs. But when you get it right you really hit the jackpot, returning thousands or even 10,000 times your money. What the bulk of VC has become (capital-wise), is institutional. These mega funds have become the JP Morgans or Goldman Sachs of Venture Capital — massive teams, investing meaningful shares of their funds in only the most proven companies, and jostling for access rather than taking brave bets when no one else believes.

The uncertainty at the early stages, with small dollar amounts and lower valuations, is a big part of what makes VC interesting. I thought this was a great post from Bilal Zuberi:

As you can see, the most valuable company started in each vintage has been distinctly outside the hot theme of its time. Most of the money always chases what is hot, but the real alpha is in thinking outside the box, identifying what is going to be hot a few years from now.

But also, at our scale, we don’t even need to catch the single biggest company of the vintage (although I certainly wouldn’t mind if we did!). Our 2019 and 2020 portfolios have one small unicorn each (~$1–2B range so far), a few others that are successful but currently at lower valuations — and our multiples for those years are 5.5x and 7x already. There are plenty of companies started each year that go on to become unicorns; we just need one or two, and our returns will be excellent. Because there are so many bullseyes we can potentially hit, we can actually take the risks of betting on early, out there, and even unpopular opportunities. We don’t have to hit every target.

Sure, those juicy management fees at the big funds must be nice (and that is what the big funds appear to optimize for). But it’s so much more fun to play for the carry. As a small operation here at UV, we have confidence in our ability to regularly deliver high multiple vintages, and therefore get into meaningful carry. Which feels great, because it aligns us with you, our LPs. We only make real money when you do.

Being small, and investing so early, also feels safer in my opinion. I know that may be a controversial statement. Investing at pre-seed and seed is the riskiest time to invest in startups, and most of the individual companies do lose investors’ money. But I think that if you do it thoughtfully, build big diversified portfolios, are sensitive to valuation, and invest patiently with a long time horizon — it’s actually pretty hard to end up with bad returns. Indeed: a portfolio of early stage VC has historically tended to outperform large/mega fund VC — and really any other asset class for that matter.

Even though we are currently in a deep bear market for early stage VC, with the vast majority of seed firms downsizing or shutting down, our returns are looking just fine. Still above 20% IRR across our portfolio of 570+ portfolio companies, over the entire life of Unpopular Ventures going back through 2019, according to AngelList:

If this is indeed “the worst market anyone can remember” for early stage VC, and we are probably fairly representative of the market considering the size of our portfolio (albeit well above average, in our humble opinions) — this is pretty good. I love this asset class.

What does worry me however, is the late stage, highly valued startups that represent the bulk of the market in dollar terms. As I wrote in the last update, the recent “K-shaped economy” has led all the money to flow into a small number of places, and with that — a small number of companies have become the most highly valued startups in history and are absorbing unprecedented amounts of capital. The trend has been toward the big, and trends like this never last forever; at some point capital rotates and other market segments outperform — and I have no doubt that at some point the money will flow back to small caps — in both the public and private markets.

Here’s what I see on the horizon, and I’ll narrow it down to 3 companies specifically: OpenAI, Anthropic, and SpaceX.

For the first two, I’ve written in our last two updates about the thin moats and negative margins at the LLM layer of AI (i.e. I think the LLMs might end up like the airlines). We’ve now seen OpenAI lose its market dominance and growth has stalled. Anthropic has become the darling, but I think their moat is just as strong as OpenAI’s. We can all observe how Claude and ChatGPT have become worse over the last few months — as each has throttled performance, likely in an attempt to reach profitability heading into their IPOs. There is growing chatter about corporate customers beginning to cut their AI budgets, because the costs are high and they are not seeing the results they hoped for. With no switching costs, and more competition coming from Google, Microsoft, Meta, China, etc. — there will be a never ending competition over price and performance that is likely to keep margins always low (if not negative).

And with SpaceX, the desired valuation is extremely high relative to the fundamentals of the business. The vision is big, and Elon may pull another rabbit out of the hat — but there is also a chance the numbers will overwhelm the narrative this time. Time will tell.

Mechanically, I think the trigger for the capital rotation will be caused by concentration. In the case of SpaceX (and soon to be OpenAI/Anthropic), the commentary has mostly been about how much they will raise in the IPO, and whether the market will be able to absorb all of it. In my opinion, that is not the issue; the issue is how much supply there will be when the lockup expires, and how many forced sellers there will be. I thought this post from Meghan Reynolds was insightful:

If this is true (I think it probably is) that many large institutions and endowments have SpaceX ownership that represents 8–20% of their entire portfolios, they are going to be forced sellers. An individual can hold that and let it ride; but it goes against an institution’s fiscal responsibilities to hold so much concentration.

Stocks are partially valued based on the fundamentals of the businesses. But the price over the short term is also heavily influenced by flows in supply and demand. No matter how good a business, if there are more sellers than buyers, the stock price will decline.

There is likely more concentration in OpenAI/Anthropic ownership than in SpaceX, given that they are younger. Concentration gets distributed over time, but that’s the thing: it requires time for the ownership to rotate from one cohort of holders to the next. So much concentration coming to market at once I think may be the catalyst that breaks the upper end of the market.

In the case of SpaceX, if you look at it only in terms of the numbers, and imagine Elon was not the CEO — the valuation should be 10–20% of the $2T they are seeking (if not lower). If they IPO at that high price, and the market decides over the subsequent 6–12 months that the numbers trump the narrative, it could bring about an 80–90% stock price decline.

Same for OpenAI and Anthropic. If growth stalls (as it already appears to have for OpenAI) because enterprise customers are not seeing the value relative to price, and the margins remain negative — they too may end up with valuations that are a small fraction of what they are currently commanding — similarly leading to steep stock price declines post IPO. Remember: software is valued highly in large part because it is high margin. If a SaaS business with ~100% gross margins gets a 10x multiple on revenue, a business with long term expected margins of 10% (that is otherwise similar) should get a 1x revenue multiple.

Tying this back to safety and risk — these feel RISKY. I feel much more comfortable holding our portfolio of 570+ early stage startups, and continuing to add to it — than I would feel owning SpaceX, OpenAI, and Anthropic. It’s impossible to predict what will happen in the short term. But long term, I firmly believe that our portfolio will outperform those 3 large caps over the coming 5+ years.

PORTFOLIO STATISTICS

SUMMARY: $85.4M of capital invested has grown into $208.5M of portfolio value. This is up $2.6M vs. last quarter ($83M/$203.5M). As always, these numbers represent quite a bit of movement under the surface — both mark ups and mark downs that have been more or less balancing each other out. Here are the detailed numbers:

Our aggregate numbers, across every investment in UV’s life:

Numbers for only the UV Syndicate:

Annualized numbers for only the UV Rolling Fund:

Statistics about our investing activity:

We rolled over ~$9k of uninvested capital from Q1, and raised another $622k in Q2.

If you are a major LP of Unpopular Ventures, have invested at least $250k to date, and are willing to sign an NDA, we will share the complete portfolio data with you. Please submit a request via this form: link

For everyone else, you can access the de-identified data here: link

Thank you for reading, and thank you for your support of Unpopular Ventures!

UV Quarterly Update: Q2 2026 was originally published in Unpopular VC on Medium, where people are continuing the conversation by highlighting and responding to this story.

UV LPs,

Welcome to our Q1 2026 LP Update. As a reminder, at Unpopular Ventures we focus on “the best companies, off the beaten path.” We invest in exceptional founders who are building high potential businesses, that are non-consensus in some way.

To help us find great opportunities around the world, we leverage a “Scout Program,” through which we share significant portions of our carried interest with anyone who helps us to identify, evaluate, and diligence companies that we invest in. If you would like to refer an opportunity, please check out our Scout Program Guidelines.

This is our quarterly update, which will be shorter than the annual update we write once a year. If you’d like to read about some of our most promising portfolio companies, how we do our portfolio tracking, or navigate to some of the past updates/content we’ve written, you can find our last Annual Update here.

Thank you all for your continued partnership!

Sincerely,

Peter Livingston and Thibault Reichelt

Unpopular Ventures

LULL

You might have noticed: we haven’t been doing as many investments lately. This is the case for both the syndicate as well as our fund. We just haven’t seen as many opportunities that we’ve been really excited about.

It reminds me of my surfing days when I was younger. In my early 20s I surfed 3–4 days a week. Sometimes the conditions were great — lots of excellent waves to surf for days on end. And sometimes they were bad — with tiny waves or wind chop. During the lulls, I would still want to surf, and I’d spend hours driving up and down the coast trying to find a spot with decent waves; I’d find the least bad option and get in the water. But I’d always end up frustrated. I should have just gone to the gym and not wasted so much effort when there weren’t waves to surf.

It’s the same with investing. There are times when there’s a swell and the surf is up — it’s easy to find lots of great investments that we have confidence in. But sometimes there is a lull, and there is simply no way to force a bad market into good investments. Right now, we’ve just been in a bit of a lull.

There are probably a few reasons for this. As I shared in the last update, I’ve been feeling uncertain about AI. And considering that AI is the biggest thing happening in tech, I have been feeling a little uncertain about tech investing as a whole.

Dan Primack (Axios Pro Rata) recently summed up a lot of what I’ve been feeling:

That’s the strange juxtaposition of the current moment. Conventional software feels risky because a lot of it is being disrupted by AI. But: most of AI seems to lack both moats and margins. So where to invest?

Since I wrote last quarter’s post, OpenAI in particular seems to be teetering toward an increasingly precarious position. Their models have lost their competitiveness against Gemini and Claude, and Elon Musk’s lawsuit is growing increasingly credible. It seems almost unfathomable, but could you imagine what would happen if they went down? The wide ranging, and second order effects of $500B of value being erased would ripple through VC, their LPs, the entire tech industry, and maybe the world economy as a whole.

They aren’t the only one either. Many “hot” companies in AI seem to have built castles on sand — valued highly, but with questionable defensibility. Cursor is a recent example that makes me worried. They have been a super hot “consensus” company, rocketing to a $32B valuation in a short amount of time. And yet, based on what I’m reading, it sounds like Claude Cowork has made them nearly obsolete overnight. We’ve had some portfolio companies of smaller scale encounter similar difficulties — initial product/market fit, rapid growth, but then become obsolete a couple years later. Everything in AI is changing so fast that nothing is certain.

Even if these companies bounce back, the fierce competitiveness further highlights the “airline” analogy I used last quarter: air travel is tremendously valuable, everyone uses it, and it changed the way we live — but airlines are terrible businesses because of the low barriers to entry and low margins. If AI remains an arms race where everyone has to keep spending crazy amounts of money to remain competitive — always worried that a competitor could make them obsolete tomorrow — margins will stay negative (i.e. the lifetime revenue of a version will be outweighed by the cost to train it), none of these companies will ever generate a profit and the valuations will plummet.

Another thing on my mind is the “K-shaped economy.” I.e. there are a small number of people and companies that are doing better than ever, while everyone else is having a hard time. We see it at the individual level — more wealth inequality than ever before, the richest are getting richer, while the poor are seeing wages and opportunities decline. We see it in the public markets, where the 7–10 biggest companies in the world keep growing in value, while all the rest are flat, if not down. We see it among VC firms, where the biggest (esp. A16Z) are raising an increasingly disproportionate share of the money, while most smaller firms are shrinking or going out of business. And of course we see it among startups too: a small number of companies are getting pumped with money at insane valuations, while for the rest, it’s the toughest market anyone can remember.

We are seeing it on the ground ourselves. As an example, recent Y Combinator companies we invested in at the highest valuations, were ironically the first to be followed with a lot more money at even higher prices. Consensus investing is what has been working. Just ape into whatever everyone else thinks is good, and you will surely be followed soon after by others at a markup.

It has made our “Unpopular” strategy not work as well in recent years. We mostly like to focus on the non-consensus, off the beaten path opportunities — and those have produced outstanding returns for us in the past. But recently, those types of investments have not been getting as much love. A lot of our portfolio companies are doing great — but fewer of them have been getting the big and fast valuation markups we grew accustomed to a few years ago. When the money in the market was distributed more widely, it was easier to invest in something that was initially unpopular but then could flip to popular in a short amount of time. I.e. if ⅔ is unpopular and ⅓ is popular (for example), you can reasonably identify ones within the ⅔ that are likely to flip to the ⅓. But if 98% are unpopular and 2% are popular, it’s much harder for companies to flip. You have to wait longer for the overlooked companies to actually build their businesses and prove themselves to be good investments over the long term.

I don’t believe these dynamics will last forever; the pendulum of the market always swings back and forth. The consensus companies have to make money eventually, or they will die, and the herd mentality as a whole will be abandoned. In a tangential vein, I thought this was an interesting point:

What if AI spells the death of asset-light businesses and the value goes back to asset heavy? Remarkably, you don’t even have to go back 50 years to see a completely different orientation. Here are the 5 most valuable companies in 1990:

1. Exxon
2. General Electric
3. IBM
4. AT&T
5. Phillip Morris

None of those were asset light. Here are the 5 most valuable today:

1. Nvidia
2. Apple
3. Alphabet
4. Microsoft
5. Amazon

The top two, Nvidia and Apple, don’t even make their own chips or devices (TSMC does for Nvidia, and Foxconn does for Apple).

I’m not sure if we will actually return to heavy industry as the bulk of our economy, but it’s interesting to think about. VC has produced great returns for the last 50 years in large part because their most successful investments were able to scale rapidly without heavy CapEx. With so much disruption, there may be a move back toward the safety of businesses that make things you can feel and touch. Or at the very least, a heavier emphasis on durable moats.

I think the bigger mismatch in the current market is that very high valuations are being paid in cases where there remains a tremendous amount of uncertainty. $500B (or more) for OpenAI and $32B for Cursor — when there is a legitimate chance either could go to zero — feels uncomfortable to me. Yes, in VC you build a portfolio, but you still have to enter at a price where you are compensated for the risk you take. The valuations that most of the YC companies raise at on demo day similarly feel unreasonably high — unlikely to produce a positive return if you invested across all of them.

Everyone explains this as “just the market.” There are indeed investors out there who are paying these prices. But there are endless historical examples of the market behaving in temporarily irrational ways. Maybe the current exuberance in a small segment of the market represents the widespread fear of missing out on AI.

The point is: at the moment we are feeling a little uncertain, and because of that we are moving more slowly and investing more selectively than we have in the past. Charlie Munger wisely said: “It takes character to sit there with all that cash and do nothing. I didn’t get to where I am by going after mediocre opportunities.” That is more of our mindset right now.

Fortunately, we already have a large portfolio with lots of companies that are doing extremely well, and we *do* have confidence in the moats of many of our biggest ones. Interestingly, our most valuable companies also have some of the clearest moats — and I don’t think that’s a coincidence.

The companies we are investing in today, for the most part remain the off-path, non-consensus, “unpopular” opportunities — that we think are more reasonably valued. Doing something that no one else thinks is a good idea, or going where others refuse to go, is one of the best moats when you are just getting started.

Here are some more general moats that we think are likely to stand against AI:

• Network effects — i.e. the product becomes more valuable the more people use it. And one major branch of this is Marketplaces. AI brings down the cost of production, but people still need to exchange goods and services with one another. Our portfolio companies Yassir, Novig, Zepto, and Yummy all benefit from this.
• Anything with a bit of hardware. It’s harder to disrupt a business with AI when you have to reach into the real world. One of our top 3 most valuable investments (the stealth one) benefits from this.
• Regulatory and licensing barriers. Our portfolio company Jeeves benefits from this. It’s not straightforward to navigate the financial regulatory frameworks of so many different countries, and acquire unique sets of approvals in each of them.
• Proprietary data sets. Companies with access to unique, continually refreshed proprietary data will be able to leverage it in ways that AI can’t access.
• Things that involve people doing things in person. Our portfolio companies 222 (in person social network) and Sniffspot (marketplace for private dog parks) come to mind here.

This is just a start, and there are of course many more. The point is: it’s clear that AI is disruptive and advancing rapidly, so we are focusing more than ever on businesses with moats that are likely to withstand AI.

With all this in mind, please continue to send us the very best “unpopular” opportunities you come across — especially ones for which there are compelling moats. We have capital, and we are eager to invest. We are looking forward to breaking out of this momentary lull.

PORTFOLIO COMPANY NEWS

Once a year, in the annual update that we publish in August, we share a detailed list of our most promising portfolio companies. We additionally aim to share noteworthy news about the portfolio every 6 months, in our winter update (this one). So here’s some portfolio news that is top of mind.

First off, the way venture investing tends to go is that a small number of the investments end up being worth a disproportionate share of the portfolio. After enough time, most VCs and Angels find that a single investment usually ends up being worth more than everything else combined. This is true even if they have lots of successful investments! As an example, Chris Sacca’s Lowercase Fund I (widely considered the best VC fund of all time) had many famous investments, including Uber, Instagram, Twitter, and others. Even so, his Uber investment was worth more than everything else combined. My own first personal angel portfolio that I built between 2012 and 2018 had a similar skew.

Our Unpopular Ventures portfolio is only middle aged (and a lot of it young), so we don’t yet have a single dominant investment yet. But we do have 3 companies that currently constitute almost half the portfolio value — and combined are also worth more than all the money we have raised in UV to date. As a result, we tend to pay more attention to those. They are:

Jeeves: Jeeves is a global business bank that provides a suite of financial services for companies that operate in more than one country. We were their first investor on $10M post, and they last raised at a $2.1B valuation. They were a little quiet for a few years, as they navigated a challenging market environment — during which they focused on business fundamentals at the expense of growth. But now that they have found their footing, they are back on an exciting growth trajectory again. Here’s the latest public info from the Founder/CEO:

Yassir: the leading super app in Francophone Africa. We led their seed rounds on $15M and $35M pre-money caps in 2019 and 2020. The company employs over 4000 team members and serves customers in 45 cities across 7 countries. They have an enormous amount of revenue — more than any other portfolio company of ours — that is continuing to grow quickly.

They raised a round last year at a unicorn valuation. We have heard that other investors in the company are valuing their stakes well above that last mark — which we think is rational, considering that the company has continued to grow and comparable companies around the world are valued at much higher multiples of revenue. Having said that, we have not actually marked our position up yet, because our historical policy has been to only mark up our investments when there is an actual transaction. When the company’s next round happens, maybe this year, we expect to see a substantial markup.

Stealth: this is an investment that wishes to stay under the radar, so we can’t name them publicly here. Those who are invested in it know which it is. The company is extremely profitable and growing. They are generating more cash than they believe they can responsibly reinvest into growth, so they recently completed another tender offer in which they bought back shares from any investors who wished to sell — at a higher price than the last one. The valuation in this latest tender was 140 times the valuation at which we originally invested, and the share price was 96x net of dilution. We did not sell, because we believe the company still has a lot of upside potential from here.

And here’s some select news about the rest of the portfolio:

222 — In person social network that facilitates real-life encounters and experiences between strangers. Growing fast, recently raised a Series A led by Mark Suster at Upfront.

Backgammon Galaxy — The world’s largest online backgammon platform, offering competitive play, AI-powered analysis, and learning tools. New round at higher valuation.

Baton Market — Marketplace for buying and selling small businesses, providing data-driven valuations and end-to-end transaction support. Expecting over $10M in revenue this year.

Belvo — The leading open finance API platform in Latin America, enabling companies to access financial data and initiate payments. Starting to hit their stride, meaningful growing revenue.

Bon (fka Bhim) — AI-powered credit platform that helps customers understand, manage, and improve credit intelligently. Returned from near-death, pivoted, new round at a markup.

Brellium — AI-based clinical compliance platform that automates chart audits to ensure quality and payer compliance for healthcare providers. On a nice trajectory to become healthcare infrastructure.

Capi Money — International payments platform that helps businesses in Africa pay overseas suppliers quickly and affordably. 3.9x growth year over year.

CopilotIQ — Connected care platform delivering AI-driven remote health monitoring for older Americans with chronic conditions like diabetes and hypertension. Great progress and recently closed a new round at a markup.

Felix — Chat-based digital platform enabling cross-border payments and remittances, primarily between the U.S. and Latin America. $90M in annualized revenue, growing fast.

Fluently — AI-powered communication coach that helps non-native English speakers improve their spoken fluency in real time. We invested on $5M, recently raised at $60M, growing fast.

GitButler (fka Sturdy) — Modern Git client that lets developers work on multiple branches simultaneously, built for AI-heavy workflows. Series A from A16Z.

Hint Health — Operating system for direct primary care, providing membership management, billing, and EMR software for provider organizations. Doing very well, growing, and developing an increasingly dominant position in their space.

Kashin — Fintech platform offering micro-lending and financial services to the informal economy in Latin America. Doing great, significantly growing revenue.

Magic Mind — Functional beverage company producing a productivity-boosting health shot made with nootropics and adaptogens. Doing great, now the most popular health shot at Whole Foods.

Nelo — Consumer payments app offering buy-now-pay-later services for online purchases in Mexico. Forgot to mention this one in the last annual update, but they are doing well.

Novig — Peer-to-peer sports prediction market that lets users trade directly against each other rather than against the house. Growing extremely fast, recently raised a big round from top tier VCs at a $500M valuation.

Ouros Energy — Lithium-ion batteries that are 3–10x more energy dense than anything else on the market. Doing great, closing a Series A.

PAX Markets — Crypto exchange on a chip. Very excited about the company’s disruptive potential.

Pluto — Next gen investing for 100M underserved Europeans. Recent up round, growing.

QFEX — Bringing perpetual futures to traditional financial markets. Launching soon.

Rampa (fka Koban) — Financial platform serving businesses in Latin America with modern treasury and payment solutions. 10x year-over-year growth.

Remedial Health — B2B pharmaceutical marketplace connecting pharmacies with drug manufacturers and distributors across Africa. Ranked Africa’s fastest-growing healthtech company.

Revv — Technology platform for auto repair shops across North America. Now used by over 2,000 repair shops across North America.

Rio — Stablecoin platform for Latin America. Preparing its Series A.

Shekel Mobility — Digital platform powering vehicle financing, distribution, and fleet management across Africa. Now operating across several countries.

SIMPLETO — HR and payroll platform for businesses in Latin America, that is the rebirth of our prior portfolio company RunaHR. We had the opportunity to invest at an attractive valuation in their first round, and were recently followed by other investors at a higher price in their seed round.

Sniffspot — Marketplace for private dog parks. Continuing to grow steadily, with significant revenue now and profitable. Very likely to survive AI.

Spin — AI-based SaaS data protection platform which monitors and backs up G Suite, Microsoft 365, and Salesforce data. Found a great Series B partner.

Stepful — Online training for healthcare workers, offering accelerated certification programs for roles like medical and pharmacy technicians. Continued strong execution.

Supercede — Reinsurance technology platform that streamlines the interaction between brokers and reinsurers. New round at a higher valuation.

Sygaldry — Bringing quantum technology to data centers. We invested at YC demo day, recently raised at a much higher price.

Taiv — AI-powered ad-tech platform that replaces generic TV commercials in bars and restaurants with targeted, relevant content. Growing fast, recently closed a new round.

Undermind — AI research engine. Already profitable.

Vibrant Health — AI-native operating system for personalized medicine. We invested in their pre-seed, and were recently followed by Phin Barnes at The General Partnership in their Seed round.

Vitalize Care — Transforming how health systems optimize labor. Big Series A and markup for us from top tier VCs.

Volantis — Semiconductor company that designs and manufactures AI chips optimized for high-performance computing needs. Making good progress, and raised more money from top investors, including Naval Ravikant.

WiseLayer — AI-powered digital workers for finance and accounting teams. Was recently acquired by Blackline in a positive outcome for us.

Yummy — Venezuela’s leading super app. Doing great, on a strong growth trajectory again. We speculate that the recent political change in Venezuela expands their potential over the coming years. As you might have seen, we have been interested in investing further here.

Zepto — India’s fastest-growing quick commerce platform, delivering groceries and essentials in minutes. The company is expected to go public this summer at a $7–8B valuation. We invested at $13M post, just 5 years ago.

PORTFOLIO STATISTICS

SUMMARY: $83M of capital invested has grown into $203.5M of portfolio value. This is up ~$1.5M vs. last quarter ($82M/$201M). As always, these numbers represent quite a bit of movement under the surface — both mark ups and mark downs that have been more or less balancing each other out. Here are the detailed numbers:

Our aggregate numbers, across every investment in UV’s life:

Numbers for only the UV Syndicate:

Annualized numbers for only the UV Rolling Fund:

Statistics about our investing activity:

We rolled over $109k of uninvested capital from Q2, and raised another $715k in Q3.

If you are a major LP of Unpopular Ventures, have invested at least $250k to date, and are willing to sign an NDA, we will share the complete portfolio data with you. Please submit a request via this form: link

For everyone else, you can access the de-identified data here: link

Thank you for reading, and thank you for your support of Unpopular Ventures!

UV Quarterly Update: Q1 2026 was originally published in Unpopular VC on Medium, where people are continuing the conversation by highlighting and responding to this story.

I am long term bullish, but short term bearish on the broad category. By long term I mean 10–20+ years, and by short term I mean the next 3–7 years.

Just about everyone is bullish on AI to some degree or another. What it can do today is already amazing, and the visions of what it could do in the future are extraordinary. It is changing how we work, learn, and interact with the world in endless ways.

So how could I possibly be bearish on any time horizon? First, a framework I frequently lean on is something called the Gartner Hype Cycle.

Every technology wave through all of history, as far as I know, has followed this pattern. A new technology is born, people recognize the potential, and they all rush in with money and talent, valuations go up, and everyone gets excited — forming the “peak of inflated expectations.” But then something happens, where it’s harder to build, or it takes longer than people expect, and the initial excitement fades — the money flows out and people temporarily lose hope — forming the “trough of disillusionment.” Then gradually over time, that original vision gets built, and reaches or even exceeds the expectations that everyone had in the first place.

A widely familiar example of this is the dot com boom and bust. All the people who got excited about the internet in the 1990s were exactly right; it just took longer than they expected. There was an initial boom, then a crash that lasted a few years, after which it all came to fruition. Pets.com and Webvan are the poster children of excess from that time, they died, and yet Chewy and Instacart today are wildly successful and worth orders of magnitude more than those initial attempts were ever valued at. Amazon’s stock price ripped 71x from IPO to peak in 1999, dropped 95% in the early 2000s, and then we all know what it did in the 20 years after.

I am not aware of a single historical technology wave that skipped the trough of disillusionment. If you know of a technology cycle that only went up and to the right — please tell me, I’d love to learn.

Bringing this to AI right now: I think we are somewhere close to the peak of inflated expectations. It’s impossible to time markets, and the ascent could go on longer than anyone expects. But sooner or later, I think we will encounter the inevitable trough of disillusionment.

The specific flaws this time, that I think will precipitate the trough, are 1) the thin margins, and 2) the minimal moats.

Regarding point 1) most are valuing AI as “software,” which is known for its high margins (usually 90%+) and because of those high margins software companies are valued at a higher multiple of revenue than most other businesses. It’s easier to generate profits from them, or recycle a large amount of that revenue to propel future growth.

But AI is low margin — at both the LLM and application layers — if not negative margin in many cases today. If you listen closely to the mainstream commentary about AI these days, there is almost zero discussion about profitability, now or in the future. People just assume it will be like other tech, where you burn a lot in the beginning and the profits will arrive later.

I think this will be harder than usual to get meaningfully profitable. The moment that investors begin to really scrutinize the margins and profitability — especially public market investors — I think there will be a reckoning.

The second, related component, is the minimal moats. All the most successful tech companies are monopolies with strong moats. I don’t see any meaningful moats in AI yet. The many LLMs all seem pretty comparable, and if that doesn’t change, they will all have a hard time charging prices that produce meaningful margins. Furthermore, China is subsidizing its AI companies, and Meta, Google, xAI, and Microsoft all have cash flows that allow them to afford to lose billions in AI indefinitely. There will be a never ending war over price and performance; I don’t see how anyone makes a profit here for a long time.

In fact, I wonder if the LLM layer of AI may turn out to be similar to the airline industry. Air travel is a huge deal:

• It provides tremendous value to society.
• Everyone uses it.
• It changed the way we live.

Kind of like what we expect from AI. And yet, the airlines are terrible businesses — due to the thin margins and competitive dynamics. Taking the analogy further, supplier concentration is a problem; they all have to buy their airplanes from Boeing or Airbus, and jet engines from GE or Rolls Royce. Reminds me of how all the LLMs have to buy their chips from Nvidia.

So it’s possible that AI will be amazing and change the world and do everything we all expect… and also not make much money.

Another parallel here might be search engines. If you remember, in the 1990s there were 20 different search engines, indistinguishable from each other, until Google came along with a better algorithm. When that became clear, everyone flocked to Google, they captured the market, got the monopoly, and all the others went to zero. Right now, all these LLMs feel like the 20 search engines that were all the same. I’m looking for the Google that finally figures out the better algorithm; if you identify it, please let me know. Maybe it will be whoever actually reaches AGI, although I’m currently skeptical that we will get there any time soon.

I’d much rather have invested in Google’s IPO in 2004 at a $27B valuation, than in the 20 different search engines in the 1990s. Same for Amazon in 2002/2003, when it was down 95% from peak. If you can think independently from the herd, and identify the solid companies with real moats during the trough of disillusionment, that is the best time to invest. The prices are low and the winners are clear(er).

Although it’s impossible to time the exact top, I do wonder if we may have seen the “emperor has no clothes” moments in the last couple weeks — that we will all look back on as the wake up call. First, the interview between Sam Altman and Brad Gerstner. For those who didn’t see it, Brad asked Sam a reasonable question, paraphrased: “You have $1.4 Trillion of spending commitments and $13B of revenue. How are you going to pay for all this?” Sam blew up at him and basically said, “you should sell your shares.”

The exchange reaffirmed my belief that the money doesn’t make sense right now, and the only way this works out for OpenAI is if they can keep the music playing. Sam doesn’t have answers to the obvious money questions. All he can do is keep everyone hyped up so they keep giving him more money at higher and higher prices.

The second moment that stood out to me was the news that OpenAI has been lobbying the federal government to backstop their loans, trying to make the case that they are “too big to fail.” If they are such a good business, why do they need government backing? This news further reaffirmed my belief that there is no moat. Sam is scrambling for something, anything, that will give him a moat, including special protections from the government. He has tried in the past under the guise of “AI safety” — trying to get the government to sanction a limited number of AI companies that they regulate and make sure are “safe” — boxing out others. Getting special financial treatment from the government is his latest attempt.

This was the first time it’s been made so clear (at least to me): the money doesn’t make sense, and there is no moat.

So how to play this?

First, keep a cool head. Don’t follow the herd and rush to invest in those triple layer OpenAI SPVs that are going around. Just because an area is hot, exciting, and everyone is talking about it — does not mean it’s a good investment. In fact, usually the thing in the news is a terrible investment (unless the news is claiming the thing is dead — then it’s often the opposite).

Second, look for the moats. I’ve painted AI with a broad brush above, and I do legitimately think the bulk of the market (especially the LLMs) is a bad investment right now. But at the same time, I have always been a believer in investing in specific companies, not broad categories — and I do think there are specific companies that represent good investments right now. There always are. Because I think the Achilles Heel in a lot of AI is the lack of moats, moats are the main thing I’m looking for in individual companies to invest in. Moats are easier to build when you venture off the beaten path, or focus on highly specific or obscure markets that most people never think about.

Third, remember that AI can be part of a business, but it doesn’t have to be the most interesting or important part of the business. AI doesn’t make a business successful; its customers, and the unique way it serves them does. Austen makes a good point here:

In other words, AI is not a magic wand. It’s simply the latest tool in the toolkit, much like the internet or computers were in past eras. No one says they are an “internet company doing X” anymore — that would be silly. If everyone is using AI in one way or another, then it’s not a differentiator. It follows that investing blindly in anything AI today is comparable to investing in anything with a .com in the 1990s. The point is: the success of a business is about so much more than whether or not it has something to do with AI.

Fourth, make sure to always keep some dry powder. As everyone says in a downturn, “be greedy when others are fearful.” But that only works if you also remembered the other half of it: “be fearful when others are greedy.” I.e. you only get to be greedy if you have the cash to do it — which comes from taking some chips off the table near the highs. There’s a reason financial advisors recommend a 60/40 stock/bond portfolio (or 70/30 if you want to be aggressive); those bonds come in handy when other investments get cheap.

Lastly, remember to keep a clear mind when the tide goes out. There will come a time when most have lost money, companies are going bust left and right, and the mainstream media will become black bearish on AI. The consensus will be that AI companies don’t make money. When that happens, if you can ignore the noise and look closely, it will be clearer than ever which ones do make money, have something of a moat, and will be here to stay. That will be the time to buy. Who knows — maybe you’ll have a chance to buy the next Amazon at a 95% discount.

Thoughts on AI was originally published in Unpopular VC on Medium, where people are continuing the conversation by highlighting and responding to this story.

Welcome to our Q4 2025 LP Update. As a reminder, at Unpopular Ventures we focus on “the best companies, off the beaten path.” We invest in exceptional founders who are building high potential businesses, that are non-consensus in some way.

To help us find great opportunities around the world, we leverage a “Scout Program,” through which we share significant portions of our carried interest with anyone who helps us to identify, evaluate, and diligence companies that we invest in. If you would like to refer an opportunity, please check out our Scout Program Guidelines.

This is our quarterly update, which will be shorter than the annual update we write once a year. If you’d like to read about some of our most promising portfolio companies, how we do our portfolio tracking, or navigate to some of the past updates/content we’ve written, you can find our last Annual Update here.

Thank you all for your continued partnership!

Sincerely,

Peter Livingston and Thibault Reichelt

Unpopular Ventures

THOUGHTS ON AI

A few LPs have asked me (Peter) about my thoughts on AI, so I thought I’d share what’s in my mind:

I am long term bullish, but short term bearish on the broad category. By long term I mean 10–20+ years, and by short term I mean the next 3–7 years.

Just about everyone is bullish on AI to some degree or another. What it can do today is already amazing, and the visions of what it could do in the future are extraordinary. It is changing how we work, learn, and interact with the world in endless ways.

So how could I possibly be bearish on any time horizon? First, a framework I frequently lean on is something called the Gartner Hype Cycle.

Every technology wave through all of history, as far as I know, has followed this pattern. A new technology is born, people recognize the potential, and they all rush in with money and talent, valuations go up, and everyone gets excited — forming the “peak of inflated expectations.” But then something happens, where it’s harder to build, or it takes longer than people expect, and the initial excitement fades — the money flows out and people temporarily lose hope — forming the “trough of disillusionment.” Then gradually over time, that original vision gets built, and reaches or even exceeds the expectations that everyone had in the first place.

A widely familiar example of this is the dot com boom and bust. All the people who got excited about the internet in the 1990s were exactly right; it just took longer than they expected. There was an initial boom, then a crash that lasted a few years, after which it all came to fruition. Pets.com and Webvan are the poster children of excess from that time, they died, and yet Chewy and Instacart today are wildly successful and worth orders of magnitude more than those initial attempts were ever valued at. Amazon’s stock price ripped 71x from IPO to peak in 1999, dropped 95% in the early 2000s, and then we all know what it did in the 20 years after.

I am not aware of a single historical technology wave that skipped the trough of disillusionment. If you know of a technology cycle that only went up and to the right — please tell me, I’d love to learn.

Bringing this to AI right now: I think we are somewhere close to the peak of inflated expectations. It’s impossible to time markets, and the ascent could go on longer than anyone expects. But sooner or later, I think we will encounter the inevitable trough of disillusionment.

The specific flaws this time, that I think will precipitate the trough, are 1) the thin margins, and 2) the minimal moats.

Regarding point 1) most are valuing AI as “software,” which is known for its high margins (usually 90%+) and because of those high margins software companies are valued at a higher multiple of revenue than most other businesses. It’s easier to generate profits from them, or recycle a large amount of that revenue to propel future growth.

But AI is low margin — at both the LLM and application layers — if not negative margin in many cases today. If you listen closely to the mainstream commentary about AI these days, there is almost zero discussion about profitability, now or in the future. People just assume it will be like other tech, where you burn a lot in the beginning and the profits will arrive later.

I think this will be harder than usual to get meaningfully profitable. The moment that investors begin to really scrutinize the margins and profitability — especially public market investors — I think there will be a reckoning.

The second, related component, is the minimal moats. All the most successful tech companies are monopolies with strong moats. I don’t see any meaningful moats in AI yet. The many LLMs all seem pretty comparable, and if that doesn’t change, they will all have a hard time charging prices that produce meaningful margins. Furthermore, China is subsidizing its AI companies, and Meta, Google, xAI, and Microsoft all have cash flows that allow them to afford to lose billions in AI indefinitely. There will be a never ending war over price and performance; I don’t see how anyone makes a profit here for a long time.

In fact, I wonder if the LLM layer of AI may turn out to be similar to the airline industry. Air travel is a huge deal:

• It provides tremendous value to society.
• Everyone uses it.
• It changed the way we live.

Kind of like what we expect from AI. And yet, the airlines are terrible businesses — due to the thin margins and competitive dynamics. Taking the analogy further, supplier concentration is a problem; they all have to buy their airplanes from Boeing or Airbus, and jet engines from GE or Rolls Royce. Reminds me of how all the LLMs have to buy their chips from Nvidia.

So it’s possible that AI will be amazing and change the world and do everything we all expect… and also not make much money.

Another parallel here might be search engines. If you remember, in the 1990s there were 20 different search engines, indistinguishable from each other, until Google came along with a better algorithm. When that became clear, everyone flocked to Google, they captured the market, got the monopoly, and all the others went to zero. Right now, all these LLMs feel like the 20 search engines that were all the same. I’m looking for the Google that finally figures out the better algorithm; if you identify it, please let me know. Maybe it will be whoever actually reaches AGI, although I’m currently skeptical that we will get there any time soon.

I’d much rather have invested in Google’s IPO in 2004 at a $27B valuation, than in the 20 different search engines in the 1990s. Same for Amazon in 2002/2003, when it was down 95% from peak. If you can think independently from the herd, and identify the solid companies with real moats during the trough of disillusionment, that is the best time to invest. The prices are low and the winners are clear(er).

Although it’s impossible to time the exact top, I do wonder if we may have seen the “emperor has no clothes” moments in the last couple weeks — that we will all look back on as the wake up call. First, the interview between Sam Altman and Brad Gerstner. For those who didn’t see it, Brad asked Sam a reasonable question, paraphrased: “You have $1.4 Trillion of spending commitments and $13B of revenue. How are you going to pay for all this?” Sam blew up at him and basically said, “you should sell your shares.”

The exchange reaffirmed my belief that the money doesn’t make sense right now, and the only way this works out for OpenAI is if they can keep the music playing. Sam doesn’t have answers to the obvious money questions. All he can do is keep everyone hyped up so they keep giving him more money at higher and higher prices.

The second moment that stood out to me was the news that OpenAI has been lobbying the federal government to backstop their loans, trying to make the case that they are “too big to fail.” If they are such a good business, why do they need government backing? This news further reaffirmed my belief that there is no moat. Sam is scrambling for something, anything, that will give him a moat, including special protections from the government. He has tried in the past under the guise of “AI safety” — trying to get the government to sanction a limited number of AI companies that they regulate and make sure are “safe” — boxing out others. Getting special financial treatment from the government is his latest attempt.

This was the first time it’s been made so clear (at least to me): the money doesn’t make sense, and there is no moat.

So how to play this?

First, keep a cool head. Don’t follow the herd and rush to invest in those triple layer OpenAI SPVs that are going around. Just because an area is hot, exciting, and everyone is talking about it — does not mean it’s a good investment. In fact, usually the thing in the news is a terrible investment (unless the news is claiming the thing is dead — then it’s often the opposite).

Second, look for the moats. I’ve painted AI with a broad brush above, and I do legitimately think the bulk of the market (especially the LLMs) is a bad investment right now. But at the same time, I have always been a believer in investing in specific companies, not broad categories — and I do think there are specific companies that represent good investments right now. There always are. Because I think the Achilles Heel in a lot of AI is the lack of moats, moats are the main thing I’m looking for in individual companies to invest in. Moats are easier to build when you venture off the beaten path, or focus on highly specific or obscure markets that most people never think about.

Third, remember that AI can be part of a business, but it doesn’t have to be the most interesting or important part of the business. AI doesn’t make a business successful; its customers, and the unique way it serves them does. Austen makes a good point here:

In other words, AI is not a magic wand. It’s simply the latest tool in the toolkit, much like the internet or computers were in past eras. No one says they are an “internet company doing X” anymore — that would be silly. If everyone is using AI in one way or another, then it’s not a differentiator. It follows that investing blindly in anything AI today is comparable to investing in anything with a .com in the 1990s. The point is: the success of a business is about so much more than whether or not it has something to do with AI.

Fourth, make sure to always keep some dry powder. As everyone says in a downturn, “be greedy when others are fearful.” But that only works if you also remembered the other half of it: “be fearful when others are greedy.” I.e. you only get to be greedy if you have the cash to do it — which comes from taking some chips off the table near the highs. There’s a reason financial advisors recommend a 60/40 stock/bond portfolio (or 70/30 if you want to be aggressive); those bonds come in handy when other investments get cheap.

Lastly, remember to keep a clear mind when the tide goes out. There will come a time when most have lost money, companies are going bust left and right, and the mainstream media will become black bearish on AI. The consensus will be that AI companies don’t make money. When that happens, if you can ignore the noise and look closely, it will be clearer than ever which ones do make money, have something of a moat, and will be here to stay. That will be the time to buy. Who knows — maybe you’ll have a chance to buy the next Amazon at a 95% discount.

PORTFOLIO

SUMMARY: $82M of capital invested has grown into $201M of portfolio value. This is roughly flat vs. last quarter ($81M/$199M — so ~$1M of appreciation). As always, these numbers represent quite a bit of movement under the surface — both mark ups and mark downs that have been more or less balancing each other out.

I am cautiously optimistic that this coming year we will finally start to see meaningful movement to the upside again. Several of our most valuable investments are doing very well under the hood, growing revenue substantially, expanding their market shares, inking big new deals — and in many cases we believe the official “marked” valuations undervalue the positions. I expect that we’ll see new rounds happen over the course of the next year that will bring some of these winning positions closer to their true values.

As I’ve discussed in past updates, it’s been a weird market the past few years — where there has been both a lot of death/destruction on the downside, and few follow on rounds to the upside. It’s a contrast from 2019–2021, when the growth stage capital was flowing freely, at high valuations, and many of the weaker companies were kept alive. I.e. before, the good companies went up and the weak companies stayed flat; now, most of the good companies stay ~flat and the weak companies go down. All those markups felt nice before, and it’s frustrating to not get the same “credit” we used to get for our good investments — but I am confident that it will all work out in time. We aren’t looking to sell any of these positions now anyway; we aim to hold for 10 years. By the time we get closer to the point of exiting our positions, I am optimistic that our winning investments will converge to their true values (which of course are continuously increasing as well). The weak companies can only die once, but there is no limit to how far the winning companies can grow.

If you would like to read about some of our most promising portfolio companies, please see our last Annual LP Update. These don’t change often enough to be worth talking about every quarter. And as usual, we are sharing the latest statistics on the overall portfolio below.

PORTFOLIO STATISTICS

Our aggregate numbers, across every investment in UV’s life:

Numbers for only the UV Syndicate:

Annualized numbers for only the UV Rolling Fund:

Statistics about our investing activity:

We rolled over $276k of uninvested capital from Q2, and raised another $810k in Q3.

If you are a major LP of Unpopular Ventures, have invested at least $250k to date, and are willing to sign an NDA, we will share the complete portfolio data with you. Please submit a request via this form: link

For everyone else, you can access the de-identified data here: link

Thank you for reading, and thank you for your support of Unpopular Ventures!

UV Quarterly Update: Q4 2025 was originally published in Unpopular VC on Medium, where people are continuing the conversation by highlighting and responding to this story.

To The Unpopular Ventures Community,

Welcome to our 2025 Annual Update. As a reminder, at Unpopular Ventures we focus on “the best companies, off the beaten path.” We invest in exceptional founders who are building high potential businesses, that are non-consensus in some way.

To help us find great opportunities around the world, we leverage a “Scout Program,” through which we share significant portions of our carried interest with anyone who helps us to identify, evaluate, and diligence companies that we invest in. If you would like to refer an opportunity, please check out our Scout Program Guidelines.

This is our annual update, which will be slightly longer than our quarterly updates. In addition to an update on our portfolio, you’ll also find an expression of gratitude to our community, a renewed case for why we believe UV is a good place to invest, disclaimers and an explanation about how we do our portfolio tracking, some of our current top performing investments, and links back to some of our other past content.

Thank you all for your continued partnership!

Peter Livingston and Thibault Reichelt

Unpopular Ventures

1. GRATITUDE

We have a lot to be thankful for, so we’ll start with a bit of gratitude. Thank you to:

OUR PORTFOLIO FOUNDERS. Thank you to the Founders and teams of all 536 portfolio companies of ours (past and present) — who allowed us to invest in their companies, and have worked like crazy to build their dreams into realities. Great Founders have endless choices of investors to take money from. Thank you for choosing us, and for allowing us to ride along as a small part of your journeys.

OUR LPS. Thank you to all of our 5400+ syndicate LPs and 100+ Rolling Fund LPs who have entrusted us with your capital. There are endless places to invest your money: dozens of asset classes, thousands of VC firms, and hundreds of syndicates and funds just on AngelList to choose from. Thank you for choosing us.

OUR SCOUTS, GUEST LEADS, AND VENTURE PARTNERS. As you know, we share carry with anyone in our network (including LPs, Portfolio Founders, and friends) — for their assistance in identifying, assessing, and negotiating one or more of our investments. Thank you to our extended family of contributors who have helped us do more than we would have been able to on our own. There are hundreds of you — you know who you are — and we are grateful to you.

THE ANGELLIST TEAM, who make the magic happen behind the scenes. Special thanks in particular to Jack Plank and Colton Smith — our managers at AngelList who we worked more closely with over the last year, who have been fantastic. An extra big thank you to Avlok Kohli, CEO at AngelList as well; we had a complex situation with one of our investments recently, and he really stepped up and helped us in a big way.

2. WHY INVEST WITH UNPOPULAR VENTURES?

It’s been a couple years since I (Peter) last made the case for why I believe UV is a good place to invest, so I thought I’d make the case anew. There are 4 main reasons: 1) skin in the game, 2) track record, 3) experience, and 4) timing.

First and foremost: skin in the game. As the Founder and Managing Partner of UV, I put my own money where my mouth is. I invest more in our own fund and syndicate than I do anywhere else: $100k per year of personal cash into our fund, and more into most of the syndicate deals we do too. I am the 5th biggest LP in UV’s fund and syndicate ($1.25M invested to date). I would like to invest more, but I am already extremely over-allocated to venture investments (irresponsibly so — most financial advisors would have a heart attack if they saw my personal portfolio) — so I have to be cautious about managing my liquidity.

Second, track record: both I and Thibault have been investing in startups for a long time, well before UV — and we have a plethora of winning investments.

Even before I started investing, I was the first employee and first engineer at a company that today has a $5B market cap (iRhythm). I later began my investing career as a solo angel, from 2012–2018, investing my own money. My first personal angel “fund” (with only my own money) has returned over 15x to date. The single best investment in that fund was made in 2015 and has returned 472x net. It was my success investing as a solo angel for so many years that gave me the confidence to start UV and invite all of you to invest with me.

My Partner, Thibault, also has an exceptional track record, including before joining me in UV. He started with some small investing as an angel in 2016, and those personal investments have already returned a realized multiple of over 5x. Since 2017 he has been making venture investments on behalf of a family office. He has done so well with them that he still invests on their behalf in addition to UV, and that family office is one of the biggest LPs across the AngelList platform. Although Thibault doesn’t have the same level of liquidity as I have, and has not been able to invest as much in UV personally, the family office that he works with is UV’s 8th biggest LP.

We started UV in 2019, and have an excellent track record here too. Our most mature vintages, 2019 and 2020, are already up 5.3x and 6.8x gross — and we expect them to keep growing much further from here. AngelList reports our aggregate IRR across the entire life of UV at 24%:

Note that the calculations behind this are not perfect — there are some investments that should be marked less, and some that are worth more (for example, Yassir’s new round has not been included — which will be a big markup for us) — but for the purpose of an apples to apples comparison with any other syndicate or fund on AngelList — this number is close to accurate.

The number has been higher in the past (even above 100% in 2021/2022), but as you likely know — the VC market has been in a down cycle the last few years and everything has slowed down. I am proud that we still have an IRR this high, at this time — which I believe is close to the bottom of this bear market. I’ll come back to this in part 4.

Third, experience. Thibault and I have experience well beyond just our winning investments, which informs our judgement today. We have both been involved with numerous startups as operators — and because of that we have deep insight into what these companies are like to build on the inside.

My personal “anti portfolio” though, is the single most valuable component of my experience. It shaped our firm’s investing strategy, and was even the inspiration for our name: Unpopular Ventures.

As I wrote in one of our recent syndicate memos, “The essence of our name and Unpopular thesis is: invest in the very best founders we can, no matter what they are doing, and *especially* if they are doing something unusual or off the beaten path. All of my biggest investing mistakes were cases when I did not invest in someone I knew was exceptional, merely because I didn’t like or understand their idea. And our biggest wins in UV have similarly all been in non-obvious ideas, markets, and geographies — but were nonetheless pursued by exceptional founders. That is the key to successful startup investing in our opinion: bet on people, not ideas. And in fact — the most unpopular ideas frequently turn out to be the best ones.”

I thought you might appreciate the color behind that, so I’ll share 5 stories.

First, the thing that made me begin angel investing was an investment I did not do in my friend’s company, Burb’n. Kevin Systrom was my fraternity brother at Stanford undergrad, and a good friend after. He and I worked down the hall from each other (at different startups) in 2007–2009 and would occasionally meet for lunch. A couple years later, I heard he was building an app called Burb’n, and was trying to raise a little money. I downloaded the app and played around with it, but I didn’t think the idea was compelling — so I dismissed it and did not reach out to him. I easily could have, and boy do I wish I had. Shortly after, he pivoted, renamed the company Instagram, and sold for $1B 18 months later (in Facebook stock at a $100B valuation — which is up another 19x since then). That was when I realized: I should invest in my friends.

The next friend who started a company was another fraternity brother, close friend, and my roommate in San Francisco for 3 years, Rainer Castillo. One day, he told me he and his co-founders were making men’s short shorts and selling them online. It was the worst idea I had ever heard. But clearly I was wrong about Burb’n, and I believed in Rainer, so I grimaced and invested a little bit. That company was Chubbies, which today is a leading men’s wear brand and exited for >$100M in cash. Not quite an Instagram-level outcome, but still a very nice win.

Based on my early success with Chubbies, and some other startups too, I waded into angel investing more and more. But I started out investing very carefully and selectively — and that turned out to be a big mistake.

David Velez was a close friend from my first year of business school. He and I later worked down the road from each other when he was a junior partner at Sequoia Capital, and I was an EIR at Kleiner Perkins — and we would occasionally meet for lunch — so I knew him well. A couple years later, I learned that he was trying to build a credit card business in Brazil. I thought “Brazil? That sounds super risky” — so I did not ask him about investing. Again, I easily could have. That company was Nubank and today has a >$60B market cap.

I guess I needed one more miss to drive the point home, because Evan Moore and Tony Xu were good friends of mine in my second year of business school. We were even co-presidents of the Stanford GSB Venture Capital Club all together. One day, Evan told me he was experimenting with an idea in food delivery. “Food delivery?” I thought, “that’s not a sexy business, and probably low margin too.” Whelp, that company was DoorDash, which today has a >$100B market cap.

Dang. If only I had invested in these friends that I knew were smart, even though I didn’t know anything about photo sharing, Brazilian banking, or food delivery. At last, I finally learned the lesson: invest in the smartest people I can, regardless of what I think of their ideas. Maybe even especially if I think the ideas are bad.

I also learned: the worst thing you can do in startup investing is not losing your money on an investment; it’s not investing in something that returns 1000x+. I would have gladly lost my money hundreds of times, if only I had invested in Instagram, Nubank, or DoorDash. This is why we invest in more companies — and take more risk — than most VCs. We have deeply internalized that the way to get the best returns is to maximize the chance that we invest in an extreme outlier — even if that means we are wrong a lot.

Fortunately, it wasn’t all misses, because in 2015 I made the investment I mentioned above that has returned 472x to date, and delivered most of my 15x personal angel fund return. Those of you who have been with us a while know what this investment was; if you don’t know and would like to, send me a note and I will tell you. Interestingly, I told everyone I knew about this investment in 2015 — well over 100 people — and in the end I was only able to convince 2 to invest relatively small amounts. My best investment was ironically also the hardest one to convince others to do. Unpopular again.

Turns out, this theme is not unique to my experience: Airbnb was famously rejected by over 80 VCs, and Google attempted to sell themselves for $1M to Excite, who turned them down. Most of the best investments are unpopular in the beginning.

The point is: I’ve been out there a long time; I’ve seen winners and losers, what they looked like at the earliest stages — and many more than I’ve shared here in just these 5 stories. That experience has been hard earned. And I believe the judgment I have developed from all that experience makes me more likely to continue making winning investments in the future — which is why I invest so much of my personal capital in UV.

The challenge though, is that those potentially best investments, the ones that sound like bad ideas in the beginning — are hard to raise money for in the syndicate format. I am quite certain that if I had tried to run a syndicate for Burb’n, Nubank, or DoorDash in their first rounds — the vast majority of you would not have been interested. This is one reason why we do so many investments from our fund, and don’t syndicate them early on; there just isn’t a lot to talk about other than a promising founder and a half baked idea.

The fourth and last thing I’ll talk about is timing.

For the first time in over 13 years, I am hearing an abundance of people express concern that Venture Capital is not a good place to invest. It’s a contrast from 2020 and 2021, when the consensus was that VC returns were incredible, and everyone was piling in. And it’s different from 2022–2024, when many were clamoring “be greedy when others are fearful!” Now in 2025, all I hear is despondence.

You can hear it in the news, as well as in the chatter on Linkedin and X. You can see it in the movement of the money: LP commitments to VC are way down, and LPs are also concentrating most of their VC/PE allocations in a handful of mega funds that they perceive to be “safe.” They are not, but hey — no one ever got fired for hiring IBM.

The last time I remember sentiment like this was in 2011/2012. It was 4–5 years after the 2007 market peak, and VC returns had indeed been weak for quite a while. If you were in Facebook or a couple other successes, you did great. But outside of those, most VCs performed terribly. This line from a 2011 NBC article is illustrative: “this decade of VC investment doom has created serious questions about whether venture capital can ever rise from the ashes.”

Turns out: 2011/2012 was an excellent time to start investing in VC.

I wonder if the timing today is similar to 2011/12. I actually think our recent vintages have quite a few good investments that will go on to deliver excellent returns — but I also recognize that the environment has been tough. Follow on capital, markups, and business growth have all been slower. Capital has pulled back, and everyone is being more cautious.

Amidst the tough environment, I am also seeing investment opportunities today that I think are particularly compelling. We allocated a big chunk of our rolling fund this past year to one of our strongest existing portfolio companies, and wrote a big fund check and ran a large syndicate for another — both of which we think are irrationally undervalued. I envision these purchases as (hopefully) similar to when Fabrice Grinda of FJ Labs went out and bought secondary Alibaba shares ahead of their IPO, or when Chris Sacca did the same in Twitter. There are incredible deals to be had right now — in large part because no one is looking.

Even though the broad market is bearish, I am feeling the most bullish I have in quite some time. I personally think we have passed the bottom in the VC market cycle, and this next decade is going to be outstanding.

Keep in mind: I am not a permabull. I actually wrote about the onset of VC winter in January 2022. But although I believed the coming years would be challenging, I also think timing the market is hard. There are always outlier startups being formed, and great investments to be made. We are always investing, through the good times and bad.

Having said that, I do think that for those of you who do like to try to time the market, now may be an unusually good time to invest in Venture Capital.

I also humbly believe that if you are going to invest in VC, one of the best places you can invest is in Unpopular Ventures. I invest heavily in it myself, and we have a long track record of finding winners. Those of you with us already: we feel lucky to have you. For those of you reading this but on the fence: we would be honored to have you with us.

3. PORTFOLIO

SUMMARY: $81M of capital invested over the last 6.5 years has grown into $199M of portfolio value. This is up vs. 1 year ago ($71M/$183M), representing ~$6M of appreciation.

On the one hand, I’m grateful that our portfolio value continues to grow. At the same time, we are clearly still in the doldrums of this VC bear market. 3% year over year growth is not what any of us invest in Venture Capital for.

Keep in mind, it’s not unique to us, nor is it unique to the VC industry alone. The entire market of companies smaller than the 10 biggest in the US has been slow in recent years:

I don’t think this trend will last forever. No tree grows to the sky. The pendulum always swings back sooner or later.

But back to our own portfolio: although I’m not happy with the recent rate of appreciation, I am also optimistic. We have a lot of companies that are doing exceptionally well under the hood, even if the markups aren’t there just yet to provide “official credit” for their performance. Venture Capital is notoriously cyclical; there are years when it’s slow, and there are years when it’s fast — money flows in, prices go up, customers buy more, and companies feel the confidence to step on the gas and grow quickly. In 2020/2021, as an example, the average VC portfolio grew in value by over 100% year over year, and there was a slew of follow on rounds and exits at high prices. That was also ironically when most people wanted to invest in VC.

The investors who reaped the rewards during that time were those who had sowed their initial investments much earlier, when the market was not hot. As I shared above, in 2011/2012 most considered VC a bad place to invest. But that is when those who made the big bucks in 2020/2021 planted their seeds.

So although the numbers are not amazing right now, I am confident that we are sowing excellent seeds that will lead to a big harvest to reap in future years.

DISCLAIMERS

As we do every year, we want to remind everyone of a few disclaimers about how we do these updates:

1. Our updates lean positive. They go out to a lot of people (5,400+ LPs), and it’s not appropriate to share negative information about our companies publicly. Furthermore, VC is a game where only the extreme positive outcomes have a meaningful impact on portfolio performance. So let’s focus on those.
2. There may be errors here. All of the internal portfolio tracking is done by Peter and checked by the team and some of our biggest LPs with a spreadsheet. It’s very possible we got something wrong. If you notice an error, please let us know. We strive to make this as accurate as possible.
3. Startup valuation is tricky and complex. Different people do it in different ways. We explain our methodology in the appendix. You may disagree with it. You are welcome to run your own analysis and arrive at your own conclusions, using the raw data (linked below).
4. The only thing that actually matters is cash on cash returns. But in startup investing, those take a long time to materialize. The intent of this analysis is to estimate how we are doing in the short term, to evaluate if we are on the right track. You can’t take any of these numbers to the bank.

PORTFOLIO STATISTICS

Our aggregate numbers, across every investment in UV’s life:

Numbers for only the UV Syndicate:

Annualized numbers for only the UV Rolling Fund:

Statistics about our investing activity:

Quarterly breakdown for the Rolling Fund:

We rolled over $42k of uninvested capital from Q2, and raised another $920k in Q3.

If you are a major LP of Unpopular Ventures, have invested at least $250k to date, and are willing to sign an NDA, we will share the complete portfolio data with you. Please submit a request via this form: link

For everyone else, you can access the de-identified data here: link

TOP COMPANIES

Here we will highlight a few of our most promising companies, organized by vintage. Note that we have A LOT more companies than these that are doing extremely well. But for the purpose of not overloading you, we are only going to name a few companies per year.

The nature of our strategy, where we invest in 100+ companies per year from our fund, is that only the investments that can return 100x+ will have a meaningful impact on our returns. We are certainly very happy when our companies produce any positive return. A 10x (or even a 3x in some cases) can produce generational wealth for the founders, and we always welcome a profit. But the reality is: we are targeting a 5x return or better for each of our vintages — and we can only get there if we have individual portfolio companies that can make a meaningful dent in that and deliver at least a 1x on the vintage (and hopefully more).

So in general, the companies mentioned below are those that have had a meaningful impact on our returns, or that we think may have a big impact in the future. The more recent vintages lean more toward the latter, given that they haven’t had as much time to prove themselves yet.

2019

Stealth — this is an investment that wishes to stay under the radar, so we can’t name them publicly here. Those who are invested in it know which it is. The company is extremely profitable, growing rapidly, and did a share buyback earlier this year — using accrued profits to buy out any investors who wished to sell. The valuation in the buyback represented a 79x return (net) to our original investment. We actually felt that price undervalued them — so we used a piece of our rolling fund to buy shares from any LPs of ours who wished to sell, at the same price. We saw it as a win/win: providing liquidity to 8 LPs of ours who wanted it, as well as an opportunity for us and our fund LPs to buy more of one of our very best companies, at a great price. We believe the investment has the potential to grow another 100x from here.

Yassir — the leading super app in Francophone Africa. We led their seed rounds on $15M and $35M pre-money caps in 2019 and 2020. The company now employs over 4000 team members and serves customers in 45 cities across 7 countries. They have an enormous amount of revenue — more than any other portfolio company of ours — that is continuing to grow quickly. Our investment has been marked up significantly, but we believe the company’s current valuation undervalues it — and we also believe the company has a lot further to grow from here. Hat tip to Thibault on this one; it was the first investment he brought to UV.

Blaze (fka Almanac) — AI that does marketing for you. This one has been a bit of a rollercoaster. We originally invested in this as Almanac, which was a completely different idea. We were one of the biggest investors in their first round on $9M pre. They have pivoted several times since, raised a $30M Series A from Tiger in 2021, pivoted again after that — and now seem to have landed on something big. They have a lot of revenue that is growing very quickly now. Thanks to Kevin Moore for the intro, and Brad Flora for co-leading the syndicate.

Sniffspot — marketplace for private dog parks. I am mentioning this one as a bit of a wildcard — because they are not super crazy big just yet, nor have they raised tons of money — but they have built a very compelling business that is profitable and growing at a nice clip. With time, I think there is a real chance this one could deliver an outstanding investment return for us, especially relative to the $7M valuation we originally invested on.

This and the first 2019 company mentioned are together part of a theme we are observing: some of our best companies don’t raise very much venture capital. It’s a contrast from the “VC Factory Model” (described by Sam Lessin) in which startups follow a charted path of seed to Series A to Series B, C etc. — and success is measured based on subsequent fundraising. That model isn’t always great for us: it results in more dilution, and sometimes these companies raise too much money, try to grow too fast, and fail anyway. The reality is that it’s actually best for us if our companies can get profitable, and grow off their own cash flows, with as little external funding as possible — even if that doesn’t give us a flashy funding event to brag about. In contrast to the big VCs that need to deploy the billions they raised — and therefore want their portfolio companies to take lots of their money — we are small, and therefore it is better for us if our companies can get big with as little money as possible.

2020

Jeeves — the global business bank for enterprises operating across 25 countries. We were the biggest investor in their first two rounds (aside from YC) on $10M and $13M post money caps, and the company last raised at a $2.1B valuation. They have not done a round in a few years, as the company is in excellent shape — with plenty of cash, 3+ years of runway, healthy unit economics, and solid growth. July 2025 was a banner month across every metric including gross profit, revenue and TPV. They closed $75M in a new credit facility and are expecting to get their first bank license in Brazil this year.

99minutos — eCommerce delivery in Latin America. We were one of the biggest investors in their seed round on a $35M cap, and they last raised an $82M Series C from Oak HC/FT and Kaszek. The public information indicates the company is doing well — with 1069 current team members according to Linkedin and continuing to grow. H/t to Ed Roman and Anish Acharya for collaboration on this.

Decentro — Fintech infrastructure in India. Although this investment hasn’t been marked up dramatically yet, the company is in excellent shape with over 1300 enterprise users in India, has built a strong foundation that is core to FinTech in India, and therefore seems poised for growth in the future. H/T to Thibault for this investment.

Hint Health — we made large investments in Hint Health in 2020 and 2021. Their revenue has grown a lot since we invested, and the company has established itself as an industry leader in Direct Primary Care. We have received secondary offers to buy our shares, which we declined. Although Hint is not the fastest growing company in our portfolio, we invested a lot at a great price, the company is in excellent shape, growing well, and we expect that over time it will deliver a very nice return on our investment.

Note that every year we typically invest in the prior years’ winners too. 2020 has significant investments in Yassir and the Stealth co, as examples. But to simplify, we will only name each company once — in the first year we invested in it.

2021

Zepto — a leading grocery delivery app in India. We invested in their first round on $13M post, and the company was last valued at $5.8B. This investment is already worth ~¾ of all the money we invested from our fund in 2021 — nearly a “vintage returner.” We believe the company will grow much further in the coming years, and if it does it will produce a nice return on our 2021 rolling fund vintage. Hat tip to Thibault for this one.

Stepful — helps those without college degrees train for entry level healthcare jobs. We invested in their first round on $13M post, and the company is now doing over $3M of bookings per month. Based on their rate of growth, we think the company may achieve unicorn status soon. H/T: Thibault.

Community Phone — simple phone service for your home or business. We invested early, and the company is doing extremely well. Thanks to Rohit Taneja at Decentro for the intro here.

Yummy — leading super app in Venezuela. We invested early on a $7M pre money cap, and although the company has had some ups and downs, it is once again on an extremely promising growth trajectory. Thanks to Steven Coulis for the intro.

Constrafor — construction procurement company that offers financing and software for general contractors. We invested early, and the company is doing very well — lots of revenue and growing fast. Thanks to Zach Kruth for the intro.

Ready — helps utility providers connect families to better services at lower cost. We invested in their seed round, they raised subsequent rounds from top tier VCs, and they have a ton of revenue that is growing nicely.

Taiv — enables customized commercials on TVs in bars and restaurants. We invested multiple times over the years, and the company is generating a lot of revenue that is growing quickly.

Terra One (fka LIBER) — AI-controlled large-scale battery systems in Europe. This is considered one of the most promising startups in Germany right now, and we think has very high potential. H/T: Thibault.

GitButler (fka Sturdy) — we originally invested in a different company here, Sturdy, which shut down/merged with a new company, Gitbutler, and part of our equity was rolled into the new company. GitButler is doing very well, and in the midst of closing a big round from a top tier VC. Thanks to Dec Kelly for the original intro here, and special thanks to Avlok Kohli and AngelList for helping us manage some complexity related to this transaction.

2022

Helius — Solana development platform. Raised lots of subsequent funding from Founders Fund, Haun Ventures, and Foundation Capital, and is doing really well. Thanks to Gian Scozzaro and Faizan Khan for this one.

Brellium — Brellium helps provider groups automate clinical and payor compliance with AI. Consistently growing its ARR. H/T: Thibault.

Novig — America’s #1 sports prediction market. We invested early, and the company is doing very well with over $170M trading volume in June. Recently raised a big Series A from top tier VCs. H/T: Thiabult.

CopilotIQ — remote health monitoring for seniors. We invested in the company’s seed round, and they are doing really well. Raised subsequent rounds from top tier VCs, has lots of revenue, and acquired another large player in the space.

nSave (fka Masref) — global bank accounts for everyone, especially those living in places where they can’t trust their banks. We invested in their first round, were followed by top tier VCs, and the company is doing great.

AltScore — AI-driven lending infrastructure, enabling any company to build, manage and offer credit products in weeks. We invested in their seed round, were followed by great VCs, and the company is doing well. H/T: Ivan Montoya

Chowdeck — food delivery in Africa. We invested very early, followed on in the company’s subsequent rounds — and they now do a large amount of volume and revenue. They started in Nigeria, and recently launched in Ghana too. H/T: Thibault

Sky Fortress — Protects Ukrainians from Russian drone and missile attacks with a network of acoustic sensors. Economist Article. We invested at a $1.5M valuation, and the company is now the leading drone detection company in Ukraine, along with Zvook, which we also invested in. H/T to Sergii Zhuk for Sky Fortress, and Thibault for Zvook. Thibault has led our effort to invest in several promising Ukrainian defense tech companies.

2023

Marathon Data — measures the revenue impact of brand with AI. We led their pre-seed, and the company has had a very successful launch — with lots of enterprise customers and fast-growing revenue.

Revv — AI powered SaaS for car repair shops. We invested in their seed round, and were followed by top tier VCs at much higher valuations. Lots of revenue that is growing fast. H/T: Thibault

Baton — Baton is the leading AI-enabled marketplace for buying and selling small businesses. Unparalleled listing quality. We invested in their seed and they have raised more money from top tier VCs and are growing fast. H/T: Thibault

Hapi — the “Robinhood of LatAm.” We invested in their seed and they have seen extraordinary growth since then. Followed at higher prices by top tier VCs. H/T: Tiago Del Rio

Felix Pago — chat based platform that allows Latinos in the US to send money abroad. Crazy growth, followed by top tier VCs at much higher prices.

Fluently — AI language coach. We invested in their pre-seed, were followed by great VCs at higher prices, and the company is seeing phenomenal growth. H/T: Sergii Zhuk

Carry — helps business owners and modern professionals build wealth. We invested in their seed and Series A, and the company is doing great.

Magic Mind — healthy beverage that improves cognitive performance. We invested in their seed, and they have seen extraordinary growth. Their product is now sold in several major supermarkets, including Publix.

222 — in real life social network. We invested in their seed round, and they are seeing rapid user and revenue growth across several geographies. H/T: Thibault.

2024

Still early for our 2024 portfolio, but here are a few that are off to a fast start:

Undermind — AI search engine for academic research. This company became profitable right after we invested — with significant and fast growing revenue. H/T: Thibault

TomaAI — AI call automation for car dealerships. We invested in their seed round, and the company subsequently raised a lot more money from top tier VCs and is growing quickly. H/T: Jack Chapman.

Conduit (fka HostAI) — AI customer service for vacation rentals. We invested in their seed, and they are seeing fast user and revenue growth. H/T: Gabriel Jarrosson

4. PAST CONTENT

Aggregating some of our past content here so it’s easy to find:

Quarterly LP updates from the last year:

• Q2 2025
• Q1 2025
• Q4 2024

All annual LP updates since UV inception:

• 2024
• 2023
• 2022
• 2021
• 2020
• 2019

Recent Interview: Peter with Brian Bell (4/2025)

Recent Interview: Peter on the OpenVC Podcast (10/2024)

Recent Interview: Peter with Burak Buyukdemir (3/2024)

Advice to First Time Founders (and Maybe All Founders) (8/2023)

A Word of Caution to AngelList LPs (10/2022)

Is Winter Coming? (1/2022) — Our post calling the onset of the bear market — before anyone else.

Thank you all for reading, and thank you for your support of Unpopular Ventures!

5. APPENDIX

Portfolio Tracking Methodology

As we originally explained in the Q4 ’22 update, we present our numbers to be in line with the SEC’s guidance that if fund managers present gross performance numbers, they must also present the numbers net of all fees. This is a bit tricky for us, because we offer numerous fee options to investors in our rolling fund — each of which would require a different calculation. 2 and 20 is the default option, but we also offer 1 and 30, and even 0 and 20 with a 6 year commitment. That last option is our favorite (we prefer long term stable capital) but recognize that 2 and 20 is the market standard. Most LPs prefer a shorter commitment duration — so we have set that as the default choice.

To produce “net” numbers, we standardize them along the 2 and 20 fee structure. To do this, we are starting with the amount that we actually invested into companies, which is “real” — regardless of whether you are signed up with management fees or not. We are then creating an “imaginary” number for capital raised, which is the amount of money we would have had to raise based on how much we invested, if all of our LPs had signed up with 2% annual management fees. But because some of our LPs are signed up without management fees — this imaginary “raised amount” does not actually represent how much we raised. It just serves to facilitate proper calculations net of 2 and 20 fees. Keep in mind: if you have a different fee structure than 2 and 20, the net numbers are different for you.

We also separate out our syndicate and fund investing. We kept them together before Q4 ’22, because our intent was to simply measure ourselves — to make sure we were doing a good job investing — and we didn’t really care about which type of vehicle the capital came from. But because the fees on these two entities are different, we have to separate them in order to produce numbers that are net of fees.

Marking to SAFE Caps

In addition to incorporating fees, dilution, and carry into our calculations, one thing we do that differs from AngelList is that we mark to SAFE caps — both up and down.

The intent is to produce more-accurate reporting. For example, we have some companies that have raised a down round after an equity round on a lower valuation cap SAFE. AngelList still holds those at the last equity round price, but we value them lower. We do it in the upside direction too.

We approach this with some nuance:

• Must be within reason. If a company raises on an outrageously high cap (or uncapped) or with a huge discount — we don’t count it. It only counts if the discount is zero/minimal, the cap is commensurate with the company’s progress, and reasonably close to what the valuation would be if it had been a priced round.
• If it’s a post money cap, we treat it as the post money valuation. If pre, we treat it as the pre-money valuation.
• We track and incorporate all of the dilution that would occur between the rounds, to the best of our ability. For example, if we invested on a $5 M post money SAFE, and the company then raises $2 M on a $12 M post money SAFE, we would mark that as a 2.0x — the delta between our post money (5) to the next pre money (10). If they then took $3 M more on an $18 M pre money cap SAFE, we would mark it up to a 3.0x → (10/5)*(18/12) = 3.
• If a priced round is signed but not closed yet, we use similar math, but also add extra dilution to account for potential option pool expansion. If there has been no priced round for a long time, we add an extra 10% dilution. If there was a priced round very recently, with the option pool already refreshed, we incorporate less.
• As soon as we receive a pro forma cap table or the exact share price, we revert to the official share price multiplied by the number of shares we have.

There is some debate about whether you can mark to a higher cap SAFE/note, or if you have to wait for a priced round before updating the valuation. AngelList, for example, only marks to priced rounds. However, other seed stage VCs say it’s common practice to mark to caps. In fact, the NVCA appears to believe it’s acceptable to use a cap as the effective valuation:

Historically, it made sense to only mark to priced rounds, because most rounds used to be priced rounds. And particularly for AngelList, with thousands of investments and a lean team — it’s most efficient for them to only mark to a share price. However, with the growing use of SAFEs, we are increasingly seeing early stage startups do 2–6 SAFE rounds over 2–4 years before eventually doing a priced round.

We certainly could wait for only priced rounds before updating our marks — but that means there will be no meaningful data to report for years. We (and probably you too) would like to estimate our investment performance on a more granular level. We think the best way to do that is to mark to reasonable caps as rough markers of valuation, while calculating the effect as if they had been priced rounds.

If another investor is investing on a higher or lower cap SAFE — they are legitimately investing on a higher/lower valuation — and we think it’s reasonable to incorporate that.

Unpopular Ventures: Annual Update 2025 was originally published in Unpopular VC on Medium, where people are continuing the conversation by highlighting and responding to this story.

UV LPs,

Welcome to our April 2025 Quarterly LP Update. As a reminder, at Unpopular Ventures we focus on “the best companies, off the beaten path.” We invest in exceptional founders who are building high potential businesses, that are non-consensus in some way.

To help us find great opportunities around the world, we leverage a “Scout Program,” through which we share significant portions of our carried interest with anyone who helps us to identify, evaluate, and diligence companies that we invest in. If you would like to refer an opportunity, please check out our Scout Program Guidelines.

This is our quarterly update, which will be shorter than the annual update we write once a year. If you’d like to read about some of our most promising portfolio companies, how we do our portfolio tracking (along with relevant disclaimers), or navigate to some of the past updates/content we’ve written, you can find our last Annual Update here.

This update includes two big milestones that we are proud of: 1) a meaningful liquidity event, and 2) we surpassed 500 portfolio companies. As usual, see further down for our latest portfolio metrics.

Thank you all for your continued partnership!

Sincerely,

Peter Livingston and Thibault Reichelt

Unpopular Ventures

Reminder: you can invest in our Rolling Fund for broad access to our portfolio, which co-invests in every new company we invest in, and invests in many more companies than you see in the syndicate. Rolling Fund LPs also receive preferential access to limited-allocation deals.

DPI! (for LPs who wanted it)

One of our strongest portfolio companies from 2019 finally completed its tender offer, using cash from accrued profits to buy shares back from any investors who wished to sell. The company remains in stealth, so we can’t share the name publicly here. We remain bullish on the company, so didn’t want to sell shares ourselves, but we recognized that some of our LPs might — so we found a way to pass through the option — allowing LPs to sell their shares if they wanted to, at the same price. We had 8 LPs choose to sell.

If an LP invested an equal amount in every investment we did in 2019, and opted to sell 100% of their 2019 position in this one company, they now have 2.65x their money back for the entire vintage. This is net of all fees, carry, dilution, etc. I.e. if they invested $100k with us in all of 2019, spread evenly across every deal, and sold, they now have $265k in their pocket.

The net multiple on the specific investment is 63.7x. Of course, most LPs didn’t invest in every deal we did, nor the same amount in every deal — so some LPs did better than 2.65x, while some got less. One LP in particular has only invested in 2 deals with us ever, was lucky enough to invest in this one and sold, and now has 21.3x his money back across all the money he ever invested with us.

Because we did not choose to sell ourselves, this will not actually be reflected in our performance metrics as DPI (distributions to paid in capital — i.e. multiple of cash returned relative to cash invested). But we are proud that we were able to provide the option of early DPI to LPs who wanted it, especially in this unusually illiquid time in the venture market. We will strive to continue to find similar opportunities for our LPs in the future.

Why didn’t we sell? We remain bullish on the company, and as you’ve heard from us before — we are long term investors. We are strong believers in the power of compound interest (or technically compound growth in this case) over a long time. That’s how you make the big money as an investor: buying great companies, and holding on to them for as long as possible.

Last year there were articles about how two Sequoia partners that had led Nvidia’s seed round, still hold most of their personal shares today, 30+ years later. When I (Peter) read that, I thought “that is going to be me.” When our top portfolio companies eventually go public, we will of course distribute the shares to our LPs — who will have the choice of what to do with them. But the shares I receive personally, from my personal investments as well as my carry — I intend to hang on to for a long time. You never know which companies are going to grow another 100x+, even after going public. It’s that long term conviction that leads to generational wealth.

But of course, not everyone shares that same long term approach. And some people encounter unforeseen life events that lead them to prioritize liquidity — which is completely understandable. We are happy that we were able to give some LPs an off ramp.

How did we do it? I.e. how did we provide the option to individual LPs, without having to sell all or part of the aggregate SPV?

We’ve been trying to invest more in this company for a while, but have not been able to, because the company literally doesn’t need the money. They are producing so much cash, more than they think they can responsibly re-invest, that they opted to do a share buyback. So it became clear that the only way to buy more shares in the company would be via secondary transactions from other existing shareholders. With this in mind, we allocated 20% of our Q4 2024 Rolling Fund to go out and try to buy secondary shares. We approached a lot of people, including former employees, but frustratingly, we couldn’t find anyone willing to sell. To make it harder, we were competing with the company, who was also trying to buy shares.

We were about to give up, and roll the uninvested funds into the current quarter to invest into other startups, when we asked AngelList if we would be allowed to offer to buy shares from LPs in our past SPVs. We thought it was a long shot, given the potential conflict of interest. We were pleasantly surprised that AngelList gave us the green light. We think this was for a two reasons: 1) we didn’t set the price, the company did (determined by an independent valuation firm) — and we merely passed through the same price, and 2) it was optional, up to the LPs to decide to sell — not us. No one was forced to sell, we were transparent that we are bullish on the company and wanted to buy more ownership, but that LPs could choose to sell to us if they wanted the liquidity. To minimize the conflict of interest further, we opted to receive no cash from our carry in the LPs’ SPV positions that were sold, and instead let our carry roll over. In the end, we were able to buy out every LP who wanted to sell, for 100% of their desired amounts.

So it was a win/win outcome. We were able to make the investment we wanted to in one of our strongest companies; although it has appreciated significantly from our original investment, we believe it could grow another 100x+ from here. And we were simultaneously able to provide liquidity to the LPs who wanted it.

500 Startups

I recall reading about Dave McClure’s 500 Startups a long time ago, and thinking that was an unfathomable goal. How could someone possibly invest in that many startups? I’m proud to report that we have now done it, crossing the 500 portfolio companies mark this quarter.

Why do we invest in so many, when most VCs pursue a more concentrated approach? Simple: we think it produces better returns.

The average VC significantly outperforms the median VC. This is due to the power law nature of startup outcomes, where a small number of the companies produce the lion’s share of the returns. I.e. if you could invest in every startup, or every VC firm — you would be quite happy with the results. But if you instead invested in one at random, or the exact median one, you would probably be unhappy with the results — because you would miss the top 10% of funds, or top 1% of startups — that made the returns of the whole group amazing. Most VCs invest in 30–50 startups per fund; how likely are they to invest in a top 1% startup, or really a top 0.1% startup — the type that can return 1000x+? Not as likely as if they invested in more companies.

500 Startups understood this, and through a diversified approach of investing in every promising startup, repeatedly produced top quartile returns. Y Combinator has done even better (albeit with a different model as an accelerator). And if you compare our returns so far against the returns of the industry — you will recognize that we are way up there as well — tracking within the top decile of VC industry returns for most of the years we have existed (especially the mature vintages 2019–2021, which are more reliable; recent vintages are too early to tell):

See the next section for our complete metrics to compare. To single out 2019 alone, the vintage that we just produced meaningful (optional) DPI for — you can see that the 2.65x DPI we just produced for LPs who invested an equal amount in every deal we did, and opted to sell — is not only way higher than top decile DPI for the vintage, but is higher than 2019 top decile *TVPI* too (2.00x).

So if investing in more startups produces better returns, why doesn’t everyone do it? Two reasons: 1) most VCs don’t have consistent, abundant, high quality deal flow, and 2) the incentives of the existing LP ecosystem don’t support it.

To the first point, we do have exceptional deal flow — thanks to our vast network of “scouts” and the great deal we offer them, combined with a good reputation that we have built over a long period of time. We believe we have better deal flow — in both quantity and quality — than the vast majority of VCs.

To the second, as we’ve written in past LP updates, a key segment of the LPs that invest in emerging managers (like us) strongly prefer concentrated VC portfolios: the “funds of funds.” These funds of funds aggregate large pools of capital and invest in lots of the smaller VC funds. The large LPs, like pension funds and endowments, see firms like us as too small for them to spend time on, so they outsource their small fund/emerging manager investing to the funds of funds. A key part of the funds of funds’ strategy is to invest part of their money in the VC funds, but then reserve most of their money to invest directly into the top portfolio companies that their portfolio VCs produce — taking over their pro rata, and investing directly in the startups without fees. They do this to justify their “value add” and therefore the fees they charge (on top of the fees of the underlying VC funds). But this strategy only works if 1) the underlying startup ownership positions are significant, representing a meaty pro rata for them to take over, and 2) there are a relatively small number of portfolio companies to keep track of.

So the concentrated approach is optimal (or at least perceived as optimal) by the funds of funds, because of their own incentives. But the optimal strategy for VCs who care about the returns of the VC funds themselves, and therefore for the LPs who are passively investing in VC — is a higher quantity of investments, in our opinion. That’s what we have done for the last 6 years, and it is really working for us — so it’s what we are going to continue to do. We are proud to have surpassed 500 portfolio companies to date. Next stop: 1000.

PORTFOLIO

SUMMARY: We had a small uptick this quarter. $77.5M of capital invested to date has now grown into $194M of portfolio value. This compares to $75M/$191M last quarter, i.e. $2.5M of additional investment and $3M of portfolio value growth — reflecting $0.5M of appreciation. The tender offer we mentioned above was already priced in last quarter.

Unfortunately, the macro exuberance we saw in Q4 and early Q1 hasn’t lasted, and the public markets have dropped back into gloom and uncertainty. Consumer sentiment has also plunged to the second lowest level on record since 1952 (source). Consumers and investors alike seem to have pulled back into conservative mindsets, waiting to see just how bad the macro tumult from tariffs and other economic decisions by the administration will be.

Some positive venture rounds are still getting done, but lots of startups are continuing to shut down or raise down rounds too. On net, our markups and mark downs more or less balanced each other out this quarter, and we imagine that that movement is representative of the overall VC market right now.

We are hopeful that the market turmoil won’t last too long, but it’s impossible to know for sure. If we get a resolution on the tariffs, and/or a change in course from the Fed, everything could flip suddenly. The last time I remember the market feeling this panicked was March 2020; then the Fed turned on the printer and everything ripped through 2021.

Having said all that, we don’t spend that much time worrying about the macro. We continue to simply make investments in the best founders and startups we can find, preferably those doing something a little unpopular or off the beaten path, and hold on to them for a long time. Great startups are always being founded, rain or shine, and the market will go through multiple ups and downs over the 10+ years we expect to hold them for. The very best companies will survive and thrive, and ultimately produce great returns for us investors.

PORTFOLIO STATISTICS

Our aggregate numbers, across every investment in UV’s life:

Numbers for only the UV Syndicate:

Annualized numbers for only the UV Rolling Fund:

Statistics about our investing activity:

Quarterly breakdown for the Rolling Fund:

We rolled over $57k of uninvested capital from Q1, and raised another $1.1M in Q2.

If you are a major LP of Unpopular Ventures, have invested at least $250k to date, and are willing to sign an NDA, we will share the complete portfolio data with you. Please submit a request via this form: link

For everyone else, you can access the de-identified data here: link

Thank you for reading, and thank you for your support of Unpopular Ventures!

UV Quarterly Update: Q2 2025 was originally published in Unpopular VC on Medium, where people are continuing the conversation by highlighting and responding to this story.

UV LPs,

Welcome to our February 2025 Quarterly LP Update. As a reminder, at Unpopular Ventures we focus on “the best companies, off the beaten path.” We invest in exceptional founders who are building high potential businesses, that are non-consensus in some way.

To help us find great opportunities around the world, we leverage a “Scout Program,” through which we share significant portions of our carried interest with anyone who helps us to identify, evaluate, and diligence companies that we invest in. If you would like to refer an opportunity, please check out our Scout Program Guidelines.

This is our quarterly update, which will be shorter than the annual update we write once a year. If you’d like to read about some of our most promising portfolio companies, how we do our portfolio tracking (along with relevant disclaimers), or navigate to some of the past updates/content we’ve written, you can find our last Annual Update here. This update will have some general thoughts from us, and then an update on the portfolio.

Thank you all for your continued partnership!

Sincerely,

Peter Livingston and Thibault Reichelt

Unpopular Ventures

Reminder: you can invest in our Rolling Fund for broad access to our portfolio, which co-invests in every new company we invest in, and invests in many more companies than you see in the syndicate. Rolling Fund LPs also receive preferential access to limited-allocation deals.

GENERAL THOUGHTS

Written by Peter, hence the switch to first person.

MARKET

AngelList recently published a report on The State of US Early Stage Venture and Startups. It’s a fantastic read, packed with informative data.

It highlights and confirms a lot of what we are seeing on the ground ourselves. 2.5 years ago (August 2022) I first shared this chart, estimating what the venture market and our portfolio might do over the coming years:

We had just been through a period where everything was up and to the right, and I speculated that as we entered a broad bear market, portfolio-level marks would probably flatten for a while or in a worst case scenario would dip, before eventually turning up again. In several updates since then, I had commented that it seemed like we were (luckily) tracking more along the blue line than the red. Lo and behold, AngelList just shared the first 2/3s of the same chart, based on real market data:

It reflects exactly that path: up and to the right through 2021, with everything ~flat since: the “blue line.” Vintages that began in 2021 and after have been flat lines.

We feel fortunate that we’ve actually been doing better than flat, thanks to a handful of outliers that are lifting the portfolio. For example, our 2019 portfolio has gone from 3.25x to 5.16x since August 2022 (thanks to Yassir and another company that remains in stealth but is now marked at 79x our investment), and our 2021 vintage is well into the green thanks to Thibault’s investment in Zepto (now at 86x, and rumored to be preparing an IPO).

But if you take out the outliers, these last 2–3 years have really been slow — especially relative to the go go years of 2020/2021. This next chart from AngelList captures the market shift from another angle:

What this shows is that starting in 2022 and accelerating in 2023, there were progressively fewer deals getting done, and the deals were progressively worse. It stopped getting worse in 2024 (thankfully), but we’ve been stuck here in the doldrums since then.

What comes next? I do think we are due for the market to turn up sooner or later. We have a lot of portfolio companies that are making great progress under the hood, but just haven’t been priced up yet. You can only hold a beach ball under water for so long.

Having said that, there is also a plausible scenario in which we plunge into an extended/deeper winter in the overall VC market. On January 27th, Dan Primack (Axios Pro Rata) wrote about the possible implications of DeepSeek on the AI Models, and in turn the VC market:

As he points out, there is still a lot we don’t know. But at the very least, this does seem to highlight that perhaps the moats for the AI models are thin. I have often wondered if the LLMs might end up being kind of like the airline businesses: extremely important, everyone uses them, provide tremendous value for society… but terrible businesses. If it’s easier than everyone expected to train a great model, the market ends up with several competitors, and these models are relatively low margin businesses when you factor in the costs to train and the short lifespans before obsolescence… perhaps OpenAI isn’t worth anything close to its latest valuation, and the tens of billions of investment have been wasted.

Considering how much has been invested, from so many top firms — that level of capital destruction would be a massive setback for the industry. If that worst case scenario came to pass, there might be a LP stampede out of the VC market, fund sizes may shrink dramatically, and VCs will be a lot more cautious for a long time. If that happens, we won’t be able to count on big and quick markups to our portfolio like we saw in 2019/20/21. The returns will only come over a longer period, as the companies grow organically and eventually sell or IPO. In that case, we might go 5+ years before seeing a change in price, vs. the price changes every 6–24 months we grew accustomed to when VC rounds were abundant. Time will tell.

What does it mean for us? Not much to be honest. We are so tiny, and we’ve always invested so early, that by nature we have had to focus more on companies that could do a lot with little. Those that had the potential to be capital efficient, and possibly reach profitability without being dependent on tens to hundreds of millions of dollars of additional capital. This is in contrast to the mega funds, which are incentivized to deploy their billions by finding companies that are likely to need hundreds of millions each. We certainly welcome it when our portfolio companies raise lots of follow-on capital; it often helps the companies grow faster and it marks up our investment value.

But we are also perfectly content when they don’t. In fact, one of our very best portfolio companies (which I can’t name due to it being in stealth), has only raised $6M total, is wildly profitable, growing fast, and has so much cash on hand from accrued profits that it is actually trying to buy back shares from its investors. Truly the dream outcome — especially in the current market. We will continue to try to invest in more companies like that.

VALUATIONS

Another thing worth commenting on from the AngelList report is this chart about valuations:

As you can see, seed and pre-seed valuations have risen dramatically in recent years. In 2015 the average seed was at $6M, and now 2022–2024 the average seed has been at ~$20M. Pre-seed started around $4–5M, and is now at $10M. Surprisingly, despite the bear market, these valuations haven’t come down. Why?

One thing worth noting is that “pre-seed” wasn’t really a thing back in 2015. The first round was just “seed.” But as seed firms got bigger, had to write bigger checks, and moved later, it created a vacuum for new, earlier investors to move in — and that gap had to be labelled something new. So “pre-seed” is probably just what “seed” was 10 years ago.

The same thing happened with Series As. It used to be that Series A was the first round of funding that a company raised. When I raised money for my startup from Kleiner Perkins back in 2011, the deal they offered was $1M on $2M post, we negotiated them up to $2.5M post, and the label on the term sheet was “Series A.” Going further back to the 2000s, or the 1990s, “Series As” were often even smaller at $250–500k.

The metrics have shifted too. Seed used to mean basically no traction, but now all the seeds have traction. It’s the pre-seeds that usually don’t (although they still often do). Series A used to mean $500k-$1M in annualized revenue, and a valuation of $20–40M, but now the expectation for Series A is more like $2–4M of revenue and a valuation of $50–100M. $500k-1M of revenue is now a “seed” in the $20–40M range. The point is, the nature of the rounds haven’t changed that much, but the labels have.

Having said that, there is valuation inflation too. The prices we see today for certain deal profiles are legitimately higher than they were before 2021. As examples, we did Zepto and Jeeves during YC at $13M and $10M post money valuations in 2021 and 2020 (with traction), and today we see very little in YC at less than a $20M valuation. YC is of course affected by their new deal, where they took over the “during-YC round” and replaced it with an extra $350k from YC on a MFN SAFE. But even outside of YC, valuations seem to be up.

The dynamic may simply be supply and demand. Amidst the market uncertainty, fewer talented people are starting companies, but there remains a lot of dry powder sitting in the VC firms. More money + fewer strong founders = higher prices.

But at the same time, a lot of companies *are* struggling to raise money these days. One thing that has frustrated Thibault and I is that not only are the valuations up, but the range of valuations at each stage seems to have tightened. The majority of pre-seeds I see, for example, are raising at a flat $10M. It used to be that some would be at $5M, some at $8M, some at $12M, and some even at $2–3M. And with that valuation variation, I could think about each one in terms of the risk/reward of that specific opportunity, at that price. There were deals I would do at $2–5M, but wouldn’t do at $10M.

But today, it seems like many (most?) companies are saying, “we are raising a pre-seed, the average pre-seed valuation is $10M, so that’s our valuation too.” No thought goes into it. Which means: we have had to raise our bar. We probably would have done more investments if we felt the price was more reasonable. But if everyone is raising at a flat high price, we have had to change our thinking from “is this a good deal at this price?” to “is this one of the top companies at this stage, thereby justifying an investment at this price that everyone seems to be raising at?”

So it’s tricky. I’d like to see valuations come back down, or at least see the range expand — which would enable us to do more investments. But for now, while valuations remain high and tight, we are pacing slower than we have in past years.

HIT “MY NUMBER.” WHAT NOW?

One last thing that’s been on my mind in recent months is that I’ve reached “my number.” Similar to what I think a lot of us do, long ago I decided on a net worth number that I felt that if I reached it, I’d have enough and wouldn’t feel like I needed to work for money anymore. I actually hit my original number a few years ago, but quickly realized it wasn’t really enough for me, especially with inflation, and I set a new number 3x higher.

I recently hit the new higher number, thanks to both investments I made during my solo investing years, as well as continued appreciation of some of our outlier UV investments. Some of it is liquid, some not, but as a diversified portfolio that I have confidence in in aggregate — I’ve reached a mental state in which I feel like I legitimately have enough.

There’s a funny dynamic in the VC career path that I think leads it to be short for most. Most people are either 1) bad at it, they strike out, and can’t raise more LP capital or are pushed out of their firms, or 2) are good at it, and make a couple monster investments that lead to them making enough money that they don’t have to work anymore. There is very little room in the middle, where you do well enough to stay in, but not well enough to make a lot of money.

I feel very lucky to have made it down path #2. I could retire if I wanted to. But I’ve given it a lot of thought over the last couple months, and decided that I love doing this, I think I’m pretty good at it, and honestly don’t know what else I would enjoy spending my time doing. I’m still young too — only recently turned 40.

So I’m going to keep at it, but I think my new mental state is changing how I invest. In particular, I’m becoming more selective. You might have noticed that I’ve put up fewer syndicates in the last few months than I have in the past. We’ve been doing fewer (but often larger) investments from our fund too.

The reasons for this are 2-fold. First, looking back at all the investments I made personally and we at UV have made together, it’s remarkable how few of them have really mattered. It’s probably 10 or fewer in total for UV, out of almost 500 companies we’ve invested in in aggregate. In another 5 years, it may be only 1–2 investments that will be worth more than everything else combined, even if everything else does reasonably well. The power law is real.

A remarkable statistic I recently learned: the number of investments in Sequoia Capital’s entire 50 year history that have returned at least 20x. Can you guess?

42. (Source)

Fewer than one per year. All of Sequoia’s elite performance is embedded in those 42 companies. It’s not unique to Sequoia either. If you look into the track record of most VCs, or most investors for that matter — you’ll see that most of their returns came from an extremely small number of investments. I read recently that most of Warren Buffet’s investing success is attributable to 5 investments of his (although there’s probably some nuance to that — depending on if you slice it by dollars, or % returns per year). You can see another example of it in the S&P 500 right now — where most of the gains in recent years have been driven by only 7 stocks.

Although we are often wrong, looking back, our current best investments did indeed feel like particularly high quality investments when we did them. Whereas a lot of the lower performing ones, in retrospect, do feel like maybe they were a little more borderline when we did them. So with this in mind, we are slowing it down, and trying to do a smaller number of the very best investments we can.

The second reason is that financially speaking, small investments in so-so opportunities won’t move the needle for me anymore — and I’m feeling that more and more. If the carry I earn from a given syndicate (or fund) investment will be worth 0.005–0.03% of my net worth, and the best outcome I can see for the investment is only 100x — thereby adding 0.5–3% to my net worth — is it really worth doing? No, the bar has to be higher. I have to both believe it’s such a great deal that we can invest a lot in it, and that the potential return could be huge: 500x+ (and really 1000x+).

The point is: the pace of investing you will be seeing for us (and especially me) will be a little different going forward — driven by both the broader market dynamics, as well as my personal mental/financial state. Fewer, bigger, and higher quality investments.

PORTFOLIO

SUMMARY: This was an up quarter for us, thanks to a couple of big markups. $75M of capital invested to date has now grown into $191M of portfolio value. This compares to $74M/$182M last quarter, i.e. $1M of additional investment and $9M of portfolio value growth — reflecting $8M of appreciation, 4.4% quarter over quarter.

The last couple of quarters have been tough for a lot of our portfolio companies, with the quantity of shutdowns and down rounds seeming to accelerate. It feels like we are finally in the depths of the bear market (as AngelList’s charts above confirm). Not only is the market tough, but a lot of the companies are finally running out of money too — forcing more official changes in price.

But of course, the way VC goes is that the returns are really driven by a small number of the investments, and it almost doesn’t even matter (financially speaking) what happens to the rest of them. So long as we get into 1 or more monster home runs, our portfolio-level returns will look amazing. And we feel fortunate to have enough of those — spread across multiple vintages too.

One small point of frustration is that many of our very best companies have also become the most secretive, which limits us in what we can share with you.

There seems to be a funny dynamic where the weakest performers and (some of) the best performers get the most quiet. The weakest ones because they are embarrassed that they can’t find a path to success, and/or just don’t have much to say. And the best ones get quiet because they become deathly afraid that if others find out how well they are doing, competitors will rise to chase them.

It’s (mostly) the companies in the middle that make the most noise. They tend to be more comfortable with us sharing news about them, and they post more information publicly. They are doing well enough to stay alive and make decent progress, but not so well that they think competitors will notice or care about their success. They usually have good reason to make more noise too — as a means to recruit employees, and/or raise additional funding.

It has been interesting to see how one of our portfolio companies is using AI agents so efficiently that it became profitable in record time and after having raised only seed capital. This company will probably never need to take on more VC funding. The less money you raise the less you dilute yourself in the process.

This could herald a new trend. Similar to how the likes of AWS and Stripe have made it much easier for startups to get going with a minimum of capital, compared to previous generations who had to buy their own servers and build their own payment rails. If more of these hyper-efficient AI-powered teams emerge, the only way to invest in their companies will be at the seed stage. This would be great news for funds like ours and bad news for the large VC firms whose business models are based on deploying large amounts of capital. We write small checks and hope for outlier returns over long periods of time. The large VC firms do not quite have the same incentives since they collect hundreds of millions in management fees on the basis of their billions of assets under management.

PORTFOLIO COMPANY NEWS

(As always, we are highlighting here only a few developments from the portfolio. It does not mean that other portfolio companies are doing any worse.)

Zepto is expected to IPO this year.

Yassir has launched its first fintech product, Yassir Pay, and is continuing to grow headcount according to LinkedIn.

Stepful is approaching unicorn territory in terms of its revenue traction.

Inversion has successfully launched to orbit and Albedo is about to do the same.

Rio is growing rapidly and expanding to Colombia.

Baton has raised its Series A.

A leading secondaries fund has offered to buy our Hint Health shares. Usually a sign that the company is doing well. As usual, we won’t be selling.

Palitronica is progressing nicely.

CopyAI executed a pivot and is finding great success in their new direction.

Yummy, after a pause, is once again on a nice growth trajectory.

One piece of news that’s bittersweet is that One Finance raised $300M at a $2.5B valuation. Unfortunately, we were forced to sell in the Walmart transaction for a little over a 2x return, at a ~$400M valuation. But good to know our bet was correct, even if we weren’t allowed to participate in all of the upside.

Altscore is continuing to make great progress after their Series A.

Blaze (fka Almanac) has found its footing after a couple pivots and is now growing fast.

Magic Mind is growing and expanding across retail locations really well.

Felix Pago raised another up round, at a much higher price.

We are invested in 2 different companies named HostAI (first, second) — both are making excellent progress.

Sniffspot, Hapi, Community Phone, Pippin Title, and The Swarm are all growing extremely well, especially relative to the amount of funding they have each raised.

PORTFOLIO STATISTICS

Our aggregate numbers, across every investment in UV’s life:

Numbers for only the UV Syndicate:

Annualized numbers for only the UV Rolling Fund:

Statistics about our investing activity:

Quarterly breakdown for the Rolling Fund:

We rolled over $145k of uninvested capital from Q4, and raised another $1.15M in Q1.

If you are a major LP of Unpopular Ventures, have invested at least $250k to date, and are willing to sign an NDA, we will share the complete portfolio data with you. Please submit a request via this form: link

For everyone else, you can access the de-identified data here: link

Thank you for reading, and thank you for your support of Unpopular Ventures!

UV Quarterly Update: Q1 2025 was originally published in Unpopular VC on Medium, where people are continuing the conversation by highlighting and responding to this story.

UV LPs,

Welcome to our November 2024 Quarterly LP Update. As a reminder, at Unpopular Ventures we focus on “the best companies, off the beaten path.” We invest in exceptional founders who are building high potential businesses, that are non-consensus in some way.

To help us find great opportunities around the world, we leverage a “Scout Program,” through which we share significant portions of our carried interest with anyone who helps us to identify, evaluate, and diligence companies that we invest in. If you would like to refer an opportunity, please check out our Scout Program Guidelines.

This is our quarterly update, which will be shorter than the annual update we write once a year. If you’d like to read about some of our most promising portfolio companies, how we do our portfolio tracking (along with relevant disclaimers), or navigate to some of the past updates/content we’ve written, you can find our last Annual Update here. This update will have some general thoughts from us, and then an update on the portfolio.

Thank you all for your continued partnership!

Sincerely,

Peter Livingston and Thibault Reichelt

Unpopular Ventures

GENERAL THOUGHTS

MARKET

While we recognize there will be mixed opinions here about the outcome of the US election, one thing we think few will disagree with is that the Trump and Republican sweep is likely to be very good for business at every level. The prospect of lower taxes and reduced regulation will be a boon for growth, profitability, M&A, the IPO market — and optimism in general. In the words of Siqi Chen:

You don’t have to take our word for it; the market is already speaking: public equities are smashing new all time highs, and Bitcoin is on fire. We are even seeing it in our little corner of the world on AngelList — with syndicate deals already appearing to raise more money on average than they were before. Historically, AngelList has been a remarkably accurate gauge of VC sentiment at large — at the macro, market sector, and even individual deal levels. VCs are people after all, no smarter than many of the sophisticated angels we have here on AngelList. So the behavior and sentiment that we see on AngelList is often indicative of what is happening in the broader VC market. And thus far, things are looking up.

Looking ahead, we believe the coming years are going to be fantastic for our portfolio. We are grateful to have had the opportunity to buy ownership in so many promising startups over the last 3 years when it has not been easy. Startups haven’t been growing as fast as they were in ‘20/21 when they and their customers were getting pumped with cash and the tide was rising. Since 2022, investors have pulled back dramatically, leading to our companies receiving less follow-on cash, and also reducing the “credit” we used to receive for our good investments in the form of quick, high value markups.

Series A is the first big funding milestone that matters to our portfolio, considering that 90%+ of our investments are at various stages of seed. The Series A market has been exceptionally slow recently, and you can see it in Carta’s data. Whereas 34–38% of startups in 2020 raised a Series A within 2 years, only 12–15% of startups from 2022 have:

Investing in 2019, 2020, and 2021 felt easy in comparison. It’s not hard to invest when almost everything you buy increases in price in a short amount of time. In contrast, these last 3 years have been much more of a grind; they haven’t been delivering that same rapid positive feedback cycle.

But at long last, we expect to finally reap the rewards of that hard work. THANK YOU to all of our LPs who have supported us through the unpopular years. We expect that you are going to be very happy with the investments you made with us — at both the fund and syndicate levels — as our portfolio matures and the market warms in the coming years.

TIME

One of the dominant narratives in VC land right now is that LPs are dissatisfied with how long it’s taking to get liquidity on their investments. For good reason: the M&A market has been slow, and the timeline to IPO is the longest it has ever been.

This table from Jason Lemkin’s recent post captures the duration problem well:

In the 1990s the timeline to IPO was 4 years. Recently, the conventional wisdom has been that it takes 10 years. But now looking at the top private companies — the ones that will really “make” the VC funds that invested in them — the time to IPO seems to have stretched to 13–14 years, and even longer in some cases.

That is a long time. I, Peter, consider myself a fairly experienced startup investor, having been at it for 12 years now (as a solo angel for the first 7) — and even that isn’t as long as the average journey to IPO these days.

But against that narrative, I’ll share an unpopular opinion: I actually like how long it takes.

Certainly, I also like it when we make an investment and we get a big cash return in a short amount of time. I will always welcome that. And if we have opportunities to return cash to our LPs without sacrificing long term returns, we will always strive to do it. We take our fiduciary responsibility to our LPs very seriously.

But zooming out, to my own experiences as an Angel Investor — I personally love the positive surprises that trickle in over a long period of time.

As an example, last month I received an update from a company I angel invested in in *2014.* I mentally wrote it off a long time ago, and hadn’t even received an update since 2020. Out of nowhere, 10 years post investment, I learned that they pivoted business models, passed $3M ARR, are growing fast, have a fresh inspiring vision, and are closing an oversubscribed venture round. My original $25k investment is now worth $125k — and could go much farther.

I love that the founder stuck with it this long, and is probably going to earn generational wealth from her efforts. She still owns ~85% of the company. A separate aspect of this that I love is that I get to be a small participant in an entrepreneur’s life’s work. She spent 10 years to get here, and might spend another 10 taking it to scale, maybe even longer. A simple action 10 years ago, combined with patience and positive support, has enabled me to ride along as a small part of her journey.

The surprise news made my day. An extra $125k investment position I didn’t even know I had. Sure — 5x in 10 years isn’t anything crazy, but it beat the S&P 500. It will be tax free too, thanks to QSBS, when they eventually decide to exit.

Now, perhaps I can talk like this because I also *did* get quick liquidity from other investments. As I said, the quick liquidity is important too. But I just love that this work I did so long ago is continuing to produce surprise investment gains even today.

I’m certain that the portfolio we have built at UV, and are continuing to build, will behave similarly. Some of these investments will pay out sooner — and we intend to return our LPs at least 1x their cash in a respectable amount of time. But I’m also sure some of them will take longer, and we will continue to receive surprise “dividends” from some of these investments for 10, 15, even 20 years.

There’s something about this that really gives me an optimistic outlook. It makes me feel good to know that some of these will surprise me far out into the future.

What else can you invest in that might give you a positive financial surprise so far out into the future? You can do something today that will have an impact on you and your family 20 years from now. In my opinion, that is a feature of VC, not a bug.

TAX TIP

Let me start with a disclaimer: I (Peter) am not a tax professional nor financial advisor, and before you take any actions related to what I’m sharing here — please consult with someone who is. My intent is to share something I have personally chosen to do, and maybe it will serve as an idea for you to investigate as well.

With that out of the way, here’s the tip: consider gifting some of your startup shares to your kids. Especially shares that are likely to be QSBS eligible.

Some people choose to give money to their kids, some don’t — and that’s a separate debate. Among those that choose to do so, there are a variety of ways to do it. One popular approach is to set up a 529 college savings plan — which allows the money to grow tax free, so long as it is only used to pay for education.

Parents typically gift their kids up to the maximum amount they are allowed to give tax free, per parent, per year — which in 2024 is $18k, or $36k from the two parents combined. If you do that for several years, early in your kids’ lives, it will have the benefit of growing tax free for 18 years, and will easily pay for their college one day.

Another thing some parents do is give their kids other assets — like stocks, bonds, or real estate. That’s nice, but it won’t grow tax free in the same way the 529 account will.

Less frequently talked about: you can also give your kids shares in startups that you own. If you give them a diversified portfolio of startups over time, that is likely to outperform a portfolio of conventional assets.

And here’s the twist: according to my understanding of US tax law, if you give your kids startup shares that are QSBS eligible — the QSBS status transfers with it. Which means when the startup eventually exits, their gains will be tax free on up to $10M per investment, per year.

So your kids will get the benefit of *both* faster appreciation, and zero taxes.

It’s nice too, because if you’re like most angels/VCs who struggle with liquidity (especially relative to aggregate paper net worth) — this is a way to give your kids valuable assets that will become liquid (and extremely valuable) by the time they are adults — without impacting your current liquidity in any way.

My wife and I have been doing this for many years now for our kids, and without getting too specific — I’ll just say that I am very happy with how this approach is setting them up financially.

Again: before you do anything here, please make sure to consult with a professional and do your own research. And if you choose to do it, make sure to document everything properly so it is all above board with US tax law.

PORTFOLIO

SUMMARY: $74M of capital invested over the last 6 years has grown into $182M of portfolio value. This is down vs. last quarter ($71M/$183M), due to more markdowns this quarter than usual. I.e. we added $3M of principal, but lost $1M of aggregate portfolio value — reflecting a total loss of $4M.

PORTFOLIO NEWS

Here are some noteworthy things that happened within the portfolio over the last quarter:

Zepto raised another up round at a $5B valuation. We invested from our rolling fund in 2021 on $13M post money. This is clearly our biggest winner from 2021 so far, and we think it’s likely to produce a strong investment multiple on that vintage. One more doubling from here will make it worth well over 1x all the money we raised in 2021 — which might happen soon, judging by how many offers we keep getting to buy out our shares. And then if it merely grows another 5x over the coming years — we think it’s plausible that this single investment could end up delivering a 5x+ return to our LPs on that portfolio vintage.

Jeeves is continuing to grow. The CEO recently said October was their best month this year, and their Brazil business is growing rapidly.

Stepful raised a very competitive Series B at a much higher valuation.

Albedo Space will launch its first satellite to space soon.

Revv raised a Series A led by Left Lane Capital.

Brellium raised a very competitive Series A.

Capi Money raised a Series A led by Creandum.

Taiv raised a Series A.

Constrafor raised a Series A.

Helius raised a large Series B at a much higher valuation.

AltScore raised a Series A.

PORTFOLIO STATISTICS

Our aggregate numbers, across every investment in UV’s life:

Numbers for only the UV Syndicate:

Annualized numbers for only the UV Rolling Fund:

Statistics about our investing activity:

Quarterly Breakdown for the Rolling Fund:

We rolled over $288k of uninvested capital from Q3, and raised another $1.25M in Q3.

If you are a major LP of Unpopular Ventures, have invested at least $250k to date, and are willing to sign an NDA, we will share the complete portfolio data with you. Please submit a request via this form: link

For everyone else, you can access the de-identified data here: link

Thank you for reading, and thank you for your support of Unpopular Ventures!

UV Quarterly Update: Q4 2024 was originally published in Unpopular VC on Medium, where people are continuing the conversation by highlighting and responding to this story.