<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Arnold Tang on Medium]]></title>
        <description><![CDATA[Stories by Arnold Tang on Medium]]></description>
        <link>https://medium.com/@arnold.tang?source=rss-12d3f45f809f------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*yBXIvpCuYzM5bKHZDTl9Ng.png</url>
            <title>Stories by Arnold Tang on Medium</title>
            <link>https://medium.com/@arnold.tang?source=rss-12d3f45f809f------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sat, 21 Nov 2026 15:14:07 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@arnold.tang/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Under the Radar #2: MELI, Latin America’s Alibaba & Amazon]]></title>
            <link>https://medium.com/@arnold.tang/under-the-radar-2-meli-latin-americas-alibaba-amazon-0a6e64b9fd4a?source=rss-12d3f45f809f------2</link>
            <guid isPermaLink="false">https://medium.com/p/0a6e64b9fd4a</guid>
            <category><![CDATA[value-investing]]></category>
            <category><![CDATA[stock-market]]></category>
            <category><![CDATA[mercadolibre]]></category>
            <dc:creator><![CDATA[Arnold Tang]]></dc:creator>
            <pubDate>Thu, 28 May 2026 04:59:02 GMT</pubDate>
            <atom:updated>2026-05-28T04:59:02.293Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/300/1*9ezi1HaAcrPazLnPOh6Jbg.png" /></figure><p>MercadoLibre. The name keeps coming up in conversations about international compounders. The shorthand people use is “the Amazon of Latin America.” That label isn’t wrong, but it misses half the picture — because MELI is also the PayPal, the logistics network, and the credit card issuer for a region where a big chunk of the population has never had a bank account.</p><h4>Two Businesses, One Flywheel</h4><p>Mercado Libre runs an e-commerce marketplace — third-party sellers, commissions, the usual playbook. But what makes it different is the logistics. In a region where delivery infrastructure is a mess, MELI built its own. Planes. Sorting centers. Same-day delivery in major cities. It’s the marketplace of Amazon with the delivery network Amazon had to spend two decades building.</p><p>Then there’s Mercado Pago. It started as the checkout button for the marketplace, and it grew into something much bigger: online payments, credit cards, wealth management, insurance. Bank coverage in Latin America is low. For a lot of people, Mercado Pago is the first financial tool they ever used. That kind of trust doesn’t transfer easily to a competitor.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*CX68SslQlwLXnvsokjWExA.jpeg" /></figure><h4>Three Markets That Matter</h4><p>Brazil drives more than half of revenue. It’s also where the fight is ugliest — Amazon and Shopee are pouring money into logistics right now. Mexico is the growth engine, and Argentina is the home market in the rearview mirror, still churning through inflation cycles where users treat MELI’s money market fund as a hedge against the peso.</p><h4>What Protects the Business</h4><p>The moat isn’t one thing. It’s four things stacked on top of each other.</p><p>First, bank penetration. In a region where large chunks of the population are unbanked, MELI is often the first point of contact with any kind of formal financial service. That’s not a customer acquisition channel — it’s a relationship that predates every competitor.</p><p>Second, logistics. You can’t just show up in Brazil or Mexico and deliver packages next day. MELI spent years building out its own network — planes, distribution centers, last-mile coverage. A new entrant spends billions and still takes half a decade to catch up.</p><p>Third, data. MELI has been watching Latin American consumers spend, borrow, and repay for over two decades. In credit underwriting, that historical data set is the moat within the moat. It lets MELI price risk in ways no newcomer can match — which means better margins on lending than any bank trying to build the same book from scratch.</p><p>Fourth, tax and compliance. Every country in the region has its own labyrinth of tax codes, import rules, and consumer protection laws. Amazon has the money to compete. What it doesn’t have is two decades of navigating Brazilian tax court. That institutional scar tissue is expensive to replicate.</p><h4>The Numbers That Tell the Story</h4><p>Revenue has been growing at roughly 40% a year for three straight years. Q1 2026 came in at 45%. That’s not decelerating — it’s actually picking up.</p><p>Profit is the tension point. In 2023 and 2024, net income roughly doubled year-over-year. Then it flattened. Amazon and Shopee started spending aggressively to gain ground in Brazil, and MELI had to match. Margins compressed. The market punished the stock — it’s down roughly 40% from its highs.</p><p>Forward P/E sits somewhere between 32 and 41x, which is cheap by MELI’s own history. The stock has rarely traded at these levels.</p><p>There is one number that bothers me, and it’s the bad debt rate. MELI has been expanding its credit card book aggressively, and the 90-day delinquency rate has risen to roughly 17.6% — well above industry norms. If a regional recession hits, that number gets worse before it gets better, and it could eat into earnings just when the market is already punishing the profit compression from competitive spend.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*9u06kpxdE3YDB5CR8jJ-UA.png" /></figure><h4>Why I’m Watching, Not Buying</h4><p>I don’t have a position. The valuation is tempting — Forward P/E near the bottom of its historical range, revenue still growing at 40%+. But I’m looking at a company fighting a war on two fronts: Amazon and Shopee squeezing margins on one side, and a consumer credit book showing early signs of stress on the other.</p><p>The new CEO took over in January 2026, right as the competitive pressure peaked. I want to see at least one more quarter — ideally two — before I decide whether that profit stall is temporary. If margins stabilize and bad debt trends improve, the current price might look obvious in retrospect. If they don’t, there’s plenty of time to stay on the sidelines.</p><p>This is a company I want to own. But wanting to own something isn’t the same as knowing when.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=0a6e64b9fd4a" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Under the Radar #1: AXON is Not a Taser Company Anymore]]></title>
            <link>https://medium.com/@arnold.tang/under-the-radar-1-axon-is-not-a-taser-company-anymore-283e37101777?source=rss-12d3f45f809f------2</link>
            <guid isPermaLink="false">https://medium.com/p/283e37101777</guid>
            <category><![CDATA[stock-market]]></category>
            <category><![CDATA[axon]]></category>
            <dc:creator><![CDATA[Arnold Tang]]></dc:creator>
            <pubDate>Mon, 25 May 2026 09:22:02 GMT</pubDate>
            <atom:updated>2026-05-25T09:26:19.937Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/221/1*IdgAK1luGgZYvtFfZaneIg.jpeg" /></figure><p>AXON used to be called Taser — you know, the gun that shoots electricity in every cop movie. In 2026, that’s maybe 20% of the story.</p><p>The real business: a cop wears an AXON body camera and carries an AXON Taser on shift → video auto-uploads to AXON’s cloud → AI writes the first draft of the incident report → prosecutors and courts pull evidence from the same platform. From the street to the courtroom, the entire workflow runs on AXON. No competitor even comes close.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*BDBZvr65nNzqIZQWwfF8pw.png" /></figure><h4>Five Things Protecting This Moat</h4><p>1. Years of police video evidence already locked in AXON’s cloud. Switching would mean migrating terabytes of chain-of-custody data. Nobody’s volunteering for that project.</p><p>2. AI report generation isn’t a nice-to-have anymore. The US is short on cops. Officers depend on AXON’s auto-generated drafts to clear paperwork. Going backwards isn’t an option.</p><p>3. Contracts run 5–10 years. Even if a competitor shows up tomorrow with a better product, they can’t get a foot in the door until the next procurement cycle.</p><p>4. It’s not just police anymore. Prosecutors, courts, and public defenders are on the same platform. AXON stopped being a tool and became infrastructure.</p><p>5. They keep stacking new AI products on top — counter-drone, automated surveillance flagging, VR training for officers. Each one tightens the bundle.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*5Uk8mCnRRqrBfEYNWFRltw.png" /></figure><h4>The Number That Matters Most</h4><p>AXON disclosed $14.3 billion in total future contracted revenue as of Q1 FY26. Full-year guidance is roughly $3.6 billion. Do the math — that’s about four years of revenue already locked. The booked-but-not-billed backlog grew 44% year-over-year.</p><p>Revenue Growth (past 3 years + current quarter):</p><p>FY2023: $1.56B +31.5%<br>FY2024: $2.08B +33.4%<br>FY2025: $2.78B +33.5%<br>Q1 FY26: $0.81B +34.0% (just one quarter)</p><p>The growth rate hasn’t budged from ~33%. That’s rare at this scale.</p><p>Net income tells a messier story:</p><p>FY2023: $174M<br> FY2024: $377M (+115%)<br> FY2025: $125M (-67%) ← SBC + Dedrone acquisition costs<br> Q1 FY26: $169M (+92%)</p><p>FY2025 looks broken, but it’s accounting noise — a one-time CEO performance award and integration expenses from the Dedrone acquisition hit the GAAP line. Q1 FY26 already snapped back.</p><h4>Q1 FY26 Earnings — Quick Hits</h4><p>- Revenue $807M, +34% YoY (beat 27–30% guidance)<br>- GAAP net income $169.3M, +92% YoY, 21% margins<br>- Adjusted EBITDA $201.6M, 25.0% margin<br>- Net revenue retention 125% — existing customers aren’t just staying, they’re expanding<br>- International hit 20% of revenue (up from 12% last year), likely hitting the 25% target years early<br>- Full-year guidance raised from 27–30% to 30–32%</p><h4>The Number That Bothers Me</h4><p>Stock-based compensation is running at roughly 17% of revenue. Annual SBC guidance: $590–620M.</p><p>Quick sanity check: hardware companies should be under 5%. Healthy SaaS is 5–12%. Above 15% means “adjusted profits” are masking real dilution. AXON’s Q1 included a one-time CEO performance grant — without that, it’s probably closer to 12–13%. I’m watching Q3 to see if that number comes down. If it doesn’t, the “hardware + SaaS” narrative has a crack in it.</p><p>Gross margin compressed 200bps last quarter (63.6% → 61.6%) from tariffs, Dedrone integration, and professional services costs. Software margins are still above 80%, but hardware + Dedrone drag the blended number down. Worth tracking — another quarter of compression and the story shifts.</p><h4>Why I’m Not Buying Yet</h4><p>Forward P/E sits around 47x (GAAP, per IBKR). If I strip out the one-time SBC spike, non-GAAP is closer to 40–42x.</p><p>I set a simple trigger for myself: IBKR GAAP Forward P/E under 40x, assuming the business fundamentals haven’t changed. At roughly $380–400 per share, I need another ~15% pullback or a couple quarters of earnings growth to close the gap.</p><p>Two things would make me reassess:<br>- Price drops into the 40x range on IBKR’s display<br>- Q2 FY26 earnings in August deliver better margin or SBC numbers</p><p>For now, I’m on the sidelines. But AXON is the most interesting stock I’m not holding.</p><p>This is not investment advice. I don’t own AXON. This is me thinking out loud.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=283e37101777" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Costco at 55x Earnings: Still the Perfect Business?]]></title>
            <link>https://medium.com/@arnold.tang/costco-at-55x-earnings-still-the-perfect-business-594c7deeacba?source=rss-12d3f45f809f------2</link>
            <guid isPermaLink="false">https://medium.com/p/594c7deeacba</guid>
            <category><![CDATA[stocks]]></category>
            <category><![CDATA[costco]]></category>
            <category><![CDATA[charlie-munger]]></category>
            <dc:creator><![CDATA[Arnold Tang]]></dc:creator>
            <pubDate>Fri, 22 May 2026 02:52:54 GMT</pubDate>
            <atom:updated>2026-05-22T02:52:54.018Z</atom:updated>
            <content:encoded><![CDATA[<h4>A personal checkup on the four moats — and what I think the stock is actually worth</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/318/1*z-IIYnQ9Ga2n__votq1-Hg.jpeg" /></figure><p>I’ve held Costco for over two years. It’s 29% of my portfolio. The stock went sideways for a long time, then rallied. I caught myself wondering: is the thing Munger called “the perfect business” still perfect?<br>Not “should you buy now.” Not “is it overpriced.” The real question for someone who already owns it: do the four pillars of Munger’s original thesis still stand in 2026?<br>Here’s what I found.<br><br><strong>The Four Pillars, One By One</strong><br><strong>1. The Membership Flywheel — Stronger Than Ever</strong><br>North American renewal rate: 92.9%. Global: 90.5%. Paid members: ~78 million, still growing 6–7% a year. Membership fees now kick in roughly 70% of operating profit. That’s the flywheel Munger described — near-pure margin, recurring, locked in.<br>But the big test happened in September 2024. Costco raised Gold Star from $60 to $65 — first hike in seven years. Renewal rates didn’t budge. I’ve read about “pricing power” in a hundred analyst reports. Seeing it actually exercised without a scratch on retention is something else entirely. This pillar isn’t just intact. It’s hardened.<br><strong>2. SKU Discipline — Same Formula, Deeper Leverage</strong><br>Still ~4,000 SKUs. Still 12–13% gross margin. The change worth tracking is Kirkland Signature: from 25% of sales in 2015 to 33% today. Every supplier on a Costco shelf now competes not just against Costco’s volume, but against Costco’s own brand on the adjacent shelf. If Costco makes its own version of your product, you lose the account — not just some margin. That’s a negotiating position most retailers can’t touch.<br><strong>3. The Give-It-Back Culture — Fine Today, One Number to Watch</strong><br>$1.50 hot dog. $4.99 rotisserie chicken. Both survived the inflation spike unchanged — these are cultural totems, not P&amp;L decisions. The margin gap vs. Walmart (~24%) and Target (~28%) is still the cleanest translation of the give-it-back philosophy into numbers.<br>What I’m watching: Costco has started building retail media — Costco Next, early ad products. If this scales, it pulls the company toward Amazon’s model, where supplier ad spend eventually shows up in shelf prices. My tripwire is global gross margin breaking 13%. That’s the earliest measurable signal of cultural drift. Hasn’t happened yet.<br><strong>4. Management — Smooth Transition, Recession Untested</strong><br>Ron Vachris took over January 2024. Forklift driver to CEO in 36 years, all at Costco. No splashy hires, no M&amp;A fantasies, no change in store expansion pace. The Jelinek playbook is running undisturbed.<br>The honest caveat: he hasn’t faced a real recession yet. Jelinek steered through the pandemic. Vachris has had calm seas. So far, so good — but the real exam hasn’t arrived.</p><p><strong>The Balance Sheet — Something I Rarely See Discussed</strong><br>Most Costco analysis skips the balance sheet because the story is about the membership model. But here’s what’s actually on it:<br>Cash and equivalents: ~$11.2 billion. Long-term debt: ~$6.5 billion. That’s net cash of close to $4.7 billion. Interest coverage is comfortably above 30x operating income. The company could pay off every dollar of debt tomorrow and still have billions left for store expansion.<br>Free cash flow runs around $7–8 billion annually. Costco returns essentially all of it — roughly $1.7 billion in dividends, the rest in special dividends and buybacks. It doesn’t hoard cash. It doesn’t over-lever. It just steadily returns capital while growing the store base 25–30 locations a year, mostly self-funded. This is the kind of balance sheet you want when retail enters a downturn — no liquidity risk, no covenant risk, no need to slow expansion to service debt.<br><br><strong>My Own Valuation — Not Just a P/E Comparison</strong><br>Saying “55x forward P/E is above the 5-year average of 38x” isn’t valuation. It’s a multiple comparison. Here’s my attempt at a real estimate.<br>I start with the economics: ~7–9% revenue growth, built on 2–3% same-store sales, 2–3% net new warehouse growth, and a modest tailwind from membership fee increases every 5–7 years. Operating margins should stay flat at around 3.5–3.7% — Costco doesn’t expand margins by policy. Share count declines about 0.5–1% a year from buybacks.<br>That gets you earnings growth in the 9–12% range — revenue growth plus a tiny buyback tailwind, no margin expansion assumed. For a company with this moat durability, I apply a 22–25x terminal multiple on earnings five years out. Discounted back at 9% (my hurdle rate for a lower-risk compounder), I get a fair value range of roughly $880 to $1,080 per share.<br>The stock is currently around $1,048. So by my math, you’re paying somewhere between full fair value and a modest premium. You’re not getting a bargain. You’re paying for quality — and hoping the moat is even more durable than the 22–25x terminal multiple implies.<br>If I’m wrong and Costco deserves a 28–30x terminal multiple because the membership model is genuinely irreplaceable, then fair value moves to $1,150–1,300. If I’m wrong in the other direction and growth slows to 5–6%, fair value drops to $750–850. The range is wide — and that’s the honest answer. Anyone giving you a single number is overconfident.<br><br><strong>What Could Actually Go Wrong</strong><br>Not a generic risk list. Here are the scenarios I think about, with rough probabilities attached.<br>Tariffs (high probability, medium impact). Roughly 30% of merchandise is imported. A sustained tariff cycle squeezes margins across retail, but Costco’s relative price advantage actually widens — its margins were already the thinnest, so every competitor gets hit harder. This is a near-term drag, not an existential problem.<br>Amazon (low probability, low impact). Ten years of trying, zero dent in Costco’s membership economics. The reason isn’t price — it’s psychology. Amazon Prime charges you for convenience. Costco charges you for access to value. A Prime member doesn’t feel compelled to “get their money’s worth” each month. A Costco member thinks: I paid $65, I’m going back. You can’t spend your way into that behavioral loop.<br>Digital disruption (medium probability, long-term impact). Costco’s e-commerce penetration is low. Sam’s Club — especially in international markets — runs circles around it on mobile ordering and same-day delivery. This isn’t a next-quarter problem. It’s generational. If younger members start renewing at lower rates, the model faces a secular challenge. Current data (92.9% overall) doesn’t show it. But I want renewal data broken out by age cohort, and Costco doesn’t provide it. That absence itself is worth noting.<br>Management failure (low probability, high impact). Vachris has been fine. But his next CEO — whenever that happens — is the bigger question. Three CEOs in 40 years is a miracle of continuity. The fourth one, whenever they arrive, is the test of whether the culture survives the founder generation.<br><br><strong>Why I’m Not Selling — And Not Buying More</strong><br>At 55x forward earnings, Costco is pricing in a lot of future excellence. When you pay that multiple and the company delivers 7–9% growth, you’re likely looking at 0–10% annualized returns over the next 3–5 years. That’s the honest math. It doesn’t mean the business is broken. It means the stock already reflects the good news.<br>But I’m not selling. The moats I checked are intact. The balance sheet is fortress-grade. The thesis hasn’t broken — and selling a compounder because “it’s expensive” is exactly the mistake Munger warned against: trying to be rich twice.<br>My five tripwires — things that would actually make me sell:<br>1. North American renewal drops below 91%<br>2. Global gross margin breaks above 13%<br>3. Kirkland Signature share falls two years in a row<br>4. CEO departs or company does a major acquisition<br>5. Total members turn negative year-over-year<br>None of these have triggered. So I hold.<br><br><em>Disclaimer: This is not investment advice. I own Costco and this is me thinking out loud about a position I manage for myself. Do your own work.</em><br> — <br><em>I publish research on the stocks I own — and the ones I’m thinking about owning next — on Substack: arnold784722.substack.com</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=594c7deeacba" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>