Decentralized Retail Trading:
The Future of Finance and Retail Trading with On-Chain Strategies
By: Patrick Gorrell, Patrick Doyle, Roman Krasiuk

“We assumed we knew nothing about investing in digital assets and we sought to find the wallets that seemingly knew everything because they had access to information that nobody else had”

- Patrick Gorrell

TLDR;

Since our early inception, FRST has used on-chain data to find specific patterns of activity that would give us a competitive edge when investing in digital assets. The mantra our team followed was simple and humble. FRST believed the wallets with money knew more than we knew. In doing so, we set out to find the wallets that seemed to be moving the markets. Although we could not detect the trades that were occurring behind the scenes of centralized exchanges like Binance, Poloniex or Kraken, we could detect the wallets that were depositing and withdrawing the most value.

Using conventional analytics tools made it nearly impossible to conduct a historical search for the types of patterns we were looking for; so we had to make our own. Along the way, we were able to identify all of the largest depositors on exchanges and follow where they were originating their funds. FRST began working with Capital Markets Trading Digital (CMT Digital) in Chicago to understand the flow of value and the personas of trading activity behind the wallets. In doing so, CMT and FRST have successfully been able to target opportunities that would otherwise go unnoticed.

Identifying Funds Heuristic

When searching through the vast number of wallets that exist on Ethereum, there are very few tools available that provide the specific wallet details necessary for spotting funds. The idea behind the heuristic is to find the largest depositing and withdrawing wallets into centralized exchanges. We assume that many large funds take advantage of price differences between exchanges so it was very likely that we could spot the movements between their accounts. About 87% of the wallets on Ethereum exist between exchanges. This means that a majority of the activity on Ethereum is token transfers from one exchange to another. By identifying wallets that act as treasury wallets, you can begin to map out all of their customer’s deposit and withdrawal wallets.

“If you want to find the funds, just look for the large depositors.”

- Patrick Gorrell

Understanding how exchange wallets work

Most centralized exchanges manage funds on account-based chains differently from UTXO chain. Ethereum’s account-based model allows for more transparency into how exchanges manage funds for their internal treasury wallets and customer custodial wallets. There are 3 basic types of wallet personas that are associated with most centralized exchanges.

1. Exchange Custodial Wallets (Customer Deposit Wallets)
2. Exchange Treasury Wallets (Exchange Primary Wallets)
3. Exchange Withdrawal Wallets

The wallets that are most important to identify are the custodial (deposit) wallets of an exchange. By locating the large depositors, we can find that funds typically move large value transfers either from one exchange directly to another or, they move funds from a wallet outside of exchanges from a cold wallet.

The second most important type of wallet to follow is the withdrawal wallets. As stated before, roughly 87% of wallets on Ethereum exist in between exchanges. We believe this is due to a large number of wallets that interact with the exchanges and them trading assets on more than one exchange. This means that a majority of the time, a withdrawal wallet is, in fact, a custody wallet to another exchange.

Typically, there are treasury wallets that are specific to deposits & withdrawals as well as types of assets on Ethereum. Binance, for example, has multiple addresses that they use to source customer funds from custodial wallets and others that are used to manage withdrawals, whereas Poloniex has wallets for different assets like Ethereum-based REP, ZRX, BAT & STN.

Transactional Patterns

After working closely with many of the large firms in digital asset trading, our team soon discovered that most funds have something called a clearing function. A clearing function is a designated individual who clears the transaction for any deposit into an exchange. Additionally, we assume that funds would be very wary of moving large value transactions without sending an initial test transaction. An easy way to predict when a large transaction is about to occur is by looking for wallets that pass enormous amounts of value that send test transactions before every new type of asset allocation to the exchange.

Transaction Filtering

The above image shows transactions for all three wallet personas we have been discussing that are within the deposit, withdrawal & treasury proximity of Binance. The columns in the table are as follows:

1. Transactions ID
2. Age of transaction
3. Token
4. Amount of tokens transferred
5. USD Value of the transaction
6. From Address
7. From Exchange Proximity
8. To Address
9. To Exchange Proximity

By tagging every transaction on Ethereum with an exchange proximity tag, you can all transactions based on their proximity to a specific exchange.

• No From Proximity -> Binance(0) denotes a large deposit transaction
• Binance(0) -> Binance(0) denotes an internal treasury transfer
• Binance(0) -> No To Exchange Proximity denotes a large deposit

In the above image, we have filtered out all of the transactions to the last 6 months by their historical value greater than $10MM USD.

Hours of Activity

Since most funds operate within the traditional working hours, a telling sign of when the funds are operational is easy to explore by using a heat map to track the hours and days of the week that transactions occur. In the example below, a heat map is used to describe the transactions of a large fund’s custodial wallet depositing into Binance. The columns represent the respective time of the day and rows are the days of the week. The cross-section of dark spots indicates that this wallet is non-operational during normal US market hours. This fund is likely operating in Asia.

When comparing the above heat map to an exchange like Binance, you can begin to understand the total volume of activity that occurs within exchange wallets as noted in the image below. The darker the color of the heat map quadrant denotes the most active periods of the wallet compared to the lighter which signifies less active periods.

Historical Volume & Holdings

Once a target wallet has been identified, the next step is trying to understand what type of assets and volume have passed through the wallet historically. It is rare to find wallets that have a one-time use. Most often you will find wallets are used for multiple assets over their lifetime. By observing the token history of a wallet, you begin to trace the investment portfolio of the wallet and begin to identify the fund’s performance over time.

Historical Wallet Relationships

Ethereum activity acts much like a social network. Wallets only interact with other wallets, contracts or entities that they are familiar with. Many times you will find that large funds have moved their primary allocation to a new wallet. We assume it is either because of security or an attempt to create privacy. FRST has created tools that make it easier to follow the flow of value from one wallet to another. The chart below shows a flow diagram of value being passed from previous wallets and being distributed out to different exchanges.

If you expand the number of relationships, it becomes apparent that funds can move investments around frequently to take advantage of price points on different exchanges. By backing tracking out of the custodial wallets and observing wallets that lay outside the perimeter of an exchange more often than not, you will find cold storage wallets that are used to house token holdings. We call these perimeter wallets and they are typically Ledgers, Trezors or some other form of cold wallet storage. Mapping perimeter wallets allow you to identify investments such as exchanges, DApps, ICO’s, token sales, lock contracts and other types of decentralized products that these funds have participated in.

Conclusion

Finding funds starts with identifying wallets that are moving large dollar values into and out of centralized cryptocurrency exchanges. Once those wallets are identified we can begin to analyze their historic holdings to understand the wallet's investments and projects & protocols that they are interested in. Then we analyze the wallets hours of normal activity. By doing this we can start to identify the wallet's geographical location. Finally, we study how the fund wallet has moved its tokens between different wallets. This is crucial to clustering other wallets that may still be owned and managed by the fund.

Tell Me More about FRST

The FRST platform was built with professional traders, data scientists, and blockchain researchers in mind. Our current customers include major digital asset exchanges, hedge funds with dedicated crypto desks, venture capital funds, audit/compliance/KYC functions within institutions, and university and government research efforts in the blockchain space.

If you are interested in getting in touch with us or learning more about our team, please visit frst.com or email info@frst.com. We are a friendly group of motivated folks, and we are always looking to add more rock stars to our team and community.

Identify Large Fund Wallets on Ethereum using FRST’s On-Chain Analysis was originally published in FRST on Medium, where people are continuing the conversation by highlighting and responding to this story.

Decentralized Retail Trading:
The Future of Finance and Retail Trading with On-Chain Strategies
By: Patrick Gorrell

“We cannot tell you what tokens to buy, but we can tell you what the traders with the best track records are buying, where they are buying and when they cash out”

- Patrick Gorrell

TLDR

Only two things matter in portfolio management: the speed at which you receive your information and the speed at which you can act.

At FRST, we are observing and communicating on-chain activity. In doing so, we help our customers see, analyze, test, model, and even imitate the best attributes of digital asset portfolios. This appeals to users who are less familiar with digital assets or with the networks upon which they travel; it allows them to understand the unique mechanisms of the blockchain environment and to learn from pertinent example actors and example portfolios.

For those with the resources, curiosity or inclination, FRST is a powerful suite of world-class tools to enable the synthesis of alpha generating strategies. For portfolio managers who prefer turnkey solutions, we offer real-time insight into the aggregated activity of the wallets with the highest Sharpe ratios

Background

FRST’s co-founders spent years understanding social trends and the power of transparency in both actions and pricing. The fundamentals behind FRST came from observing and developing the largest influence management and pricing platforms in the world at the time. We were tasked with developing statistics used to price specific influencer posts based on their audience demographics. In short, we would graph the entire audience of someone like Kim Kardashian to find out how many mothers between the ages of 18–25 would consume a single post worth $400k.

Beginning in October 2016, my co-founder Jonas Frost and I began tracking wallets on Ethereum that behaved like venture capital. We would diligently aggregate transfer data that would allow us to follow these wallets and assume what price they were getting their positions at. During the ICO craze in 2017, we had become so efficient in tracking these wallets that we were able to take positions with better pricing right behind them and could also begin finding other trading behavior they participated in. Jonas and I funded FRST with the proceeds of trading using on-chain data and personifying the wallets of different blockchains to understand what their behavior meant. We then set out to share the strategies we developed, knowing their success is fundamentally tied to the number of people that would follow them.

Hypothesis

Modern developments of social media have made it hard to avoid the relevance of reflexivity and the role it plays in price and demand. Defining a model to evaluate the responsiveness of markets when information is shared about and the observable movements of money from one position to another is now a possibility. Analytics quantified with transparent markets provides a perfect opportunity to observe the reactions of price and social response as time elapses.

“Timelessly valid generalizations that can be used reversibly to explain and predict economic phenomena.”

— George Soros, The Crash of 2008 and What It Means: The New Paradigm for Financial Markets, p. 53

Reflexivity

A vast majority of investment decisions do not come from qualified advisories. Because of the nature of modern and social media, it is safe to assume that the market is more reactive than proactive as well as reflexive and uneducated. Cause and effect share a circular dependency that amplifies the supply and demand for goods and services. Influencers in the blockchain and cryptographic currency domain tend to have a bias, and it drives their motivations to share decisions that they make. Acting on incomplete information, a herd mentality follows under the influence of these market motivators to drive boom and bust cycles, such as the ICO craze of 2017.

“I can state the core idea in two relatively simple propositions. One is that in situations that have thinking participants, the participants’ view of the world is always partial and distorted. That is the principle of fallibility. The other is that these distorted views can influence the situation to which they relate because false views lead to inappropriate actions. That is the principle of reflexivity. For instance, treating drug addicts as criminals creates criminal behavior. It misconstrues the problem and interferes with the proper treatment of addicts. As another example, declaring that government is bad tends to make for bad government.”

— George Soros

Because of inherent properties of reflexivity, combined with the low barrier of entry into blockchain markets and shadowed price fluctuations, ICOs tend to multiply at a much greater rate than most independent blockchains. Risk is much greater when influence drives token values more than the solution that’s been created to address the underlying problem.

Value Drivers

In reflexive markets, the key driver is exposure and believability. In the industry of social media, we call this social currency. If an action is taken by a specific actor of influence (i.e., a top VC Investor, with a track record of proven performance, decides to buy $1MM of Ethereum with 80% of their portfolio), then it is safe to say that they feel roughly 80% confident that Ethereum will gain value over other asset or tokens they hold. These fundamental types of analysis are powerful drivers because they are easy to explain and to share with others.

Incentivized Performance

Naturally, investors tend to share their wins and hide their losses. With on-chain performance metrics, the influential traders are likely to perform better based off of their historical activity. It is very likely that certain performers can game their on-chain signals for personal profit, but the likelihood that they will continue to capture an audience is very low. To continually benefit from your trading decisions as an influential trader, you will constantly be measured against the performance of others.

Network Multiplier

Similar to the rate at which social media platforms grow, decentralized strategies can grow in similar virality. There is no proprietary trading mechanism that is required. The more exposure a strategy has, the more volatile it becomes with shorter frequency. Strategies with a large audience may experience more frequent price movement based on the activity. Supply and demand will create more periodic cycles of trade between professional, as well as retail, traders.

Creating Influential DeFi Strategies

Social influence is the most powerful form of marketing, as it directly ties an audience to a specific person of authority.

“The value of influence is derived from an ability to convey the value of an idea or thing to an audience that values their opinions.”

— Patrick Gorrell

By using on-chain verifiable instruments, it is possible to create indexes of value or authority that others can follow and invest behind. A well-formed public strategy will perform better as the number of people who participate in it grows.

Metcalfe’s law states “the effect of a telecommunications network is proportional to the square of the number of connected users of the system”. We can see this effect over time by simply using Google Search Trends compared to the price of Bitcoin.

Influence in ICOs Is No Longer a Factor

In 2017, overhyped ICOs led to the rapid rise and eventual decline in digital asset prices across the board. ICOs are the biggest proponent of these events due to their unregulated offerings and promises to investors. Large marketing budgets and lofty promises exacerbated this volatility, and today nearly 100%. Since 2017, nearly 100% of ICOs have failed to deliver, have dissolved, or have had no practical purpose in the market.

The people who are responsible for such actions are either being prosecuted or are no longer taken seriously by the average retail investor. As federal agencies begin to investigate these organizations, the transparency of blockchains makes it easy to follow the holdings of these individuals while they solicit their tokens to the greater public.

An Era of Transparent & Decentralized Finance

If given the opportunity to invest in a privatized traditional market product (Goldman Sachs or Vanguard) or an open and transparent product with a similar yield, it’s quite obvious that the open strategy would be the preferred choice. As the development of new finance products, such as MakerDAO and the subsidiary lending products that stem from these platforms continues to grow exponentially, the option to manage your own investment becomes more compelling every day.

Decentralized Retail Strategies

To understand the value of on-chain strategies, it’s important to first understand how they are derived and how to measure performance. A great example is a Maker DAO collateralized debt position, better known as CDP. A CDP is basically a credit card against your existing ETH. By depositing ETH into a CDP, you can draw a loan in the form of Maker DAO that is a stable currency meant to hold the equivalent of $1.00 USD value.

DeFi in Motion

Traders often use CDPs to borrow or to leverage against their assets to trade with. This allows them to maintain any upside of the underlying asset (ETH in this case) while still being able to trade with the value of it using DAI to purchase assets. CDP’s are similar to having a credit card against your ETH deposits and you pay interest on what you borrow.

FRST Trader Index Chart: FRST Platform

Above displays a chart of best performing CDP in the MakerDAO ecosystem. blue represents the leverage against the underlying asset, and orange represents the price of ETH at the same time. What you see in this image is a perfect leading indicator of confidence in the price of ETH. Using the leverage provided by MakerDAO, this trader was able to turn roughly $8MM dollars into $23MM in about 100 days by recursively buying ETH with its leveraged positions. The initial spike, in the beginning, was an initial draw used to purchase more ETH and further fund its position.

DeFi Strategies That Work

By analyzing all of the MakerDAO CDP’s, we can measure their performance among other related wallets that are bound to the ecosystem, such as Dharma and Compound, to create strategies that use 100% on-chain activity.

The two charts above are a backtested DeFi strategy created at FRST for retail investors to follow. The left indicates the strategies performance (blue) over the price of ETH (orange), and the right indicates the position (blue) versus the price of ETH (green). The wallet behavior and market data prove to be more than the sum of their parts in this case. Additionally, the personification of this activity makes it easy to group similar patterns of behavior to performance.

Strategies That Scale

As of today, there is no such thing as on-chain high-frequency trading. The best that can be afforded is network coverage and a highly reputed node that listens and broadcasts information to different networks. Pre-block rendering of transactions and annotations are key components to the future of decentralized finance and trading.

Currently, almost all trade is done between the top Cryptocurrency Exchanges. Strategies can be extended to any of the centralized or decentralized exchanges carrying assets that can be collateralized. Because these signals are public knowledge, anyone can verify the activity. Therefore, the determining factor in maximizing profitability is the delta between the point the signal is issued and the speed of the response.

Special thanks to:
Gregory Magadini, The player behind the initial insight @ FRST.

Gregory Berzin, Data Scientist @ FRST.

Hans HODL, Sr. Quant Researcher at Ikigai Asset Management for the help. https://twitter.com/hansthered

Using on chain data to find and follow the most performant traders on Ethereum was originally published in FRST on Medium, where people are continuing the conversation by highlighting and responding to this story.