What Most of Us Don't Get About Fake Volumes
Before working for CoinMarketCap, I was a derivatives trader that became absorbed by the crypto markets at the tail end of 2018. As I was trading Bitcoin futures, I became aware of the burgeoning problem of volume inflation — or otherwise called “wash trading.” This was present in the crypto derivatives exchanges I was trading on and was more rampant in spot-exchanges that I saw listed on CoinMarketCap.
I recall being new to crypto and all, checking prices and the different exchanges on CoinMarketCap and wondering, “How are these exchanges getting so much volume?” Mind you, it was the crypto winter then, with Bitcoin falling from its all-time highs and enshrouded in a fog of pessimism. The volumes didn’t add up. I opened up the order books of exchanges with humongous volume, and there it was, clear as day, the time and sales firing away at prices in between the best bid and ask — a clear sign of wash trading.
The prevalence of volume inflation within the cryptocurrency space kept growing. CoinMarketCap, as the world’s most highly-trafficked cryptocurrency information website, was also being lambasted for showing these inflated volumes. It was during this time that the founder, Brandon Chez and investor of Hashtag.Capital (a fund that I was running with a few friends) reached out and sought our advice for solutions. I thus started my journey in CMC, researching and ideating ways to combat one of the most significant problems plaguing the cryptocurrency space.
The Problem With Using Volume in the Cryptocurrency World
The use of volume makes sense in a world with strict regulatory oversight — high volumes would indicate liquid markets with a good number of traders. However, if market malpractices like wash trading are not regulated — such as in the case of cryptocurrency — volume quickly becomes a poor indicator.
What about using computers to spot suspicious trading activity to flag inflated volume? After all, in the traditional markets, brokers and exchanges are able to flag wash trading by monitoring for patterns such as two accounts going back and forth on the same market for large amounts of trading volume. However, that requires access to account-ID data, and such information is usually kept private and confidential by exchanges. Third-party aggregators like CoinMarketCap would have no means of obtaining that information. Besides, brokers and exchanges in regulated financial markets are required by law to conduct market surveillance, unlike in cryptocurrency markets.
There is also the issue of scalability. While Bitwise did an excellent job with the report released in early 2019, flagging out errant exchanges, they did so with static data and isolated their study to only 81 exchanges and only Bitcoin market pairs. This is far from being representative of the entire cryptocurrency space, which consists of 300 exchanges and over 20,000 market pairs with varying degrees of volume, liquidity and market size.
To address the root of the volume inflation problem, the regulatory landscape has got to evolve to outlaw market malpractices. Yet, we know that the world is not going to regulate crypto exchanges efficiently overnight. So then, what can data aggregators do to get as close to addressing the volume inflation problem that continues to be so prevalent today?
Our Take on a Solution
What the cryptocurrency space needs is a sustainable solution that could crunch live data from more than 300 exchanges and over 20,000 market pairs with varying degrees of volume, liquidity and market size.
In the face of a lax regulatory landscape, with no access to account-ID data and having to deal with a massive number of exchanges and market pairs, the mission to solve volume inflation is an uphill battle.
Instead of looking at what was out there, the CoinMarketCap team and I went back to the drawing board to answer the very core question: “What are the factors necessary for good, organic volume in trading?”
It came down to two things — liquidity and traders. If volume indicates liquid markets and a high number of traders, the reverse must also hold.
Liquidity refers to the ease of being able to buy and sell an asset without affecting the price of the asset, and there are different methodologies to calculating liquidity. For instance, one could track orders at 1% or 2% of the depth of the best bid and ask price, or look at the entire order book in totality — but they each come with their limitations that could be easily gamed by errant exchanges.
Ultimately it needs to go back to what matters most to the retail users — minimizing slippage. The recent Liquidity Score by CoinMarketCap does just that. It showcases markets with the least slippage, empowering retail users to quickly identify exchanges and market pairs where they can trade most cost-effectively.
Estimating the Number of Traders
Estimating trader count is an even more challenging problem. The best party to report this figure would be exchanges — but to allow them full control over this metric without any way to verify would defeat the purpose of it. Errant exchanges that inflated volumes would surely inflate this figure.
The next best alternative would be web traffic data. As cryptocurrency trading is still retail-driven, the Web Traffic Factor (that combines a series of traffic data variables) becomes a good approximation of traders on exchanges, as yet another replacement for volume in a space where volume can’t be relied on.
Where It All Ends Up
By feeding a combination of the Liquidity Score, Web Traffic Factor and volume into a machine learning algorithm, the cryptocurrency space will derive a model that will give us yet our best chance at identifying exchanges and market pairs in the most objective, accurate manner possible today. This is our best chance at fighting volume inflation.