Is the Crypto Currency “Market” about to have a Fakebook moment? That’s the moment that people realized that 90% of users, traffic, and content on Fakebook was bots, spam, and other fakes (and ultimately stopped using it). As we explained in Splitting Pennies, the financial world is not as it seems, but now it appears that Crypto is a mirage wrapped inside of an enigma. It appears data quality in the Crypto Currency world is much more fake than faked Government economic data (there is at least substantiation for it), the falsely inflated stock market, and other market anomalies. First let’s start with this article that originally appeared on Medium, from a Crypto trader that was suspicious about fill quality:
In this piece I will expose why I believe more than $3 billion of all cryptoassets’ volume to be fabricated, and how OKex, #1 exchange rated by volume, is the main offender with up to 93% of its volume being nonexistent. I’ll endeavour to prove it by analyzing publicly available data. When I set out to datamine for this piece, I had no idea I would end up talking about fake volume. I initially meant to gather data about cryptoassets liquidity, that could be a complement to volume. I thought it would prove an interesting indicator when assessing the value of an asset. Read the whole article here.
This article deserves a full read. The results are astounding. Now as far as independently confirming them, we haven’t done that – but it confirms what we’ve been experiencing ‘in the field’ so to speak. Without revealing much we’ve been developing what we can call Crypto HFT. This is now a running joke as HFT in Crypto is impossible. Crypto exchanges mostly use REST API which is for web servers:
Representational State Transfer (REST) is an architectural style that defines a set of constraints and properties based on HTTP. Web Services that conform to the REST architectural style, or RESTful web services, provide interoperability between computer systems on the Internet. REST-compliant web services allow the requesting systems to access and manipulate textual representations of web resources by using a uniform and predefined set of stateless operations. Other kinds of web services, such as SOAP web services, expose their own arbitrary sets of operations.
See an example here from ITBIT. Before we get into our REST bashing, REST is OK for many functions such as updating apps with data, web server interaction such as RESTful Web APIs and other HTTP related handling. REST is not designed for trading and is not appropriate to trade over. Just like it’s not appropriate to trade on your mobile phone, because data can be delayed, orders can be rejected, it’s just unprofessional. Why the exchanges offer REST API was at first puzzling but after reading this article and doing some background research it all started to make sense: Crypto really is a pump and dump scam from many angles. First, the exchanges inflate their volume to lure customers as if there is ‘deep liquidity’ when in fact, you can’t place an order for more than 2 or 3 BTC without getting fill issues. That’s not to say that ALL exchanges are fake, the article goes on to mention Bittrex is mostly clean. The point is that there is no ‘data source’ like there is for other markets, so it leaves many to rely on sources like coinmarketcap.com which are good sites but hardly institutional grade (sorry, dude). That’s why Bloc10 founded Total Cryptos, a site dedicated to Crypto Data, price Data, information, and more. The sad news is that we are only beginning to unravel this enigma. There is no Bloomberg in Crypto. So there is no way to really audit information from these exchanges except for this ad hoc statistical analysis done by this trader. And that’s not the only problem!
A straightforward data point - the total supply of bitcoin hit 17 million. But as with most things in crypto, it wasn't so simple. Every 10 minutes or so, miners find a block of transactions and the network adds 12.5 new bitcoin to the total supply as a reward for the finders. And each reward has been logged on the blockchain since bitcoin launched in early 2009. As such, it seemed like a number - a milestone - the industry could trust. But as some celebrated once the mark was hit on bitcoin data provider Blockchain's website, others took to Twitter to rain on their parade. Jameson Lopp, Casa engineer and the creator of Statoshi.info, another public-facing bitcoin data site, tweeted: "Today I've learned that a lot of data sources are incorrectly reporting the total bitcoin supply. We haven't actually hit 17 million BTC yet."
Again, without a central authority, there’s no way to really even start to calculate precise numbers. But what’s a few thousand BTC here and there? These 2 data issues are not related, but they both showcase a trend in Crypto that is plaguing markets and businesses. It can by why 77% of CIOs have ‘no plans’ for Blockchain in their organization. Blockchain is a great idea, but really isn’t ready for the Enterprise. So going back to our trading issue, what we found is that prices just simply disappear. Orders aren’t rejected – they just don’t exist. It’s difficult to quantify this phenomenon as there are no ‘error codes’ – the orders just don’t go through. Now, there are a number of groups algo trading Cryptos we aren’t claiming that it’s impossible to algo trade Cryptos, we are on the HFT side so that makes algo trading much more difficult. Simply that, the data is bad. Is it faked intentionally, or is it just bad because they don’t check it, or it’s because of bad programming – these are all possibilities and the straight answer is we would only know the answers to these questions if they opened up their operations to us which they won’t do. What we can attest to is that this is eerily familiar to the days of FX algo trading in 2006 – 2010 when dealers were ‘taking the other side’ and allowing demo accounts to do really well which lured customers to make deposits only to find out that the live environment was nothing like the demo. In one memory we remember having our account put on the ‘losing’ server (which is the live server that never wins) and after complaining to our high profile connection they moved us to the ‘winning’ server and guess what our strategy did ok. But, these FX dealers, although unscrupulous, were lightly regulated, and finally the practice was ended (in some cases due to customer lawsuits like the Ass-price-slippage case against FXCM). Meta Trader 4 allowed brokers to build a business based on the false hope of having an algorithmic strategy ‘automatically’ rake in the profits for you, as the MT4 demo was setup such that it was easy to ‘hack’ the backtester to create fantastic looking results on poor quality data that would be impossible to achieve in real market conditions. Specifically, we refer to this image on GitHub: Sounds like the last 12 years, fellow FX traders – live results not matching up with demo results. But here we have maybe found the culprit – bad data. Unfortunately, there isn’t really a solution for this problem, at least one that individual traders & investors can solve. The practical solution is when institutions start getting into Crypto, like Peter Thiel announced recently.
tZERO is about to release a Securities Crypto trading platform, set to revolutionize Crypto trading on Wall St. The likely culprit is probably a mix of both 1) Poor technology and even more poor developers hired with an hourly rate as low as possible with no experience building trading systems and 2) fluffing up the data to make it look more ‘institutional’ to lure investors who might have some extra funds in their 401k for example. It’s something to be analyzed further and certainly a big warning to all Crypto day-traders to take into account in their strategies, and likely the data quality decreases on the less liquid, less popular Crypto pairs.