December 18th, 2017 — December 18th, 2018.
One year after having reached its all-time high at $19’891, the king of cryptocurrencies, bitcoin, is trading in the $3’500 area. The decline of $16’481 amounts to a loss of 82.50%.
High volatility has become a constant and we had to get used to large daily fluctuations. No cryptocurrency made it out of 2018 unscathed. The average decline was in the order of 90%. Ether lost its place to Ripple in terms of market cap and some cryptocurrencies that seemed very promising (e.g. ADA, XMR, MIOTA) slumped out of the top 10. The daily volumes of transactions, that had reached record numbers last January, fell by 25–30% on the main coins, while even going practically to zero on several minor tokens.
This scenario opened the way to a significant polarization of opinions. Sitting at opposite extremes we can observe the catastrophists on one hand and the resurrectionists on the other.
The catastrophists argue that the current situation was a foreseeable inevitable outcome, given the (lack of) intrinsic characteristics of cryptocurrencies (no store of value, no medium of exchange, no unit of measure, hence no currency and no intrinsic value).
Conversely, the resurrectionists affirm that this is just another correction like many others that have occurred in the past, some even more catastrophic (percentage-wise). Therefore, they firmly believe that the upward trend will resume.
However, at this point the main question, even before wondering if and when there will be a recovery and from what levels the price of the BTC will rebound, is if the entire cryptocurrency universe is about to disappear.
My answer is clear-cut: “No, the world of virtual currencies is not going to disappear”.
Because based on the last 12 months of data collection, fundamentals of the two largest public blockchains, Bitcoin and Ethereum, appears to be not only sound but even slightly improving. Fundamentals of a cryptonetwork boil down to the health of the supply-siders and demand-siders. Supply-siders are the folks who provision the network’s service (in a PoW blockchain a supply-sider is a miner), and demand-siders are the ones who use the service (common users).
In a recent article, Chris Burninske, who introduced with Willy Woo the Network Value to Transactions ratio (NVT) observed that fundamentals of legitimate cryptonetworks fell less than prices, and significantly so. For instance, in the case of Bitcoin, 59% of the network activity remains since the peak, while for Ethereum the figure is around 48%. The proxy of value for Bitcoin implies a 65% drawdown in network value is justified, well below the 82% decline in Bitcoin’s network value we currently witness in the market. In the case of Ethereum, the proxy of value implies a 77% drawdown in network value is justified, again much less than the 93%drawdown we currently see.
Bitcoin is currently processing ~250,000 transactions per day, and Ethereum ~500,000. Although the value of the respective native tokens issued on the Ethereum blockchain continued to slide over the last few months, the number of daily transactions remained stable. In numbers, Burninske reports that while BTC and ETH lost 82% and 93%, respectively from the peak, daily number of transactions only fell by 41% and 52% respectively.
Even though fundamentals may look healthy, an analysis based solely on them won’t produce reliable price predictions. In facts, hardly anyone (or no one) actually knows what will happen to the prices of the cryptoassets. Forecasting Cryptocurrencies price trends is particularly challenging due to the lack of a methodological framework to do so. Markets are then left under the influence of exogenous forces and are more exposed to manipulations. This are a typical traits of new asset classes in their early stages.
In the traditional financial world, predictions are based on two main analytical pillars: fundamental and technical analysis. In terms of investment styles, Momentum strategies, based on decades of data analysis, have assumed a key role and are concurring with quantitative investment strategies for the predominant position. In fact, algorithmic trading is widely used in every developed financial market and several emerging ones.
Unfortunately, these models do not work that well when applied to Cryptocurrencies, mostly due to a lack of historical data to feed into these analytical frameworks on one hand and to large price swings which makes smoothing techniques more difficult to apply.
This is an important issue since how quants like to say “garbage in, garbage out”, meaning that bad inputs automatically produce bad (even worse) outputs.
To wrap it up, Burninske observes that “in crypto we currently bicker not just over which models should be used, but which data should we input in whichever model will be chosen. This double issue caused the insane whiplash of the markets”.
STEPS TOWARDS A SOLUTION
The fundamental challenge of forecasting crypto-assets prices is that we have no clue about the factors that influence their price. Most crypto-analyst inferred that there is a correlation between the price of cryptocurrencies and market factors such as the behavior of fiat currencies, commodities and specific groups of stocks. Even though little has been done to validate those intuitions statistically, they were widely accepted since they made sense from logical point of view.
It turns out that those intuitions might be completely wrong. Two researchers from Yale University (professor Aleh Tsyvinski and Ph.D. candidate Yukun Liu), sought to “formulate and investigate potential predictors for cryptocurrency returns”. In their paper, they analyzed years’ worth of past price data for Bitcoin, Ripple and Ethereum.
Specifically, the paper focuses on the correlations between the price of cryptocurrencies with four main factors that have been universally assuming to play a role in their prices:
- Stocks Prices
- Fiat Currencies exchange rates
- Commodities prices
- Macro-economic factors (non-durable consumption growth, durable consumption growth, industrial production growth, and personal income growth)
Their study reached the conclusion that there is no statistically significant correlation between the prices of cryptocurrencies and the prices of stocks, currencies, commodities or macro-economic indicators. Tsyvinski and Liu raised the hypothesis that two cryptocurrency-specific factors seem to greatly affect the prices of cryptos, notably “momentum” and “investor attention effect”.
Momentum is a classic factor for almost every relevant asset class. Conceptually, momentum quantifies the propensity of an asset to continue to trend in the same direction, or to increase/decrease in value after it experience minor rises/declines. In simple terms, “If things go up, they continue to go up on average, and if things go down, they continue to go down”. The report found that if the price of bitcoin increased sharply over a week, it would be likely to continue to increase for the following week.
Investor attention can be summarized as the amount of interest and hype around cryptocurrencies. “It is measured by factors such as the number of online searches and posts by investors” the authors said, defining the latter as the “Investor Attention” factor.
The results make a lot of conceptual sense. However, it does not represent a disrupting discovery in the cryptocurrencies forecasting field. It could appear rather obvious to many market participants that the above factors would have played a significant role in short term price trends. But when we try to think to their practical use to implementing investment strategies, we do not reach any statistically significant conclusion.
Cryptocurrencies are a new asset class. Consequently, it is rather difficult to list them under any other existing asset classes. But, there are only few doubts about the fact that they are securities. Therefore, within years, cryptoassets could possibly walk through the gate of structured products’ universe.
May be just as an intermediary stop, before reaching adulthood and aspiring to sit on the throne left vacant by the unstoppable decline of fiat currencies.