Note: The original text is from deribit, written by Avi Felman and Adi Dommaraju.

“This is network effects, and without reflexive bubbles, network effects are impossible. The more we believe that the software and Internet century will be defined and governed by networks, especially if we believe that network participants will also have a part of it, then we will most likely Be prepared for a lot of reflexivity." - Alex Danco

The actual value of Bitcoin is in the hands of its holders, and many people try to apply valuation metrics to Bitcoin through on-chain metrics, S2F and other methods - but its core value comes from the acceptance of different groups of people around the world. The value of BTC comes from personal subjective value judgments on it. The more people who think bitcoin is valuable, the more valuable it is (so people often use Metcalfe's Law to value bitcoin).

This uncertainty in Bitcoin’s valuation is both an asset and a liability, as it can create a FOMO atmosphere as easily as it can convince people that it will go to zero. Individuals tend to value Bitcoin based on their surroundings, such as in a market where friends, family or a hairdresser are talking about the value of Bitcoin, they are more likely to accept the value of Bitcoin, but when they listen to it on TV By the time people talk about the "bubble bursting," they tend to avoid Bitcoin.

This investor behavior creates a strong reflexive nature. Once an uptrend starts, it is likely to continue until the number of Bitcoin participants reaches a climax, and the same is true for a downtrend. This dynamic can lead to wild market volatility from month to month and year to year, exacerbated by the prevalence of technical analysis as many in the market use the same simple metrics to evaluate Bitcoin.

Simply looking at the chart below can intuitively grasp the trending nature of an asset, and you can see fairly clear visual market trends with the naked eye.

As a trader, it is clear that trend following is a core component of winning in these markets due to the behavior of crypto investors. As with anything in life, it pays to look at data to test our empirical hypotheses.

There are a number of statistical tests that can be used to determine the autocorrelation (i.e., the nature of the trend) of a time series. One of the most basic tests of autocorrelation is the Hurst index, originally developed by Harold Edwin Hurst to determine the optimal dam size for the Nile, Now the index has been used for the more useful task of determining the autocorrelation of stocks.

In the table above, we show two-year Hurst prices for various assets, and unsurprisingly, Bitcoin emerged as the most autocorrelated asset. Stocks are relatively mean reverting, while oil and bitcoin show a higher probability of continuing trends during these timeframes.

The Hurst index of course doesn't tell the whole story, because it doesn't quantify the magnitude of the trend - so if an asset is only up 1 basis point a day, but every day it goes up, you'll find it has a very high Hurst index, which is useful for trading Assets are useful, but not helpful for understanding the size of the trend.

The easiest way to visualize the size of a return is through a histogram, the flatter the histogram, the greater the asset's average move.

From the graph above, we can clearly see that the returns of Bitcoin and Ethereum are relatively spread out, which indicates that the average volatility is large. Combined with the Hurst exponent above, we can make reasonable inferences that trends in cryptocurrencies are more general and larger than those in traditional assets, which should work well for trend-following strategies (rather than mean return strategy).

To further confirm the fundamental exploration above, we performed some backtests to explore whether simple trend-following strategies are indeed more effective in crypto markets than in traditional markets. We define the basic trend following strategy as follows:

1. Only do long;

2. Created a price Z-score of the 20-day moving average;

a. The Z-score is a standardized measure of distance from the mean, where 1.8 means the price has a standard deviation of 1.8 from the mean and 1.8 from the 20-day moving average;

3. Enter long positions when the Z score is higher than 1.8 (indicating that the asset has shown strong momentum);

4. When the Z value is lower than 1.8, exit the long position, indicating the loss of kinetic energy;

You can see above that crypto assets like ETH and BTC have better returns and a cleaner performance compared to traditional financial assets, crude oil and SPX. Our original theory seems to have some merit.

To make this more reliable on a higher timeframe, we tested Bitcoin's forward returns for a normal week, and the weeks after a large (>1std) positive move. As can be seen from the table above, after a strong week for Bitcoin, the following week is more likely to have above-average returns.

These tests bring us back to our original point about reflexivity. Unlike the cash flow producing companies that make up SPX or the physical oil that can be consumed as energy, crypto networks like BTC and ETH derive their value from usage and confidence, both of which are driven by emotion. In short, as Soros said, higher prices lead to better fundamentals, which in turn lead to higher prices. This concept of reflexivity is actually a very useful active trading tool.

This article comes from Tao of Yuan Universe, reproduced with authorization.

This article comes from Tao of Yuan Universe, reproduced with authorization.