The Dog Days of Crypto
Summary
→Remember the adage 1 year in human life equals about 7 years for Fido?
→Ever wondered how many days in crypto is equivalent to the S&P 500?
→Crypto investors are being baptized by fire.
As a child my parents taught me that one year in my short life was equal to 7 years to my best friend’s life. This was a beginning of explaining the sad existence of mortality to my young soul. How come my dog was fun and youthful when I was 4, was suddenly old and slow when I was 9?
The reason for this is that the speed of maturity or experience is different for us all. It is no different in the investment world. It is well known that bonds are far less volatile than equities, or in other words equity investors experience similar levels of volatility in shorter periods of time. However, with the introduction of cryptocurrency as an asset class, equity investing is no longer the most volatile of the public asset classes.
The idea for this fun research project came to me in bits and pieces. First, I experienced the decline of 20k Bitcoin down to 6k. While high volatility was expected, the short time frame of the decline (about 1.5 months) was drastic. Then during my day job as an equity portfolio manager I experienced the largest point decline in the Dow Jones Industrial Average ever in early February. However, this decline emotionally felt benign given my experience investing in crypto. This led me to explore the differences in volatility between crypto and equities, and the interesting topic of volatility equivalent time between the two asset classes.
The first step of this process was to get the daily returns for Bitcoin. Cryptocurrency data in general is scarce which makes quantitative analysis incredibly difficult. However, I was able to get the daily returns going back to 04/29/2013 through the “crypto” R package that is fed by Coinmarketcap.com. Daily returns are based upon UTC time.
The standard deviation of daily returns for Bitcoin is an astronomical 4.52% from 4/29/2013 through 3/4/2018. It is important to note that this this data is based upon close to close (or midnight to midnight), and volatility during the day can often be substantially higher.
If we assume Bitcoin’s return distribution is normally distributed (it isn’t) then we should expect to see daily moves greater than +/-4.52% about 2 to 3 times per week. The below chart details how often a 1,2, and 3 standard deviation event should occur.
Daily volatility has spiked YTD for Bitcoin to about 6.53%. Market sentiment has been swinging dramatically out of proportion with what the expected volatility suggests. A move of 5% and sentiment shifts to to the bulls and when Bitcoin drops 5% then it is expected that the bottom will fall out. I would caution into reading too much into very normal volatility (for this asset class).
As a sidebar this is further evidence that people need to be cautious in using the square root rule to annualize volatility. The S&P 500 has a daily volatility of 1.10% since 1988, which turns into 17.44% annualized using the square root rule (multiplying 1.10% by the square root of 252). 17.44% smells about right. If we used the same rule on the 4.52% daily Bitcoin volatility (multiplying by square root of 365 as crypto never sleeps) then we’d get 86.35%.
Back to Fido. Given we know the daily Bitcoin volatility is 4.52% I wanted to find out exactly how long of a daily time interval it would take a S&P 500 investor to experience that same sort of volatility.
The return of an interval is simply calculated using the value of the S&P 500 at the beginning of the interval and the end. Then I would repeat the return calculation using the same interval size until no more intervals could be created. For example for a 3 day interval I first calculated the return of 1/5/1988 through 1/7/1988 for the return for the first period for the 3 day intervals. Then I calculated the return for the 3 day interval of 1/8/1988, 1/11/1988, and 1/12/1988. I repeated the procedure through March 2018 resulting in 2537 independent 3 day intervals. At that point I took the standard deviation of those 3 day intervals. The result was an interval volatility of 1.86%. Rinsed and repeat for intervals of daily through 30 trading days.
The result is that one day of Bitcoin volatility is nearly equivalent to volatility of an interval of roughly 23 trading days for the S&P 500. So how much volatility experience have crypto investors racked up? It would take a S&P 500 investors about 100 years to experience similar volatility as a crypto investor would in just 3 years! Just to be clear, I’m solely talking about volatility experience and not overall experience or skill as an investor.
Given this was just a fun project and crypto is all about meme’s I’d like to close with one of my favorites.
While most equity investors raise their noses at the experience of crypto investors, perhaps it is time to recognize them for what they are: grizzled veterans.
