Media is misinformed about money, bitcoin and volatility

You might have seen the latest headlines about the Winklevoss twins, Co-Founders of Gemini Trust, claiming that “bitcoin is better at being gold than gold.” These quotes are from a Bloomberg interview from June 6, 2016. “...bitcoin has been a lot more stable than gold. It has evidenced a lot less volatility.” And as evidence they cited “…the Wall Street Journal actually wrote a story in May in how it was more stable than gold over the last 28 days.” And finally ”It’s actually better at being gold than gold.”

The problem is, none of that is actually true. And it seems that nobody in the media made any effort to fact check their claims either. Because if they actually looked at the data, they would have realized that bitcoin’s price volatility is several orders of magnitude higher than that of gold over pretty much any timeframe (including the entire if rather short history of bitcoin).

So let’s start by simply plotting the daily price changes of gold vs the daily price changes of bitcoins. This gives you already a pretty good understanding which one is more volatile and should give you a hint what you probably will find if you bother to do an actual statistical analysis.


Volatility: Gold vs bitcoin


Running the numbers, we find that the standard deviation of bitcoin is much higher over any time-frame we looked at. The annualized standard deviations of daily returns over 1 week, 1 month, 3 months, a year, 3 years and since 2010 (the start of bitcoin price history in Bloomberg) are all a lot higher for bitcoin prices compared to gold. None of this data is particularly hard to get and calculating the standard deviation can be done in a few seconds either using Bloomberg’s built-in analytics or in an ordinary Excel spreadsheet.


Volatility 2: Gold vs bitcoin


So what about the WSJ story where the newspaper apparently reported in May that Bitcoin had been more stable than gold for 28 days? We are not really sure. There was article in the WSJ on April 19, 2016 with the title “Is Bitcoin Becoming More Stable Than Gold?” In that article, the Author (Stephanie Yang) claims that “The last 24 days mark the longest period in which bitcoin prices have been less volatile than gold prices, going back to 2010”. The article doesn’t exactly state what that means. Looking at the daily absolute price changes in percent over the given time-frame (24 business days from March 16 to April 19), one can see that the changes in bitcoin prices exceed those of gold on several occasions and vice versa. The annualized standard deviation of gold and bitcoin over that exact timeframe are both about 16%, with bitcoin’s volatility being a fraction of a percent lower which might be what has prompted the author to write that story. But then, the standard deviation of a short 24 days of daily returns is not something you typically find in a fund prospectus. Nobody will ever say, “I see the backtesting of the returns of your product shows that the 24-day standard deviation was unusually low. That is fantastic.” It is a completely arbitrary time-frame. Needless to say that the 24-day rolling annualized standard deviation of daily returns (hard to imagine this ever catches on) is now back in the high 40s.


Daily returns: gold vs bitcoin


We are not saying that bitcoin’s price volatility cannot fall below gold’s volatility at times. But we would say that it should happen first before you claim so. Calculating the standard deviation of returns is easy and thus any responsible media source should fact-check such claims before simply reporting it without comment. Price returns and volatility are pure math. It’s not like a lot of areas in economics where econometrical tools are abused to “prove” all kinds of things people want to believe are true. There is no interpretation of the results, they are what they are. To derive from a 24-day window that bitcoin is a better store of value than gold after having experienced >50% daily price swings over the past few years is nonsense. By way of comparison, gold has been used as money for 6,000 years and we have great data for daily pricing going back 600 years. Looking at that data, the maximum price swing in any given day was 32%.And that was directly linked to the US government sharply devaluing the USD. Since the official demonetization of gold in 1971, the largest daily price move was 13%. 


If you watch the full Bloomberg interview you also realize that the Winklevoss twins don’t seem to really understand what gold is, what a store of value is, and thus what money is. Gold is simply counter-party risk free money that maintains its purchasing power over time. What makes it so useful as money is that prices of goods that the average consumer needs have remained stable in gold over very long time periods. A BigMac costs pretty much the same in gold today as it did 30 years ago (while in USD it has gone up 300%). The same is true for the average house, car, a gallon of gasoline, etc. Despite referring to gold (and implicitly bitcoin) as money, the Winklevoss twins imply in the interview that gold and bitcoin are things INVESTORS seek in times of crisis. But gold is not an investment. Stocks and bonds are investments, where the investor takes a view on the company and the management and is hopefully rewarded with a return. Gold is simply money, and as a store of value works much better than any fiat currency ever has over a sustained period. Hence gold is not something that should be of interest only to money managers and people with a trading account who increase their gold allocation when they expect prices to go up. Gold is useful to everybody trying to save money and protect wealth over time due to its remarkably stable purchasing power. Anything proposed as money, including bitcoins, should be able to do that, but gold is the only form of money that has a proven track record in both ancient and modern times. Bitcoin’s 8-year history is but the bat of a historical eyelash, a monetary experiment rather than a monetary proof of value.