How diversified is the Autopilot?
We talked about the Herfindahl Index (a concentration index) back in December, which showed that it is 5x less concentrated than just holding Bitcoin. It’s an “ok” index, but a somewhat simplistic one that doesn’t take into account how correlated those assets are to each other in terms of prices.
So this month we wanted to talk about another way of looking at it, the “Diversification Ratio” (D-ratio). This expresses the weighted average volatility of the portfolio instruments as a multiple of portfolio volatility. The higher the D-ratio, the more risk is reduced by combining assets that behave differently, rather than concentrating in assets who all more or less behave the same way, price-wise. And by picking up negative correlations between asset prices to reduce portfolio risk.
The result is the following: Bitcoin has a D-ratio of 1, as the portfolio risk will be equal to the risk of the single instrument BTC (duh! wink ). An Index strategy like Crypto20 has an average D-ratio of 1.2 over 2020, meaning limited diversification was offered since there is a concentration in the largest assets that tend to behave very similarly.
A broad universe like S&P500 has an average D ratio of 1.7 over 2020 by using market cap weights. The Crypto Autopilot had an average of 2.2 D-ratio, a substantial decrease in risk by diversifying more than an Indexing Strategy. In the short run (single-year 2020), especially when one asset is really on a tear, the benefits of spreading out might not be immediately apparent but in the longer run (multi-year) there is a proven link between higher D-ratio and outperformance since the portfolio picks up codependence as well. That’s why we keep saying that a good holding horizon for the Autopilot is 2.5 to 5 years.
