Diversification Benefits for Quantitative Fund Management
By Ron DiRusso, Contributor for Lucy Labs
Most investors would agree that diversification can be healthy for a portfolio. For one thing, provided the asset mix is uncorrelated, diversification reduces overall risk, without necessarily reducing returns. Of course, that depends on the mix of assets chosen, and the number of assets in the portfolio. Too many assets can sometimes reduce the overall diversification benefits.
Let us look at a typical use of diversification in asset management. Risk Parity is a portfolio strategy that uses risk to weight a portfolio. Typical portfolios have always used dollar weighting to balance portfolios. The 60–40 equity to bond ratio is the most familiar example of dollar weighting. Risk Parity, on the other hand, uses risk weighting, so it allocates equal risk weights to various uncorrelated asset classes, such as equities, bonds, commodities, and alternatives like real estate. Risk weighting allows correlation, or lack thereof, to work to reduce the risk and potentially enhance the return of the overall portfolio. The benefits of using a diversified approach are many, including lower volatility, more consistent returns, and less dependency on one given asset.
In the world of quantitative hedge funds, we look for ways to create income for our clients that is driven from a theme, such as momentum, or trend following. Trend following is a popular strategy that profits from the natural tendency of autocorrelation, meaning if a market is moving in one direction, it will continue to move in that direction. Trend following strategies can be diversified in several ways.
For one thing, the asset mix in a trend following strategy can help with diversification. If all of the assets in the model are highly correlated, the strategy will provide less diversification than if an uncorrelated mix of assets is chosen. Additionally, the frequency of the models that are used can affect returns. Typical trend following models can use either moving averages, or breakouts to determine a trend. Sometimes mixing these techniques can help with diversification. Also, blending frequencies helps with diversification.
In the chart below the orange line is the signal generated from a 50/200 day MA crossover while the blue line is the signal from a 5/20 day MA crossover. You can see how they often cancel each other out. That canceling effect reduces the overall volatility of the strategy, but sometimes with lower returns if the averages are not all similarly profitable.
Another way to diversify away from pure trend following is to overlay other types of strategies. Some examples might include a mean reverting strategy like market making. Likewise, more fundamental, statistical arbitrage strategies can help diversify away from pure trend-following.
Option strategies can also help with diversification. Trend following, along with fundamental and market-making strategies, are all purely directional. They trade spot or forward/futures markets as their underlying asset. They can also be described as Delta One strategies. Options, on the other hand, can be used as a separate asset class if approached as distinct from the underlying assets. An option strategy can have a very different return profile than the underlying asset.
Using relative value (RV) options strategies can create a diverse overlay to a directional portfolio. The inputs to a directional strategy are usually related to price, whereas options strategies tend to focus on a derivative of price, volatility. Volatility can go higher when prices go lower and vice versa. It creates an orthogonal universe to pure price action, although in the end we all depend on prices.
Options traders use supply and demand to find opportunities in the volatility space. Even for the same asset, different maturities can have different volatility levels. The chart below is a typical example of the term-structure of volatility for Bitcoin options. It shows that shorter term volatility is lower than longer term. In this term-structure, it might be advantageous to purchase short-term options with lower volatility levels and sell longer dated options against them.
Likewise, different strikes can have distinct volatility levels. The chart below depicts the volatility level for different strikes for options that expire for the same maturity, approximately two months. One can see that at the money options (at the lowest point in the chart) have a lower volatility level than options that are out of the money. The overall premium is lower for out of the money options, but on a volatility basis, they can be much more expensive. Options traders can use this information when there are kinks in the curve, meaning that one particular strike is much more or less expensive on a relative basis than others. Traders can sell that kinked strike, and purchase another strike against it. They can then manage the position at a portfolio level.
Options have a property called time value, which is the amount of volatility or value of time left in the option. Time value tends to decay in a non-linear fashion. This feature of options can be used to create a yield play that can benefit a portfolio when markets quiet down or remain in a range. When an option has little time left, it decays more rapidly than when it has a longer term to expiration. This means that in general, selling shorter dated options can be a way to generate more yield.
In short, there are many ways to trade options that could provide a non-correlated return to more directional strategies. These techniques are used widely in the equities and other markets, but can be extended to the crypto world to create a more balanced and robust return for our clients.