Steady, Low Vol & Skew

Steady Vol inversely weights holdings in an asset with respect to current volatility.

Initially there was a little confusion about how Steady Vol was different from Low Vol strategies. Low Vol strategies weight individual constituent stocks by their volatility whereas steady vol weights periods (days, months etc.) it's all about timing.

Recently, I realised, that all we are ever trying to do is reduce portfolio volatility, higher returns are really just a side show compared to this deeper goal.

So, just for kicks I applied Steady Vol to a Low Vol index, so we end up inversely weighting both by constituents and on an intertemporal basis.

Interestingly the Sharpes are almost identical at about 1.

What's different is the Skew. As I alluded to before, the skew for Low Vol is a terrible -0.4, which means your upsides are far more limited than your downsides, but Steady Vol has a skew of +2.

Let's try to identify what that looks and might feel like.

The two Sharpe Trajectories are pretty similar, as they both have almost the same Sharpe ratio.

However Steady Vol stems massive losses early on in February 2009 and then during Summer 2011, quickly reacting to volatile conditions and deleveraging, but it also manages a massive profit over a short period of time at the beginning of 2012.

That's what positive skew feels like - minimising losses while being open to large windfalls when available.

Sharpe Trajectories make it so much easier to see that the efficiency gains Steady Vol makes over Low Vol in the 2009-2012 period are lost in the second half of the backtesting period.

Steady Vol leverages up during the historically low volatility period and incurs magnified profits and losses. Inefficient.

Steady Vol logic for the Lazy Backtest IDE is here.