The Skew-a-Day Strategy

The Sentimental Skew Strategy is my test bench for the Backtesting IDE.

I have documented the building of the strategy (while building the IDE).

The strategy uses the CBOE implied skew index as an indicator to long or short the S&P 500.

It was also instructive to understand the temptations strategy builders often succumb to.

A subtle tweak here and there can make all the difference.

While I was building and testing the IDE I noticed outsized returns if I used a horizon (or holding period) of 20 rather than 21 business days.

Over 25 years the Sharpe becomes 0.65.

The Skew Strategy is moving far beyond the benchmark of the S&P with a Sharpe of 0.5.

Now there is about a 1 in 600 chance of trading noise versus a 1 in 100 chance with the previous strategy.

We rarely invest in anything for 25 years however, better to generate future parallel universes and understand shorter to medium time frames.

The 20 day horizon beats the 21 day original strategy in over two thirds of each of the last 25 years.

Not a once off windfall.

Small tweaks causing large changes in performance make me wary. More investigation is needed.

One thing to take note of, the Skew index interpolates the 30 calendar day point on the 'implied skew curve'.

Does a 20 day horizon make a better match to those 30 days?

Or maybe a slightly more responsive holding period makes all the difference?

Perhaps a bit of both.

Or perhaps I have subliminally tried enough strategy combinations to randomly hit upon fool's gold!

More next time.