Big 'O' Sharpe changes backtest starting dates day by day until the lowest Sharpe is found.
When strategies rebalance on periodic basis it turns out that such very small changes cause very large differences in results.
Big 'O' Sharpe is the pessimistic grumpy brother of the happy-go-lucky Sharpe ratio. On the plus side if you do get a good Big 'O' Sharpe number, you're onto a good thing.
Before I take this weapon to more popular strategies I am culling my own. Skewerage and Momentum++ have fallen already, now for Steady Vol (a timing strategy which inversely weights holdings by current vol).
By bumping decision days by 1 day offsets I have found the best and worst results of Steady Vol applied to the S&P 500 over the last 50 years.
The 'Best' and 'Worst' Sharpe Trajectories are identical but make decisions just two days apart each month.
Over the same period of time the S&P 500's Sharpe ratio is 0.4.
'Best' has a Sharpe of 0.55, a 40% improvement. But 'Worst' has a Sharpe of 0.35. A 10% drop.
If I invest in this strategy I'd rather be quoted the worst Sharpe number, i.e. the Big 'O' Sharpe, and be pleasantly surprised if things work out well, wouldn't you? Which is why Big 'O' Sharpe is really the only relevant Sharpe ratio anyone should use.
Any recommendations for which strategies I should point this Big 'O' Sharpe weapon next? Let me know.