Including Strategy Shutdowns in Sharpe Ratio Evaluation

Summary

The note explains that whether to include a shutdown period in a strategy’s Sharpe ratio depends on why the strategy stopped trading and what the statistic is intended to measure. If a discretionary trader deliberately moves to cash, that choice is part of the strategy and the idle period can reasonably remain in its performance record.

If an algorithm is unable to access the market because of an operational interruption, excluding that period may better isolate the strategy’s performance while it could trade. However, investors may request a reasonable estimate of how it would have performed during the interruption, particularly to check that poor periods are not being omitted selectively. The discussion offers a decision framework rather than a universal rule, and does not specify how to estimate missing returns or adjust annualization and risk-free rates.

Key ideas

• Include time in cash when the decision to stop trading is part of the strategy.
• An external market-access interruption may be excluded when evaluating tradable-period performance.
• Investors may ask for simulated returns during shutdowns to guard against selective omission of bad periods.
• The purpose of the Sharpe ratio determines how shutdown time should be treated.

Tags

# Should the strategy shut down period be considered when calculating Sharpe?


# Should the strategy shut down period be considered when calculating Sharpe?












When computing the Sharpe for a strategy, should the shut down period be considered?

## Answer by LazyCat (score 4, accepted)

https://quant.stackexchange.com/a/57892

It usually depends on:

- the reason why the strategy was shut down

- what are you using the sharpe ratio number for

Examples:

- you're a discretionary trader and at a certain point decide to go all in cash for the next month. It's reasonable to include shut down period into calculation, since the decision was a part of your strategy

- you run an algo strategy and suddenly loose access to the market for the next month. For the purpose of evaluating the strategy it makes sense to exclude that period altogether. At the same time if you have a way to reasonably simulate the performance during that month the investors may ask to include that in the overall stats to make sure you're not excluding periods with bad performance.

Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.