How Hedge Funds Assess Sharpe Ratio Claims
Summary
The document explains how a prospective hedge fund employer or investor may evaluate a person’s claimed Sharpe ratio. The usual evidence is a return time series, preferably from live trading and at daily frequency. Some reviewers may accept a combination of live and backtested performance when the two are reasonably consistent. People who previously worked at funds may also describe monthly strategy returns and their role, though responsibility for a strategy may still depend on trust.
It contrasts this informal review with standardized performance reporting. GIPS provides conventions for presenting investment returns, including time-weighted returns and disclosures about composites, benchmarks, account discretion, and gross or net results. The document says GIPS does not prescribe a Sharpe ratio calculation or prevent all ways of making the ratio look better, such as illiquid or serially correlated returns, changing the measurement horizon, or certain return swaps. Independent third parties can verify GIPS compliance, but investors still choose their own performance measures.
Key ideas
- Reviewers commonly request a strategy return history, preferably from live trading and at daily frequency.
- A blend of backtested and live returns may be accepted when the results are consistent.
- Attribution of past fund performance to an individual may rely partly on trust.
- GIPS standardizes return presentation and disclosures but does not provide specific Sharpe ratio guidance.
- Sharpe ratios can be distorted by illiquidity, serial correlation, measurement choices, or structured swaps.
Tags
Full text
# How is someone's Sharpe ratio recorded and communicated? # How is someone's Sharpe ratio recorded and communicated? When I read about, say, some hedge fund wanting people with such-and-such Sharpe ratio, how is that ratio recorded and communicated to the interested party? I mean, do people just take it on faith that if I say my Sharpe ratio is X, is it so? Or is there some formal process where this can be recorded? ## Answer by Tal Fishman (score 3, accepted) https://quant.stackexchange.com/a/3463 The hedge fund will usually ask people to provide a time-series of their strategy's returns, preferable from live trading and preferably at a daily frequency. Sometimes they will accept a mix of backtested and live results, so long as they are not too different from each other, both in theory and in practice. Some funds break down their monthly returns by strategy, and people who have worked at such funds in the past can provide this information and indicate which strategy was their responsibility, but then it will be taken "on faith" that this person was, in fact, the primary person responsible for a given strategy. Bottom line, I do not think there is any formal process for recording this information, and the standards of the potential employer/investor will vary widely. ## Answer by Ram Ahluwalia (score 5) https://quant.stackexchange.com/a/3468 The GIPS standards are increasingly used for presenting investment returns in a standardized fashion across equities, real estate, private equity, fixed income and other asset classes. The GIPS standards rely on, among other things, chain-linking time-weighted returns and they require specific disclosures including carve-outs, net or gross performance, treatment of discretionary accounts, and definitions of composites and benchmarks. There is no specific guidance on Sharpe ratio which is a shame since the ratio can be gamed by trading illiquid (serially correlated assets), by extending the measurement horizon, or engaging in a return swap where you pay out both the best and worst performing months to a dealer. However, by presenting the returns in a GIPS compliant fashion an investor can evaluate against performance metrics of their choosing. A firm can also verify GIPS compliance by an independent third party.
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.