Sharpe has limit, might not be best, but due to long-short nature of HF, it is a way to measure. So, is it good or not? I think mainly good, but with limitation, but hell, in tests, Sharpe seems so bad for HF. Any one?
Sharpe uses the total risk (Standard dev.) that assumes normality. HF returns are, in general, non-normal.
Downside deviation is a better measure for HF: (((min(0,return - threshold return)/N-1)))^.5
Sortino is better for HFs, but try to find one that doesnt give you their Sharpe…as far as this test is concerned…Do not use Sharpe to evaluate Non-Normal strategy retursn, ie. HFs
for some reason I have a hard time remembering this but I think that Sharpe is not appropriate because it uses standard deviation, Downside deviation is more appropriate because it doesn’t penelize the manager for the volatility of positive returns… am i right?
I know this was debated as well, can someone confirm or deny? Sortino = (return - Min)/stand dev of portfolio) RAY’s = (return - Rf)/downside dev)
I’m so glad I read this.
^NO Sortino = (return - MAR)/Downside Deviation or Targetdeviation Ray’s SF = (Return - MAR)/Std Dev
3rd & Long Wrote: ------------------------------------------------------- > I know this was debated as well, can someone > confirm or deny? > > Sortino = (return - Min)/stand dev of portfolio) > > RAY’s = (return - Rf)/downside dev) Sortino uses downside deviation. What is Ray?
Sharpe can be “gamed” little, sharpe will look better for longer period of time, since std. is lower given longer period. Plus, sharpe doesn’t consider liquidity factor, it assumes all investment are liquid. More to remember…
3rd & Long Wrote: ------------------------------------------------------- > I know this was debated as well, can someone > confirm or deny? > > Sortino = (return - Min)/stand dev of portfolio) > > RAY’s = (return - Rf)/downside dev) I thought (return - min)/ std dev is Roy’s Safety first rule
so both used MAR though right?
sorry, RAY, ROY…tomato/tomatoe…
also there was a whole list of items that are wrong with Sharpe for Hedge funds: 1. it can only be used on separate basis 2. it’s dependent on time, if you use a longer time period sharpe goes up 3. it’s usefulness has not been proven in empirical historical data 4. … can’t remember the rest!!! there were at least 8 items!!!
bigwilly Wrote: ------------------------------------------------------- > ^NO > > Sortino = (return - MAR)/Downside Deviation or > Targetdeviation > > Ray’s SF = (Return - MAR)/Std Dev Since Ray’s SF focuses on shortfall risk, I would use RL instead of MAR since MAR assumes a positive return while shortfall can be a negative return.
Iwona pass Wrote: ------------------------------------------------------- > also there was a whole list of items that are > wrong with Sharpe for Hedge funds: > > 1. it can only be used on separate basis > > 2. it’s dependent on time, if you use a longer > time period sharpe goes up > > 3. it’s usefulness has not been proven in > empirical historical data > > 4. … > > can’t remember the rest!!! there were at least 8 > items!!! ^ i dont even remember these, btw, these ratios are absolutely a**hat ratios, who actually looks at a hedge fund presentation and thinks “hmm, sortino ratio looks good”
abacus…for Roy MAR = Minimum Accepted REturn. Usuallyl you dont use a Negative return in Roy’s.
bigwilly Wrote: ------------------------------------------------------- > abacus…for Roy MAR = Minimum Accepted REturn. > Usuallyl you dont use a Negative return in Roy’s. I think there was an example in schweser where a -ve number was used. Will check on that But how to we factor client’s statement ‘I don’t want to lose more than 5% of my portfolio’ vs. ‘I would like a min 5% return’
Maximum acceptable portfolio loss of 5% over the next year?
The 2 critical measures for hedge fund performance evaluation are drawdown and downside measure like Sortino ratio