In: Finance
Sharpe’s and Treynor’s. Which measure are you most likely to use ? What benchmark would you choose to compare the performance?
The use of the Sharpe’s ratio has been more mainstream. Sortino’s is rarely mentioned in undergrad textbooks; however, it is used by practitioners. Sortino’s Ratio is only penalizing for harmful volatility. In your opinion, do you think it would be appropriate to utilize this ratio? Would you use it to evaluate ?
Both Sharpe ratio and Treynor ratio are important for financial modelling. Sharpe ratio ( (return-risk freerate)/std dev) measures the additional return a fund will generate with respect to the diversifyable risk we takes whereas Treynors ratio (return-risk free rate)/portfolio beta) deals with the extra return a fund will generate per additional systematic risk taken in terms of beta.
However if we have to choose an important ratio among these 2, Sharpe ratio will be preferable as it deal with diversifiable risk in terms of standard deviation. This can be easily mitigated and modified using different assets in a portfolio and by varying asset weights. ie: This is the basic optimization that should be done for every portfolio and hence as a first priority we should consider optimizing sharpe ratio.
Sortino Ratio ( (return-risk free rate)/DR) where DR is the downside risk which only considers the negative variation volatility. Volatility or std deviation in Sharpe ratio consider risk on both sides. ie If an asset return increase significantly, the std deviation will show higher value, but as investors we are only concerned with downside variation. We will be having problem only if a stocks return go downside. As a solution to this Sortino ratio was introduced which considers only downside risk. and optimizing using these values can give more better results.
However, calculating only downside risk is a very tedious and time consuming process. Even though this is an appropriate way of measuring a portfolio, the usage of Sortino ratio is not common due to its complexity in calculation.