and how does over 45% roi and same volatility get a sharpe of 1.43??

I have the same question. My algorithm returns 25% in first year with a volatility of 3%. Why is the Sharpe ratio only 1.68?

I thought sharpe ratio = (return - risk free rate) / volatility. If you assume zero risk free rates then it should be just return / volatility.

It has been frustrating for me since they did not calculate sortino correctly, and then it was beta, and now it's backtest sharpe. I don't know how it is possible to mess up so bad on such simple calculations.

Unless.. they are now charging us for shorting stocks. Even then, if 25% return with 3% vol gives a sharpe of 1.68 that means that they are charging us a whooping 20% for charging stocks. mmm.. doesn't make sense. Maybe they just got the calculations wrong.

Yeah, I think the calculations are wrong. I'm getting some werid results as well.

I am wondering if this little function is the culprit:

def _adjust_returns(returns, adjustment_factor):  
    """  
    Returns a new pd.Series adjusted by adjustment_factor. Optimizes for the  
    case of adjustment_factor being 0  
    Parameters  
    ----------  
    returns : pd.Series  
    adjustment_factor : series / float  
    Returns  
    -------  
    pd.Series  
    """  
    **if isinstance(adjustment_factor, (float, int)) and adjustment_factor == 0:  
        return returns.copy()**  
    return returns - adjustment_factor  

There was Backtester update.

My test shows wrong calculation of Sharpe and Sortino Ratios in Bactester.

Hi all,

A lot of questions in this thread, sorry if I don't get to all of them.

@toan
Averaging sharpe ratios across different time scales will always give different results. Here's a paper on the relative error when the sharpe ratio is averaged across different time scales as opposed to calculated once as (mean (daily risk-adjusted-return)) / std(daily risk-adjusted-volatility) * sqrt(252)

@Pravin
See the calculation above, this error is due to inconsistent averaging across time scales. Using the daily returns will always give back different values than a single CAGR-based calculation. Neither methodology is "wrong", however.

If there's anything else I can clarify, please feel free to ask.

Best,
Lotanna Ezenwa

Lotanna,

you should not be getting 4.4 sharpe with 27% return and 1.43 sharpe with over 45% return when the volatility is the same no matter how you calculate it.

Lotana,

Can you comment this metrics:

My test shows wrong calculation of Sharpe and Sortino Ratios in Bactester.

Pyfolio metrics:

Backtest Months: 26
Backtest
annual_return 0.14
annual_volatility 0.09
sharpe_ratio 1.54
calmar_ratio 1.95
stability 0.93
max_drawdown -0.07
omega_ratio 1.30
sortino_ratio 2.31
skewness -0.07
kurtosis 2.39
information_ratio 0.03
alpha 0.14
beta 0.05

Becktester Metrics:

Total Returns
33.8%
Benchmark Returns
16%
Alpha
0.13
Beta
0.06
Sharpe
0.37
Sortino
0.55
Information Ratio
0.02
Volatility
0.09
Max Drawdown
-7.4%

Hi Lotanna,

When you say:

See the calculation above, this error is due to inconsistent averaging across time scales.

Does it mean there is an error currently in the calculation and is going to be fixed?