|
PureBytes Links
Trading Reference Links
|
I'll plug in the Sharpe Ratio. Thanks.
I failed to mention a few things:
The strategies are applied to intraday data.
This equity curve is based on trading two stocks, each using a different
strategy. I matched stocks and strategies based on a high RR, low standard
error, and relative negative covariance. I multiplied all my RR's by 1000
because I wanted to see whole numbers. Also, I didn't account for any
slippage, which is a weakness. I hope that 900+ trades in four years of
changing markets is statistically significant, but perhaps it's not.
Here's another catch: These strategies are not super robust: if I apply the
same strategies to different stock combinations, I can get unsatisfactory
results. You could say that each market is a different animal, and you come
up with a strategy for that animal, but these beasts are known to change
into something else.
At 05:40 PM 1/3/2002 -0500, Bob Fulks wrote:
>At 10:33 AM -0800 1/3/02, Craig wrote:
>
> >Test based on trading two markets simultaneously. Trader is always in
> the market, and 100% invested. Total capital = $100,000, buying and
> selling $50,000 for each market. Non-cumulative (No profits are
> re-invested.) Commissions are subtracted from equity curve.
> >
> >I would appreciate any comments about this equity curve. What is wrong
> with it, or what is lacking in terms of a measurement or ranking value.
>
>Try calculating the Sharpe Ratio in Excel:
>
> monthly_profit_% = monthly_profit / $100,000); [Looks like about 20%]
>
> average_monthly_profit_% = AVERAGE(monthly_profit_% (over all months);
>
> a_excess_return = 12 * average_monthly_profit_% - 5%;
>
> a_standard_deviation = SQRT(12) * STDEV (monthly_profit_% (over all
> months);
>
> Sharpe = a_excess_return / a_standard_deviation;
>
>A Sharpe Ratio over 1 is "OK". A Sharpe Ratio over 3 is great.
>
>Bob Fulks
|