R-squared of profitable daily algo strategy

Out of curiosity I tried to calculate the Sharpe ratio on 1 week of daytrading.
I noticed that my Sharpe ratio improved if I deleted some of my very profitable trades.

Somebody advice me to use the Sortino ratio instead as daytrader.
But now I have another problem:
  • I took all trades from 1 week
  • I calculated the Sortino ratio
  • Result was that Sortino ratio could not be calculated as there were no losing trades...
As there were no losing trades, so no negative excess returns, the downside risk was zero. And later dividing by zero did obviously not work.

Conclusion: to calculate the Sortino ratio I should at least make one losing trade.
 
damien.dualsigma said:
Good job... what is your average holding period and return per trade?
I would say keep doing what you are doing
More...

Holding times go from 5 minutes to 4 hours and 30 minutes (for that specific week). Sometimes holding time is from opening till close.

To me long holding times are not equal to big risk. Long holding is because there is a strong trend and the profit is huge too. Long holdings always have a stop to protect part of the profit just in case.

Returns I don't post anymore. Too much negative reactions. People on ET expect you to lose money, not make money.
Returns are also not a good reference to check the quality of a system. Drawdown, leverage, products traded... all these things are important to have a realistic and correct idea about the quality of a trading system.
 
Ok I understand that for some mysterious reason people hate it when you tell them it was you who took their money.... so forgetting about the overall strategy, what is your target profit and stop loss for each one of these short term trades? Do you use trailing stops?
 
damien.dualsigma said:
what is your target profit and stop loss for each one of these short term trades? Do you use trailing stops?
More...

I always put a fixed stop, so I know before opening a trade already what my (theoretical) max loss will be.
I trade mathematical models, no charting, no price action or anything else.
I have no profit target. The mathematical models tell me what to do and give me an indication of the probability that the signal is profitable.
I only put a trailing stop when probabilities are not high enough or when the open profit is big.
 
You must log in or register to reply here.
Back
Top