12 Comments

Nice, I have a question ,what is the basket of stocks are you using for backtest, it is the whole US stock market (excluding OTCs ) ? Or it's just S&P ? And if it's S&P I suppose it doesn't have survivorship bias

Expand full comment

It's the whole US stock market (excluding OTCs ), delisted + listed stocks from 2007.

Expand full comment

Thanks Niv and one more question, how is the size allocated per trade, it is equally devided on how many setups are on each specific day ?

Expand full comment

The risk is 0.5% of the account per trade.

There was no limit to how many signals it can take per day - only limit is portfolio size up to 1.5x leverage.

Expand full comment

Again, great effort Niv. What I find most difficult in trading EPs is position sizing with the high volatility in the first hour. Did you test different entry points and stop loss percentages?

Expand full comment

I tested different stop losses yes - many things works. I tested different entries on an intraday level a long time ago but I'm currently not going into that rabbit hole as buying the open seems to be enough. Will be smarter after some live testing.

Expand full comment

can you share the sharpe ratio of the strategies, and of the combined? they must be quite uncorrelated so i assume the Sharpe of the combined is much improved

Expand full comment

Great write up! Can you confirm if ROC is rate of change? So rate of change of 30% usually less EV?

Expand full comment

Rate of Change % for the past 30 days. I didn't give the exact number where EV decays to keep some edge to myself ;) But if you recreate it I'm sure you can find the optimized number

Expand full comment

Maybe whisper to me? Lol

Expand full comment

Hi Niv, great post as usual ! Correct me if I am wrong but isn't the max DD on the EP backtest 30% ? Atleast thats what I see from the image but I might be wrong. Thanks for the post as always, great, simple to digest info !

Expand full comment

Absolutely, I copied it from momentum breakout and forgot to change the max DD %. Thanks for letting me know it's fixed now

Expand full comment