Are actuals pit?

Played around with actuals in combination with earnings estimates Rate of change.
Looks a bit hefty to me (all liquidt stocks, mostly mid to big caps).

Please coment. TY

P123, please respond: results look a bit good to be true…

Those are highly liquid stocks!

What backtest parameters are you using? I’m not seeing anything unusual here.

Interesting. Def worth keeping an eye on :slight_smile:

@yuvaltaylor Check out 2020 - Present

This one is on highy liquid stocks + variable slippage. Just not used to strats that are higher cap and print that kind of backtest.

So based on your answer I understand that actuals are PIT!

Great data add!

Interesting sim. I had trouble getting similar results until I realized that Force Positions into Universe=False was needed. Setting that to true trashes the equity curve. I suspect the Universe changes a lot. I need to look at that.

The outperformance AlgoMan puts a finger on happened between October 2020 and February 2023. Before and after those dates, the performance was congruent with that of the S&P 500. If you substitute for the rule frank("ActualGr%PQ(#EPS)",#all,#DESC,#ExclNA)>80 the rule frank("(EPSActual(0,Qtr)-EPSActual(1,QTR))/EPSActual(1,Qtr)",#all,#DESC,#ExclNA)>80 you get a very similar result. So I don't think, as JudgeTrade implied, that there's any lookahead bias here.

Impressive start -

1 Like