Musselmant posted:"Did your experiments with P123'S longer term backtester periods generate a few good criteria combos you can share?"

1) Screens that started their backtest in 1999 (peak of the internet explosion)performed much, much better than the same screens that started in 2001 (internet collapse). The P123 Zweig screen had a total gain of 3,919% starting in '99. It had a 600% total gain when starting in '01.

2)An excellent performing public P123 screen over the long term has been Quality_Value with an 83,000% return (32% CAGR), -77% MaxDD. 1.36 Sortino ratio. It does rely on P123's custom Ranking formula.

Country(“USA”)
Universe($ADR)=FALSE // Only USA stocks
Frank(“OpMgn%TTM”,#Industry,#desc)>=50
Frank(“OpMgn%TTM”,#All,#desc)>=50
Frank(“OpMgn%5Y”,#Industry,#desc)>=50
Frank(“OpMgn%5Y”,#All,#desc)>=50
Frank(“AstTurnTTM”,#Industry,#desc)>=50
Frank(“AstTurnTTM”,#All,#desc)>=50
Frank(“ROI%TTM”,#Industry,#desc)>=50
Frank(“ROI%TTM”,#All,#desc)>=50
Frank(“ROI%5Yavg”,#Industry,#desc)>=50
Frank(“ROI%5Yavg”,#All,#desc)>=50
Frank(“IntCovTTM”,#Industry,#desc)>=50
Frank(“IntCovTTM”,#All,#desc)>=50
CurFYEPSLow>0 // Earnings will be positive this year
FOrderOLD(“Pr2SalesTTM”,#All,#Asc,#Previous,TRUE)<=8

3. My most successful screen has been what I call BenchmarkPlus!. It uses the standard Benchmarking the Dow parameters plus i) it stays out of commodities and highly regulated industries, and ii) it requires ROC > CAPM Discount Rate. Longterm CAGR 26%, maxDD -56%, Sortino 1.36.

Universe(NOOTC)=TRUE & Universe($adr)=false
Universe(MasterLP)=false and SalesGr%TTM!=NA
AvgDailyTot(20)>500000 // Enough liquidity to trade
Price > 1 // No penny stocks
// Stay out of commodities and highly regulated industries
!GICS(FINANC) and !GICS(BASICM) and !GICS(UTILIT)
!GICS(REOPER) and !GICS(BLDRAW) and !GICS(AGRLIV) and !GICS(ACROPS) and !GICS(CCOALL)
EBITDAPY>0 and EBITDAPY2>0 and EBITDAPY3>0 and EBITDAPY4>0 and EBITDAPY5>0 //no negative earnings
CurFYEPSMean>0 & CurFYEPSMean!=NA
NextFYEPSMean>CurFYEPSMean
$ROC>$CAPMDiscRate/100
CurFYEPSMean>1.10*EPSActual(0,ANN) //Growing EPS
ROE%5YAvg != NA & ROE%TTM != NA
Price<$MktToBookLow*(ROE%TTM/ROE%5Yavg)*BVPSA //priced below historical norms
Pr2SalesTTM != NA
FOrderOLD(“Pr2SalesTTM”,#All,#Asc,#Previous,TRUE)<=4

Taz