Skip to main content
No. of Recommendations: 16
I track post-discovery CAGR for some screens. Below are 69, a few with timing, but no blends. Most hold for 21 market days. No standard screens. This is just a list I keep to track screens I might want to investigate further.

The backtest start date is when the screen was posted, capped at 20 years ago. Post-discovery results start on different dates, so a market index is needed for comparison. I use a S&P 1500 Equal Weight index (blend of tickers RSP, EWMC, EWSC). All results below end 20181231, and use 0.4% friction. I use the depth (number of holdings) described in the discovery post.

delta is CAGR minus index. years is backtest length. sort is (delta * years).

Short screen results are the screen picks held long. So lower CAGR is better for short screens. Short screens have no friction, and delta is index minus CAGR. Higher delta is better. Short screen names end with '_s'.

Screen                                             CAGR  GSD  Sharpe  StartDate  yrs  SP1500EW  delta   post   sort  depth    settings    type
Short_HardPS_s_20110626_GeekGod -17 50 -0.26 20110627 8 10 27 231169 201 4 h21i1 short
NOIS_s_20160329_DrBob2 -21 34 -0.74 20160329 3 8 29 261725 81 5 h252 short
FireAway_20140125_musselmant 17 21 0.90 20140127 5 7 10 248107 51 5 h21i1f0.4 screen
H52EarnPSmo_20140519_numerous 16 27 0.73 20140519 5 6 10 250863 45 5 h21i1f0.4 screen
ROICi_20160506_kimardenmiller 19 23 0.94 20160506 3 8 11 262264 30 10 h21i1f0.4 screen
kindaLikeWER_20180930_borisnand 11 23 0.55 19981231 20 10 1 271966 29 200 h21i1f0.4 index
MagicFormi_20140119_MarginOfBuffett 11 17 0.75 20140121 5 6 5 247981 26 8 h253i1f0.4 screen
ROICi_BCC_DSP_20160506_kimardenmiller 17 22 0.85 20160506 3 8 9 262264 24 10 h21i1f0.4 timing
100ClubNoFin_20021214_DrBob2 12 22 0.64 20021213 16 11 1 137183 20 5 h63f0.4 screen
WER 11 25 0.49 19981231 20 10 1 4044 14 125 h21i1f0.4 index
Tortoriello_20150405_mark19601962 8 18 0.51 20150406 4 4 3 255743 12 10 h253f0.4 screen
Volatility_Fscorei_20150415_DrBob2 7 16 0.53 20150415 4 4 3 255931 12 5 h21i1f0.4s screen
Short_HardPS5_s_20180826_GeekGod -76 45 -3.78 20180827 0 -46 30 271766 10 5 h21i1 short
WER_PE_BCC_20130929_rgearyiii 9 13 0.73 20130930 5 8 1 245553 6 10 h1 timing
Zacks_PE_Beta_20121208_rgearyiii 12 14 0.94 20121210 6 11 1 240468 6 10 h21i1f0.4 screen
mi_3PT_SCV_pihA_20170728_borisnand 3 17 0.24 20170728 1 -1 4 267224 6 10 h21i1f0.4 screen
kindaLikeMTUM_20160423_borisnand 9 21 0.55 20160425 3 7 2 262119 6 10 h126i1f0.4 screen
StockMarketWinners_20170808_Whoop8 3 12 0.29 20170808 1 0 4 267298 5 25 h253i1f0.4 screen
mo_PSrel5_20160127_musselmant 14 21 0.77 20160127 3 13 2 261161 4 100 h5i1f0.4 screen
TinyTitans_20150401_numerous 6 17 0.41 20150401 4 5 1 255681 4 33 h21i1f0.4 screen
ROE_20180730_rrr12345 -26 20 -1.58 20180730 0 -34 7 271531 3 48 h84n3i1f0.4 screen
H52vrCOP_20150925_borisnand 10 25 0.49 20150925 3 9 1 258489 3 5 h21i1f0.4 screen
SP500_DolVol_20170304_borisnand 2 13 0.16 20170306 2 1 1 265706 2 100 h253i1f0.1 screen
SP1500_BiotechStaples50mc2i_20161108_borisnand 8 16 0.57 20161108 2 7 1 264367 2 50 h63i1f0.4 screen
SP1500EqualWeight 10 24 0.47 19981231 20 10 0 263443 0 1500 h63i1 index
Mo_PCF_20140303_musselmant 6 22 0.37 20140303 5 6 0 249030 -1 5 h21i1f0.4 screen
RoeEyH52_20170725_mungofitch -3 17 -0.17 20170725 1 -2 -1 267195 -2 20 h63i1f0.4 screen
Pim_Van_Vliet_(Shareholder_Yield)_20180907_Whoop8 -56 24 -3.73 20180907 0 -48 -8 271817 -3 20 h63i1f0.4 screen
BigJumpTunedi_20140522_TGMark 6 36 0.31 20140522 5 6 -1 250937 -3 5 h21i1f0.4 screen
TrendingValueBeta_20151016_musselmant 6 22 0.38 20151016 3 7 -1 258890 -3 5 h21i1f0.4 screen
OptimistsRule_20180106_borisnand -19 34 -0.64 20180108 1 -12 -7 269012 -7 5 h21i1f0.4 screen
rankPefDvMo_20180720_borisnand -50 25 -3.02 20180720 0 -34 -16 271490 -7 10 h21i1f0.4 screen
kindaLikeSP1500EW_20180930_borisnand 9 22 0.47 19981231 20 10 0 271966 -8 1500 h63i1f0.4 index
BetaFscore_20150416_musselmant 2 20 0.18 20150416 4 4 -2 255931 -8 5 h21i1f0.4 screen
Up5X3i_20160112_TGMark 8 24 0.41 20160112 3 11 -4 260746 -12 10 h21i1f0.4 screen
SPY_BCC_20130929_rgearyiii 5 8 0.67 20130930 5 8 -2 245553 -12 1 h1i1 timing
SoWR_20130211_BarryDTO 7 17 0.53 20130211 6 10 -2 241539 -13 10 h21i1f0.4 screen
PE_RRS_vol_20160222_quantforlife 7 18 0.45 20160222 3 11 -5 261345 -14 89 h21i1f0.4 screen
SPY_NHNL_20130929_rgearyiii 5 9 0.58 20130930 5 8 -3 245553 -14 1 h1i1 timing
SurpriseEstimatesi_20180127_rdutt -24 24 -1.23 20180205 1 -8 -17 269478 -15 10 h10f0.4 screen
H52vrPOI_20140905_numerous 2 18 0.16 20140905 4 5 -4 252826 -17 10 h21i1f0.4 screen
RSPE_20160225_ilmostro 5 19 0.33 20160225 3 11 -6 261376 -18 20 h21i1f0.4 screen
WER_PE_20150429_DoesMIWork -1 19 0.03 20150429 4 5 -5 256197 -19 20 h21i1f0.4 screen
H52EarnMC_SI_20151011_borisnand 1 17 0.12 20151012 3 7 -6 244801 -20 25 h21i1f0.4 screen
SP500EW_BCC_20130929_rgearyiii 4 7 0.52 20130930 5 8 -4 245553 -20 500 h1 timing
SP500EqualWeight 9 22 0.44 19981231 20 10 -1 152894 -23 499 h63i1 index
SmallValue_20140822_numerous 0 21 0.07 20140822 4 6 -6 252573 -25 5 h21i1f0.4 screen
SP1500Fscore_20140826_numerous 0 19 0.05 20140826 4 6 -6 252657 -26 10 h21i1f0.4 screen
SP500EW_NHNL_20130929_rgearyiii 3 8 0.36 20130930 5 8 -5 245553 -26 500 h1 timing
Industry123i_20160118_musselmant 3 19 0.24 20160119 3 13 -10 260865 -29 5 h21i1f0.4 screen
RiftA_20140111_numerous 0 17 0.08 20140113 5 6 -6 247780 -30 10 h21i1f0.4 screen
EarnHighMo_20151228_borisnand -3 18 -0.09 20151228 3 8 -11 260325 -32 25 h21i1f0.4 screen
Marketsmith_20140115_rgearyiii -1 20 0.01 20140115 5 6 -7 247879 -35 5 h21i1f0.4 screen
RiftYbe_20140111_numerous -1 19 0.01 20140113 5 6 -7 247780 -36 10 h21i1f0.4 screen
WER_PE_NHNL_20130929_rgearyiii 1 17 0.10 20130930 5 8 -7 245553 -37 10 h1i1f0.4 timing
Zacks1 8 26 0.38 19991220 19 10 -2 842 -42 150 h21i1f0.4 index
WER_PE_20060320_MainiacJoe 5 29 0.27 20060320 13 8 -4 249091 -46 11 h21i1f0.4 screen
InmoHighEarni_20151004_borisnand -7 22 -0.28 20151005 3 8 -15 258656 -48 5 h21i1f0.4 screen
CashCowROE_SIP2_20151223_mungofitch -8 31 -0.18 20151223 3 8 -16 260242 -48 5 h21i1f0.4 screen
YieldRRS_20100903_mungofitch 6 29 0.36 20100903 8 12 -6 226652 -49 5 h21i1f0.4 screen
WER_PE_20130929_rgearyiii -2 21 -0.02 20130930 5 8 -10 245553 -50 10 h20i1f0.4 screen
Top50pctDolVol 7 25 0.35 19981231 20 10 -3 256567 -56 999 h21i1f0.4 index
EvBuying_20131209_DrBob2 -5 32 -0.04 20131209 5 6 -11 247042 -56 5 h21i1f0.4 screen
ValueTimes_20140111_numerous -6 20 -0.24 20140113 5 6 -12 247780 -59 10 h21i1f0.4 screen
BigJump_20131211_DrBob2 -7 37 -0.08 20131211 5 7 -14 247102 -69 5 h21i1f0.4 screen
TrendingValue_20120222_tetranomad -1 23 0.06 20120222 7 10 -11 235816 -76 15 h21i1f0.4 screen
WER_RSPE_20080625_MainiacJoe 3 25 0.22 20080625 11 10 -8 210273 -79 20 h21i1f0.4 screen
SP500MktCapWeight 6 21 0.31 19981231 20 10 -4 247888 -82 500 h63i1 index
gSPY 6 21 0.30 19981231 20 10 -4 247888 -82 500 i1 index
Print the post Back To Top
No. of Recommendations: 8
Results for 2018 (from 20171229 to 20181231) for the top 12 screens above, sorted by CAGR.

Screen                                 2018_CAGR
FireAway_20140125_musselmant 7
MagicFormi_20140119_MarginOfBuffett 4
Volatility_Fscorei_20150415_DrBob2 -2
kindaLikeWER_20180930_borisnand -7
WER -7
Tortoriello_20150405_mark19601962 -7
SP1500EqualWeight -10
H52EarnPSmo_20140519_numerous -10
ROICi_BCC_DSP_20160506_kimardenmiller -11
ROICi_20160506_kimardenmiller -15
100ClubNoFin_20021214_DrBob2 -21
NOIS_s_20160329_DrBob2 -46
Short_HardPS_s_20110626_GeekGod -56
Print the post Back To Top
No. of Recommendations: 1
Results for 2018 (from 20171229 to 20181231) for the top 12 screens above, sorted by CAGR.

Screen 2018_CAGR
FireAway_20140125_musselmant 7
MagicFormi_20140119_MarginOfBuffett 4
Volatility_Fscorei_20150415_DrBob2 -2
kindaLikeWER_20180930_borisnand -7
WER -7
Tortoriello_20150405_mark19601962 -7
SP1500EqualWeight -10
H52EarnPSmo_20140519_numerous -10
ROICi_BCC_DSP_20160506_kimardenmiller -11
ROICi_20160506_kimardenmiller -15
100ClubNoFin_20021214_DrBob2 -21
NOIS_s_20160329_DrBob2 -46
Short_HardPS_s_20110626_GeekGod -56


Amazing, my screen made the list. Maybe I should post more screens.
Print the post Back To Top
No. of Recommendations: 0
If you have gtr-speak links for the top ones it would be great; gtr has changed since the screens were created.
Print the post Back To Top
No. of Recommendations: 1
Thank you for the very interesting work, Don. For just the long screens >= 3 years post-discovery the average delta looks to be about -3.65:


Screen CAGR GSD Sharpe StartDate yrs SP15EW delta post sort depth settings type
FireAway_20140125_musselmant 17 21 0.9 20140127 5 7 10 248107 51 5 h21i1f0.4 screen
H52EarnPSmo_20140519_numerous 16 27 0.73 20140519 5 6 10 250863 45 5 h21i1f0.4 screen
ROICi_20160506_kimardenmiller 19 23 0.94 20160506 3 8 11 262264 30 10 h21i1f0.4 screen
MagicFormi_20140119_MarginOfBuffett 11 17 0.75 20140121 5 6 5 247981 26 8 h253i1f0.4 screen
100ClubNoFin_20021214_DrBob2 12 22 0.64 20021213 16 11 1 137183 20 5 h63f0.4 screen
Tortoriello_20150405_mark19601962 8 18 0.51 20150406 4 4 3 255743 12 10 h253f0.4 screen
Volatility_Fscorei_20150415_DrBob2 7 16 0.53 20150415 4 4 3 255931 12 5 h21i1f0.4s screen
kindaLikeMTUM_20160423_borisnand 9 21 0.55 20160425 3 7 2 262119 6 10 h126i1f0.4 screen
mo_PSrel5_20160127_musselmant 14 21 0.77 20160127 3 13 2 261161 4 100 h5i1f0.4 screen
Zacks_PE_Beta_20121208_rgearyiii 12 14 0.94 20121210 6 11 1 240468 6 10 h21i1f0.4 screen
TinyTitans_20150401_numerous 6 17 0.41 20150401 4 5 1 255681 4 33 h21i1f0.4 screen
H52vrCOP_20150925_borisnand 10 25 0.49 20150925 3 9 1 258489 3 5 h21i1f0.4 screen
Mo_PCF_20140303_musselmant 6 22 0.37 20140303 5 6 0 249030 -1 5 h21i1f0.4 screen
TrendingValueBeta_20151016_musselmant 6 22 0.38 20151016 3 7 -1 258890 -3 5 h21i1f0.4 screen
BigJumpTunedi_20140522_TGMark 6 36 0.31 20140522 5 6 -1 250937 -3 5 h21i1f0.4 screen
BetaFscore_20150416_musselmant 2 20 0.18 20150416 4 4 -2 255931 -8 5 h21i1f0.4 screen
Up5X3i_20160112_TGMark 8 24 0.41 20160112 3 11 -4 260746 -12 10 h21i1f0.4 screen
SoWR_20130211_BarryDTO 7 17 0.53 20130211 6 10 -2 241539 -13 10 h21i1f0.4 screen
PE_RRS_vol_20160222_quantforlife 7 18 0.45 20160222 3 11 -5 261345 -14 89 h21i1f0.4 screen
H52vrPOI_20140905_numerous 2 18 0.16 20140905 4 5 -4 252826 -17 10 h21i1f0.4 screen
RSPE_20160225_ilmostro 5 19 0.33 20160225 3 11 -6 261376 -18 20 h21i1f0.4 screen
H52EarnMC_SI_20151011_borisnand 1 17 0.12 20151012 3 7 -6 244801 -20 25 h21i1f0.4 screen
WER_PE_20150429_DoesMIWork -1 19 0.03 20150429 4 5 -5 256197 -19 20 h21i1f0.4 screen
SmallValue_20140822_numerous 0 21 0.07 20140822 4 6 -6 252573 -25 5 h21i1f0.4 screen
SP1500Fscore_20140826_numerous 0 19 0.05 20140826 4 6 -6 252657 -26 10 h21i1f0.4 screen
Industry123i_20160118_musselmant 3 19 0.24 20160119 3 13 -10 260865 -29 5 h21i1f0.4 screen
RiftA_20140111_numerous 0 17 0.08 20140113 5 6 -6 247780 -30 10 h21i1f0.4 screen
EarnHighMo_20151228_borisnand -3 18 -0.09 20151228 3 8 -11 260325 -32 25 h21i1f0.4 screen
Marketsmith_20140115_rgearyiii -1 20 0.01 20140115 5 6 -7 247879 -35 5 h21i1f0.4 screen
RiftYbe_20140111_numerous -1 19 0.01 20140113 5 6 -7 247780 -36 10 h21i1f0.4 screen
InmoHighEarni_20151004_borisnand -7 22 -0.28 20151005 3 8 -15 258656 -48 5 h21i1f0.4 screen
CashCowROE_SIP2_20151223_mungofitch -8 31 -0.18 20151223 3 8 -16 260242 -48 5 h21i1f0.4 screen
YieldRRS_20100903_mungofitch 6 29 0.36 20100903 8 12 -6 226652 -49 5 h21i1f0.4 screen
WER_PE_20130929_rgearyiii -2 21 -0.02 20130930 5 8 -10 245553 -50 10 h20i1f0.4 screen
WER_PE_20060320_MainiacJoe 5 29 0.27 20060320 13 8 -4 249091 -46 11 h21i1f0.4 screen
EvBuying_20131209_DrBob2 -5 32 -0.04 20131209 5 6 -11 247042 -56 5 h21i1f0.4 screen
ValueTimes_20140111_numerous -6 20 -0.24 20140113 5 6 -12 247780 -59 10 h21i1f0.4 screen
BigJump_20131211_DrBob2 -7 37 -0.08 20131211 5 7 -14 247102 -69 5 h21i1f0.4 screen
TrendingValue_20120222_tetranomad -1 23 0.06 20120222 7 10 -11 235816 -76 15 h21i1f0.4 screen
WER_RSPE_20080625_MainiacJoe 3 25 0.22 20080625 11 10 -8 210273 -79 20 h21i1f0.4 screen
Averages 4.15 21.90 0.30 4.95 7.70 -3.65


Are we willing to say that delta is somewhat of a proxy for how well we think MI has worked over the past 5 years? (average post-discovery period for these screens)

BTW: Is there still a way to get your S&P 1500 Equal Weight index with the current Yahoo! problem?
Print the post Back To Top
No. of Recommendations: 1
Are we willing to say that delta is somewhat of a proxy for how well we think MI has worked over the past 5 years? (average post-discovery period for these screens)

I am not willing to say that. No one is using all of these screens. I think they have to go through some screen selection process before evalutaion. Also, these screens are just the ones that I took interest in, and so are biased.

The standard MI screens might be a better proxy. They trailed the market on average in 2017 and 2018. These are mostly value and momentum type screens, and growth and not-momentum has done better the last few years. Maybe some MI screens are needed that don't use momentum or value multiples.

 Averages2018   CAGR  MDD  SDcg
avgLongScreen -16 -32 10
avgShortScreen -39 -53 12
avgIndex -7 -23 0.6

https://boards.fool.com/1-year-backtests-to-2018-december-34...

 Averages2017   CAGR  MDD  SDcg
avgLongScreen 16 -18 12
avgShortScreen 21 -25 20
avgIndex 21 -4 1.2

https://boards.fool.com/1-year-backtests-to-2017-december-32...

BTW: Is there still a way to get your S&P 1500 Equal Weight index with the current Yahoo! problem?

I try to avoid using Yahoo quotes. In GTR1, I use the following (earliest start 19941003):

http://gtr1.net/2013/?~SP1500EqualWeight:h63i1::sp1500.a:et1...

and for slightly longer backtests (earliest start 19870302):
http://gtr1.net/2013/?~kindaLikeSP1500EW_20180930_borisnand:...

Portfolio Visualizer Portfolio Analysis Results (Jan 2011 - Dec 2018) for the blend of the three equal weight ETFs:
https://www.portfoliovisualizer.com/backtest-portfolio?s=y&a...

Screen                                 CAGR   TR     GSD     MDD    Sharpe  Beta  AT
SP1500EqualWeight 10.11 116 18.79 -25.84 0.64 1.11 0.3
kindaLikeSP1500EW_20180930_borisnand 10.20 117 18.55 -25.41 0.65 1.10 0.4
PortfolioVisualizer_SP1500EqualWeight 8.61% 94 14.37% -21.87% 0.62 0.95


Year  GTR1   kindaLike    PV       avg_M*  RSP (Price)  EWSC (Price)  EWMC (Price)    delta
2011 -1.36 -0.47 -2.55% -2.54 -0.67 -6.72 -0.22 1.19
2012 18.16 16.52 14.87% 14.88 17.16 11.27 16.2 3.29
2013 38.98 40.16 37.83% 37.85 35.53 43.04 34.97 1.15
2014 9.04 7.77 8.45% 8.48 14.06 1.2 10.19 0.59
2015 -4.54 -2.66 -6.17% -6.17 -2.66 -10.89 -4.95 1.63
2016 23.93 20.98 20.76% 20.75 14.5 25.24 22.52 3.17
2017 14.34 17.67 13.89% 13.89 18.51 10.11 13.05 0.45
2018 -9.70 -10.55 -10.33% -10.41 -7.82 -10.97 -12.45 0.63


Portfolio Visualizer results match M* results.
delta = {SP1500EqualWeight} GTR1 results minus Portfolio Visualizer results.

The GTR1 results are consistently higher. This could be due to trading expenses. Or rebalancing. A more carefully constructed GTR1 screen could probably replicate RSP, but I don't need that much precision. {kindaLikeSP1500EW} is good enough for my uses.
Print the post Back To Top
No. of Recommendations: 0
... these screens are just the ones that I took interest in ... The standard MI screens might be a better proxy.

Averages2018 CAGR MDD SDcg
avgLongScreen -16 -32 10
avgIndex -7 -23 0.6

Averages2017 CAGR MDD SDcg
avgLongScreen 16 -18 12
avgIndex 21 -4 1.2


Interesting, so using the standard MI screens the 2017 delta is -5 and 2018 is -9? Happen to 2014 - 2016 handy in that same summary average format?

Thanks for the index urls, I did not have those bookmarked.
Print the post Back To Top
No. of Recommendations: 0
No summary data for earlier years, but full portfolio equity curve data is available at:
https://boards.fool.com/sip-screens-average-portfolio-values...
Print the post Back To Top
No. of Recommendations: 7
No one is using all of these screens. I think they have to go through some screen selection process
before evalutaion.


I took a quick look at how the screens sorted by how they had performed up till 1/2/2014 and then
their results from 1/2/2014 till 12/31/2018. Long term performance it the primary factor I have used
in screen selection. None of the top 10 screens equaled the SP500 equal weingt.
Only 8 of 59 screens beat the S&P 500 equal weight. Note 1.073 = 7.3% CAGR 2014 thru 2018

                      1/2/2014  12/31/2018  Gain   CAGR
SP500EqualWeight 759 1082 1.424 1.073
VG_Horse 125666 153206 1.219 1.040
Net_Nets_Grahamified 37359 36484 0.977 0.995
AdvancedHelper 34732 42524 1.224 1.041
LowPS_ 34671 52603 1.517 1.087
Advanced 29951 33331 1.113 1.022
Silver_Parachute 29248 30688 1.049 1.010
HighOnVolume 29092 29870 1.027 1.005
Blue_Skies 17011 22138 1.301 1.054
VG_Zebra 11786 11143 0.945 0.989
RS_100 9012 9779 1.085 1.016
YldEarnYear_SI 7870 7399 0.940 0.988
WK_Voom 7101 10452 1.472 1.080
Low_Mult 6454 9507 1.473 1.081
3pt_Relative_Value 6057 7289 1.203 1.038
78RPM 4925 4473 0.908 0.981
Gentle_Screamers 4847 3706 0.764 0.948
P_S_I_Love_You 4282 5304 1.239 1.044
LowPSvol 3539 3378 0.955 0.991
GS_PCF 3334 3640 1.092 1.018
High_Relative_Value 2644 3875 1.466 1.079
Dipstick 1990 853 0.429 0.844
POI 1621 2964 1.828 1.128
BI 1465 1502 1.026 1.005
Melange 1383 1636 1.183 1.034
Quality_Earnings 1191 1304 1.095 1.018
GSX 1142 966 0.846 0.967
GS_Mungo 995 1091 1.097 1.019
GS_Mungo_Voom 897 963 1.074 1.014
Shrinkage 825 1069 1.295 1.053
Small_Value 808 785 0.971 0.994
Fried_500 740 976 1.319 1.057
Value_at_the_Top 563 622 1.104 1.020
Rabbitt 535 544 1.016 1.003
Up5X3 376 555 1.477 1.081
POG 302 322 1.067 1.013
Incoming_Cash 292 315 1.078 1.015
Turnarounds 283 300 1.060 1.012
GSX2 263 351 1.337 1.060
SP500MktCapWeight 260 394 1.515 1.087
FCF_26 218 288 1.320 1.057
Microcap_Momentum 195 122 0.623 0.910
Money_Flow 167 169 1.015 1.003
Up_5_ 152 197 1.303 1.054
PIH_Naked 110 123 1.118 1.022
Bob 98 143 1.462 1.079
ARS 75 70 0.929 0.985
OptiMan 49 60 1.211 1.039
CANSLIM_26 47 54 1.155 1.029
Steady_Growth 34 42 1.254 1.046
Z26saTA 33 25 0.756 0.946
S_P_Peg 33 35 1.069 1.013
Zweig_26 27 28 1.026 1.005
WER 24 31 1.276 1.050
SP1500EqualWeight 10 13 1.347 1.061
Dilution_s 3 1 0.354 0.812
The_Mirror_s 3 2 0.690 0.928
Negative_FCF_s 1 1 0.645 0.916
HIAR_s 1 0 0.690 0.929
Current_Ratio_s 0 0 0.613 0.907


RAM
Print the post Back To Top
No. of Recommendations: 1
I found some 2016 data. This shows the average MI standard screen return was about the same as the index for 10 years to 2015, and in 2016.

Averages2016    CAGR_10yr  R1a  R1b
avgLongScreen 8 15 16
avgShortScreen -17 4 3
avgIndex 8 13 13


CAGR_10yr is the CAGR from 20051230 to 20151231 with 5 picks from each screen.
R1a is the CAGR from 20151231 to 20161230 with 10 picks from each screen.
R1b is the CAGR from 20151231 to 20161230 with 5 picks from each screen.
https://boards.fool.com/10-year-backtests-to-2016-32056663.a...


But there was a report of a MI CAGR 17.1% versus 6.8% CAGR for SPY from 2006 to 2013. So outperformance was possible. But just not from picking random MI standard screens.
year  TGMark    SPY   delta
2006 12.8% 10.6% 2.2%
2007 14.9% 5.1% 9.8%
2008 -6.0% -36.7% 30.7%
2009 35.6% 26.4% 9.2%
2010 38.7% 15.1% 23.6%
2011 -22.7% 1.9% -24.6%
2012 7.4% 16.0% -8.6%
2013 79.0% 32.3% 46.7%

https://boards.fool.com/8-years-of-mi-31055159.aspx


And Zee's Gold screens showed some good results from 2005 to 2012.
https://boards.fool.com/zees-sip-gold-7-years-30426910.aspx
Print the post Back To Top
No. of Recommendations: 2
I took a quick look at how the screens sorted by how they had performed up till 1/2/2014 and then their results from 1/2/2014 till 12/31/2018. Long term performance it the primary factor I have used in screen selection. None of the top 10 screens equaled the SP500 equal weight. Only 8 of 59 screens beat the S&P 500 equal weight. Note 1.073 = 7.3% CAGR 2014 thru 2018

Very helpful, RAM. So, if we average the top 10 long-term performing screens for a proxy of "careful MI picking" we get a past 5-year delta of -4.7%?:

 	
2014-2018 CAGR
SP500EqualWeight 7.3%

VG_Horse 4.0%
Net_Nets_Grahamified -0.5%
AdvancedHelper 4.1%
LowPS_ 8.7%
Advanced 2.2%
Silver_Parachute 1.0%
HighOnVolume 0.5%
Blue_Skies 5.4%
VG_Zebra -1.1%
RS_100 1.6%
Average 10 screens 2.6%


BTW: ~what period of time is "long-term"?
Print the post Back To Top
No. of Recommendations: 3
BTW: ~what period of time is "long-term"?

When you asked that I had a senior moment, I ditched my work and wasn't positive exactly how
I had calculated it. So redid the calcs:
Found start date and CAGR from start date till 1/2/1014 for each screen.
Sorted screens by CAGR till 1/2/2014.
Then found CAGR from 1/2/2014 to end of 2018 for each screen.
Came up with a slightly different set of top 10 screens.

Date                  StartDate  PreDiscv  CAGR Pre 13  CAGR_Post13
VG_Horse 03/02/87 31.4 35.1 4.0
Net_Nets_Grahamified 03/02/87 31.4 30.0 -0.5
AdvancedHelper 03/02/87 31.4 29.7 4.1
LowPS_ 03/02/87 31.4 29.6 8.7
Advanced 03/02/87 31.4 29.0 2.2
HighOnVolume 03/02/87 31.4 28.9 0.5
Blue_Skies 03/02/87 31.4 26.7 5.4
VG_Zebra 03/02/87 31.4 25.3 -1.1
RS_100 03/02/87 31.4 24.2 1.6
YldEarnYear_SI 03/02/87 31.4 23.7 -1.2
WK_Voom 03/02/87 31.4 23.3 8.0
Low_Mult 03/02/87 31.4 22.9 8.1
3pt_Relative_Value 03/02/87 31.4 22.6 3.8
78RPM 03/02/87 31.4 21.8 -1.9
Gentle_Screamers 03/02/87 31.4 21.8 -5.2
P_S_I_Love_You 03/02/87 31.4 21.3 4.4
Rabbitt 09/02/97 21.0 20.8 0.3
LowPSvol 03/02/87 31.4 20.6 -0.9
GS_PCF 03/02/87 31.4 20.3 1.8
High_Relative_Value 03/02/87 31.4 19.4 7.9
Microcap_Momentum 10/01/01 17.0 19.1 -9.0
Up5X3 09/02/97 21.0 18.8 8.1
Silver_Parachute 03/02/87 31.4 18.1 1.0
Melange 12/01/88 29.7 18.1 3.4
POI 03/02/87 31.4 17.6 12.8
POG 09/02/97 21.0 17.6 1.3
Incoming_Cash 09/02/97 21.0 17.4 1.5
BI 03/02/87 31.4 17.2 0.5
GSX2 09/02/97 21.0 16.8 6.0
Quality_Earnings 03/02/87 31.4 16.4 1.8
GSX 03/02/87 31.4 16.3 -3.3
GS_Mungo 03/02/87 31.4 15.8 1.9
GS_Mungo_Voom 03/02/87 31.4 15.4 1.4
Shrinkage 03/02/87 31.4 15.1 5.3
Small_Value 03/02/87 31.4 15.0 -0.6
Fried_500 03/02/87 31.4 14.7 5.7
Money_Flow 09/02/97 21.0 14.3 0.3
Up_5_ 09/02/97 21.0 13.8 5.4
Value_at_the_Top 03/02/87 31.4 13.7 2.0
Bob 09/05/00 18.1 13.4 7.9
WER 01/03/92 26.6 12.8 5.0
PIH_Naked 09/02/97 21.0 12.1 2.2
Turnarounds 03/02/87 31.4 11.2 1.2
Dipstick 03/02/87 31.4 10.7 -15.6
FCF_26 03/02/87 31.4 10.3 5.7
SP1500EqualWeight 10/03/94 23.9 10.1 6.1
ARS 09/02/97 21.0 10.0 -1.5
SP500EqualWeight 03/02/87 31.4 9.7 7.3
SP500MktCapWeight 03/02/87 31.4 8.3 8.7
OptiMan 09/02/97 21.0 7.9 3.9
CANSLIM_26 09/02/97 21.0 7.6 2.9
Z26saTA 12/31/01 16.8 7.4 -5.4
Zweig_26 12/31/01 16.8 6.2 0.5
Steady_Growth 09/02/97 21.0 5.9 4.6
S_P_Peg 09/02/97 21.0 5.8 1.3


I spent a considerable amount of time evaluating the results on the Portfolio123 site. The
post discovery results on their for pay screens haven't done much better. Even evaluated
screens posted by people I knew had some solid economic backgrounds, not much better.
The last 5 years have especially been bad for value all mechanical screens, not just value.

RAM
Print the post Back To Top
No. of Recommendations: 3
Came up with a slightly different set of top 10 screens.

Nice work. So maybe a -4.9% delta:


Top 10 Std 2014-2018 StartDate CAGR Pre 13 CAGR Post 13
SP500EqualWeight 7.3

VG_Horse 03/02/87 35.1 4.0
Net_Nets_Grahamified 03/02/87 30.0 -0.5
AdvancedHelper 03/02/87 29.7 4.1
LowPS_ 03/02/87 29.6 8.7
Advanced 03/02/87 29.0 2.2
HighOnVolume 03/02/87 28.9 0.5
Blue_Skies 03/02/87 26.7 5.4
VG_Zebra 03/02/87 25.3 -1.1
RS_100 03/02/87 24.2 1.6
YldEarnYear_SI 03/02/87 23.7 -1.2
Average Screen 28.2 2.4
Delta -4.9



For the Zee Gold list I'm getting a -2.1%:

Zee Gold 2014-2018 CAGR
SP500EqualWeight 7.3%

P_S_I_Love_You 4.4%
Silver_Parachute 1.0%
High_Relative_Value 7.9%
Up_5_ 5.4%
Average Screen 5.2%
Delta -2.1%

Print the post Back To Top
No. of Recommendations: 0
But there was a report of a MI CAGR 17.1% versus 6.8% CAGR for SPY from 2006 to 2013. So outperformance was possible. But just not from picking random MI standard screens.

Mark, it would be awesome to hear about your results 2014-2018, since the above post goes exactly to the same starting line we're discussing here, 1/1/2014.
Print the post Back To Top
No. of Recommendations: 1
I ran the first six screens above and confirmed the results. SP1500EqualWeight CAGR was 6.1.

GTR1 Results from 20140102 to 20181231. CAGR_10deep is screen 10 deep, CAGR_5deep is screen 5 deep.

screen                CAGR_Post13  CAGR_10deep  CAGR_5deep
VG_Horse 4 3.9 4.0
Net_Nets_Grahamified -0.5 -1.0 -6.0
AdvancedHelper 4.1 3.0 -2.4
LowPS_ 8.7 6.4 8.9
Advanced 2.2 2.9 1.0
HighOnVolume 0.5 -0.9 -4.9
average 3.2 2.4 0.1


CAGR_Post13 is from the post above and is for screen depth 10. (Results are slightly different than CAGR_10deep due to portfolio rebalancing differences. See below for details.)

The average CAGR_Post13 for all 52 screens was 2.1. Average of the best ten CAGR_Post13 was 7.9. There were only 7 screens that had higher CAGR_Post13 than SP1500EqualWeight: POI, LowPSplus, Low_Mult, Up5X3, WK_Voom, High_Relative_Value, and Bob. It would have been difficult to select from this list of screens and beat the market over the last 5 years.

details:
--------------------------
GTR1 reported CAGR:
21 portfolios with start and end values: P01start, P01end
CAGR01 = P01end/P01start - 1
CAGR = average of CAGRxx from 00 to 20

CAGR = (P00end/P00start + P01end/P01start + P02end/P02start + ... + P20end/P20start)/21 - 1

---------------------------
using average portfolio value:
PortAvg = average of Pxx from 00 to 20

CAGR_calc = PortAvgEnd/PortAvgStart - 1

CAGR_calc = (sum of PxxEnd from 00 to 20)/(sum of PxxStart from 00 to 20) - 1
-------------------------------

example1: 3 portfolios
P00start = 1
P00end = 1.1
P01start = 1
P01end = 1.1
P02start = 2
P02end = 2.4

CAGR = 13.33%
CAGR_calc = 15%
---------------------------
Print the post Back To Top
No. of Recommendations: 1
H52EarnPSmo_20140519_numerous
a monthly screen from the list in this lind, redone as a bullet screen with the lowest priced stock as last step, has NO average losing years 1997 forward before friction, despite requiring $1 million in ave. daily volume and price>$5; it would have bad losing years if you picked the worst start date, but may be a good candidate for averaging purchases over time.
After friction still has no average losing years and a 47.2% CAGR and beta .85 sharpe .86.

http://gtr1.net/2013/?~H52EarnPSmo_20140519_numerous:h21i1::...

pre<>
avg min max sd
CAGR: 54.18 24.03 80.71 13.91
GSD(1): 82.51 71.23 108.22 10.73
DIGSD(1; 0%): 76.38 69.19 89.53 5.29
LDD(1; 0%): 38.58 35.88 43.13 2.10
LDDD3: 19.98 11.08 29.55 5.11
MDD: -73.60 -86.12 -59.69 8.120
UI(1): 30.26 22.80 39.93 5.09
Sharpe(1): 0.92 0.61 1.23 0.17
Beta(1): 0.85 0.73 0.96 0.06
TI(1): 70.38 48.72 95.44 11.66
AT: 11.49 11.16 11.86 0.14

19971231 10.99
19981231 37.79
19991231 110.52
20001229 53.00
20011231 127.23
20021231 52.39
20031231 150.60
20041231 283.83
20051230 10.06
20061229 202.04
20071231 34.30
20081231 32.84
20091231 47.70
20101231 41.21
20111230 45.57
20121231 179.68
20131231 55.06
20141231 22.01
20151231 34.35
20161230 163.15
20171229 15.06
20181231 27.64
20190108 6.87

If you redo my bullet above as a 5 stock screen by changing the prior step to 10 stocks, with .4% friction has a sharpe of 1.13 and a beta of .8 and a cagr of 37.5%. Last 3 years combined would be 28.6% cagr so pretty good in a weak period for other screens.

http://gtr1.net/2013/?~H52EarnPSmo_20140519_numerous:h21i1f0...


Print the post Back To Top
No. of Recommendations: 1
I ran the top 6 screens again, but with no friction. Average CAGR improved from 2.4 to 5.7. That is close to SP1500EqualWeight CAGR 6.1. These screens matched the market, but did not cover their trading costs over the past 5 years.

Screen (10 deep, no friction)  CAGR  AT
Advanced 7 10
AdvancedHelper 7 9
HighOnVolume 3 10
LowPSplus 11 11
Net_Nets_Grahamified 0 3
VG_Horse 6 5
average 5.7 7.9


Maybe with the big move to index funds,ETFs, and quant funds, the market has moved more in lockstep the last 5 years. This could be an improvement in market efficiency such that MI screens will not work. Or maybe the inefficiency spring has just been compressed, and there is a big move ahead.
Print the post Back To Top
No. of Recommendations: 0
This could be an improvement in market efficiency such that MI screens will not work. Or maybe the inefficiency spring has just been compressed, and there is a big move ahead.

Those are the million dollar questions. The SiPro universe has sure been weak for quite a while.
I wonder if the same is generally true for the standard VL screens.


Mark
Print the post Back To Top
No. of Recommendations: 2
Maybe with the big move to index funds,ETFs, and quant funds, the market has moved more in lockstep the last 5 years.

I was just thinking how it would be informative to compare performance for the same 5 years to the classic ETF strategies here, e.g. GEM and QTAA. But with the current Yahoo outage we can't actually do that, can we?

I guess this screen approximates BCC on RSP for a 6% CAGR vs RSP of 7%?
https://gtr1.net/2013/?~SP500EW_sell_if_All_Bearish_Don27298...
Print the post Back To Top
No. of Recommendations: 1
I've been catching up on MI while self-isolating. Do you have a similar table for all the standard SI Pro screens, with both the CAGR and GSD calculated for each screen since discovery? From that data I'd like to calculate if the best screens are just getting lucky from random variation, or if they are statistically significant.

I saw a similar exercise for sports analytics on fivethirtyeight.com. If you have 5 sports teams that are equally good, you can predict the distribution of their win-loss records for the entire season. If they are all equally good (e.g the game is just rolling dice), they probably won't all be tied. Some will finish ahead of the others because of pure luck. If one of the teams is much better than the others and isn't just lucky, it will have much more wins than is would be predicted based on the probabilities from random variation.
Print the post Back To Top
No. of Recommendations: 1
Standard screens' Portfolio values are in a download link in post 273008.
Print the post Back To Top
No. of Recommendations: 2
Print the post Back To Top