Post New

Has anyone done any research into what, specifically, caused YEY to flatten out a decade ago, then come roaring back in the 2020's? I've heard mention that inclusion of MLPs may have contributed to the first demise, but I assume they're still in the current stock selection universe.

If we can identify a measurable trend shift or critical metric epoch, then we could improve our returns by moving between YEY and the new ROE_Cash screen rather than staying fully invested in both through thick or thin times.

Taz

Has anyone done any research into what, specifically, caused YEY to flatten out a decade ago, then come roaring back in the 2020's?

Just a thought. I Closed out all VL screens in Mar-Apr 2014 including YEY. Could it be that most MI'ers and GRT1 switch to SiPro have something to do with YEY going "FLAT"?

GD_

MLPs didn't affect YEY much. YEY includes US common (10,11), foreign (12), REIT (18, 48), ADR (30, 31), and LP (71,71).

styp.a = 0!10!11!12!18!30!31!48!71!72

Screen                   CAGR  SAWR  GSD  MDD  Sharpe    From       To
1500ew                    11     8   22   -60   0.50   19870302  20240108
YldEarnYear_SI            25     9   32   -61   0.84   19870302  20240111
YldEarnYear_SI_noMLP      25    10   31   -58   0.84   19870302  20240111
YldEarnYear_SI_UScommon   15     7   29   -69   0.59   19870302  20240111
 
1500ew                     9     7   23   -42   0.48   20140110  20240111
YldEarnYear_SI             7     5   37   -59   0.33   20140110  20240111
YldEarnYear_SI_noMLP       8     6   36   -57   0.38   20140110  20240111
YldEarnYear_SI_UScommon   10     7   34   -60   0.44   20140110  20240111

https://gtr1.net/2013/?~YldEarnYear_SI:h21f0.4::st...
https://gtr1.net/2013/?~YldEarnYear_SI_noMLP:h21f0...
https://gtr1.net/2013/?~YldEarnYear_SI_UScommon:h2...
https://gtr1.net/2013/?~1500ew:h63f0.4::pref%28sp1...

Has anyone done any research into what, specifically, caused YEY to flatten out a decade ago, then come roaring back in the 2020's? I've heard mention that inclusion of MLPs may have contributed to the first demise, but I assume they're still in the current stock selection universe.

If we can identify a measurable trend shift or critical metric epoch, then we could improve our returns by moving between YEY and the new ROE_Cash screen rather than staying fully invested in both through thick or thin times.

Or you might want to read the book Fooled by Randomness by Nassim Taleb.

Elan

Or you might want to read the book Fooled by Randomness by Nassim Taleb.

Were folks fooled when we thought it worked? or when we thought it didn't work?

Seems like there are three main possibilities to explain these observations about screen performance:
- fooled by randomness
- post discovery or herd effect
- conditions at certain times favor a strategy and at other times they don't

YEY would seem not to be in the second camp, as it performed well for years after it was discovered (as I recall - have not used it or studied it).
I guess it could be the first, or the third.

For the first camp, we'd have to have randomness on the scale of years to explain the observations. I guess that is possible.
Usually, I'm mentally in the third camp. One could run ANOVA type statistical analyses on YEY performance using different possible factors.
For example whether fed policy/interest rates are rising, falling, or stable. Is performance correlated to employment numbers, other macro economic conditions.
Or to the dominance of mega-caps? Or even which party is in power? (I presume not).
These larger scale factors will come and go; interest rates will rise and fall, and mega-caps will not always be in favor. So who knows.

Coincidentally, I happened to be looking a few days ago at the Big_Jump screen which there was a thread about back in 2015.
Robbie and Ray were going at it, and Robbie posted this http://www.datahelper.com/mi/search.phtml?nofool=y... which showed a 42% return from 1987-2015.
What do you think it was from 2015 to the end of the pandemic? It crashed, 0% return through to the bottom of the pandemic in late March 2020.
And from then until now? It's back at about 46% CAGR, indistinguishable from the earlier period 1987-2015.

SiPro screens are much harder to interpret than something like YEY.
But my feeling is that screens can suddenly "work" again, that it's not just randomness.
I don't trust my feelings though, so a deep dive into screen performance would indeed be nice.


Mark

My simple minded check for randomness is to never invest in a screen that doesn’t outperform
with a hold of a much larger number of equities than you intend to invest in. YEY has still done
well from 1987 to 2024 with daily start tests holding 50 stocks with 0.4% friction.
CAGR 15.2%
SAWR 8.6%
MDD -55
UI 8.96
AT 4.3
http://gtr1.net/2013/?!!QlpoMTFBWSZTWUUuQ08AAWZfgB...

RAM

Robbie and Ray were going at it, and Robbie posted this http://www.datahelper.com/mi/search.phtml?nofool=y... which showed a 42% return from 1987-2015.

Sorry, that link doesn't work.

Robbie's post 255260 on that screen was 3-1-2015 and this is the URL he posted and referred to top 10 having CAGR 42.42% (now it shows a tad higher, through 2-27-2015):
https://gtr1.net/2013/?~BigJump%28I%29:e20150227s:...

This is the same screen, from 2-27-2015 through the pandemic bottom 03-16-2020, top 10 0%:
https://gtr1.net/2013/?~BigJump%28I%29:s20150227e2...

And this is from the pandemic bottom until now, top 10 45%:
https://gtr1.net/2013/?~BigJump%28I%29:s20200316s:...

I think all the quant ETFs have a lot to do with the problem.

The CAGR for 2014 thru 2023 is 6.88

Seems like there are three main possibilities to explain these observations about screen performance:
- fooled by randomness
- post discovery or herd effect
- conditions at certain times favor a strategy and at other times they don't

I think it's a combination of all three. A screen may be fundamentally sound, but the random variability of returns over periods of one year or even five years, or more, is greater than what we tend to believe. And conditions may change, but you can only see such changes in hindsight, and they will not be predictive of the next change.

Elan

If we can identify a measurable trend shift or critical metric epoch, then we could improve our returns by moving between YEY and the new ROE_Cash screen rather than staying fully invested in both through thick or thin times.

How about a version of WWL, what's working lately? Check to see which of the two screens has performed better over the last six or 12 months and go with the winner.

DB2

DrBob2: How about a version of WWL, what's working lately?

Here the first is a simple WWL comparison just taking the 4,3,2 weighted sum of the last 3 years at the end of the year
and infesting in largest sum for the next full year gives a 21.2% CAGR over the last 24 years.
Second is just taking the larger of ROE or YEY last end of year return and investing in the higher one gives a 26.3% CAGR.
Third and fourth are just the ROE and YEY results for same period, YEY's return is good but with volatility and a 55% MDD.

 	      ROE_Cash YEYr   ROE_Cash	YEYr     WWL	WWL    ROE_Cash	YldErnYr Blender
			      3y_Wt	3y_Wt   4+3+2	1Yr			
19971231	-2.52	16.96							
19981231	101.62	24.98							
19991231	95.94	-2.79	683.6	97.7	1	1	1	1	1
20001229	2.38	39.67	500.6	200.3	1.024	1.024	1.024	1.397	1.343
20011231	-12.98	55.99	147.1	337.4	0.891	1.597	0.891	2.179	1.733
20021231	-9.03	30.46	-70.3	369.1	1.162	2.083	0.810	2.842	1.954
20031231	28.74	88.60	61.9	557.8	2.192	3.929	1.043	5.361	3.151
20041231	11.69	44.46	114.9	504.6	3.167	5.676	1.165	7.744	4.039
20051230	11.85	36.64	139.9	457.1	4.327	7.756	1.303	10.581	4.590
20061229	16.18	48.58	123.7	393.2	6.429	11.524	1.514	15.722	5.815
20071231	20.15	103.76	152.8	634.1	13.099	23.482	1.819	32.035	9.076
20081231	-31.68	-22.97	-33.9	316.6	10.090	18.088	1.243	24.676	6.854
20091231	50.70	20.15	148.0	219.2	12.123	21.732	1.873	29.648	9.290
20101231	19.53	48.99	166.9	210.5	18.063	25.976	2.239	44.174	11.94
20111230	-0.25	16.43	159.0	253.0	21.030	30.244	2.233	51.431	13.30
20121231	8.68	44.62	73.0	325.8	30.414	43.739	2.427	74.380	17.16
20131231	51.33	55.35	230.9	388.1	47.247	67.947	3.673	115.545	25.46
20141231	11.19	4.82	216.1	274.6	49.526	71.224	4.084	121.119	27.55
20151231	3.11	7.32	148.7	154.5	53.153	73.436	4.211	129.990	28.22
20161230	17.50	2.26	101.7	40.6	54.353	75.093	4.948	132.923	30.61
20171229	37.66	4.32	209.4	38.7	74.824	103.377	6.812	138.663	35.66
20181231	-12.69	-16.07	97.2	-46.8	65.330	90.259	5.948	116.379	34.42
20191231	47.01	27.52	225.3	70.5	96.042	132.692	8.744	148.402	42.32
20201231	31.48	-6.92	241.6	22.7	126.274	174.459	11.496	138.135	48.26
20211231	24.46	48.97	286.3	230.2	157.165	217.138	14.308	205.781	66.22
20221230	-39.69	18.29	-22.4	206.2	94.790	256.851	8.630	243.416	64.16
20231229	48.70	5.71	124.7	175.7	100.206	271.527	12.832	257.3	74.386
									
				CAGRs	21.2	26.3	11.2	26.0	19.70%

Don't think this proves much of anything as we are just playing with the ones we already know are the winners.
What WWL needs is a algorithm that takes into consideration all screens at a point in time and ranks them by their probibility
of success in the next month.

RAMc

Post New

Software as a Service Gems FAQ

---

Contact Shrewd'm
Contact the developer of these message boards.