Tiny Titans

Has anyone used the tiny titans of AAII? It claims to have returned 24% a year.

Has anyone used the tiny titans of AAII? It claims to have returned 24% a year.

I tried a version that used essentially the same parameters except aiming for less slippage and volitility of the Tiney Titans which I called Mid-Titans. . . Looked good on backtest but didn’t hold up post discovery. Here is borasnand’s GTR1 version of Tiney Titans posted last year.
http://gtr1.net/2013/?~TinyTitans_20150401_numerous:s2007101…

CAGR from 20071015 is 17.76 but notice the GSD is 34.4, UI is 20 SAWR is only 8.8.
Might be a candidate for a young risk taker but there are better choices for seniors like me.

RAM

6 Likes

Thanks. A little disappointing. Looking at aaii’s site, you would think you could become fabulously wealthy using it, but my 15 years here have taught me that is never the case.

1 Like

Looking at aaii’s site, you would think you could become fabulously wealthy using it, but my 15 years here have taught me that is never the case.

That’s because nobody writes any articles “This strategy is a pile of crap, it will lose all your money.” Also, nobody writes any articles “This strategy worked great [only] one time in history.”

10 Likes

{Tiny Titans} results depend on low dollar volume stocks, and so I haven’t used it. I like that it applies the momentum criteria early, and the use of low PS. The AAII annual results are close to the gtr1 results:

        Tiny Titans
 year       AAII     gtr1
 2021       67        52
 2020       42        56
 2019       31        38
 2018       13         0
 2017       -7        -2
 2016       21        35
 2015       -7        -5
 2014       26        28
 2013       54        86
 2012       20        35
 2011       -23      -10
 2010       21        35
 2009       71        46
 2008       -56      -49
 2007        2         6
 
  avg       18        23
  sd        34        34
 cagr       13        18
correl      1.00     0.91

Adding a dollar volume liquidity criteria shows that most of the excess returns are below daily dollar volume of $500k.

  adv>=    CAGR  GSD
 **0   25   31**
 **100,000   25   34**
 **250,000   22   38**
  500,000   14   42
  750,000    9   46
1,000,000    7   52
2,000,000  -15   86
5,000,000  -20   85

My friction estimate for this low dollar volume range is about 1.5%, and so I will use that for {TinyTitans}. The market cap limit of $250 million is passing fewer stocks now than 20 years ago, and so I switched to use bottom 50%. With the high friction, a quarterly hold time works as well as monthly. With these changes, {TinyTitans_2022a} CAGR since 20061229 was 6, which is less than {SP1500equalweight} CAGR 10, and less than {TinyTitans_20150401_numerous} CAGR 18. TinyTitans might be fool’s gold.

http://gtr1.net/2013/?~TinyTitans_20150401_numerous:h21f0.4:…
http://gtr1.net/2013/?~TinyTitans_2022a:h63f1.5::PS:gt0:Sale…
https://gtr1.net/2013/?~SP1500EqualWeight:h63f0.4::pref%28sp…

        Tiny Titans  2015  2022
 year       AAII     gtr1  gtr1  SP1500ew
 2021       67        52    47      31
 2020       42        56    23      14
 2019       31        38     3      25
 2018       13         0   -14     -10
 2017       -7        -2    -3      14
 2016       21        35    20      24
 2015       -7        -5    -8      -5
 2014       26        28    11       9
 2013       54        86    56      39
 2012       20        35    24      18
 2011       -23      -10   -21      -1
 2010       21        35    28      26
 2009       71        46    13      46
 2008       -56      -49   -39     -36
 2007        2         6    -5      -1
 
  avg       18        23     9      13
  sd        34        34    25      21
 cagr       13        18     6      11
correl      1.00     0.91  0.83    0.88
16 Likes

Thanks for reporting your investigations into the Tiny Titans screen and the gtr1 versions thereof.

The market cap limit of $250 million is passing fewer stocks now than 20 years ago, and so I switched to use bottom 50%.

That seems a huge change, if I understand correctly, from microcap to including stocks up to the median market cap. I wonder what that 50% cutoff is? Maybe a change that would hold truer to the intent of the screen while increasing the number of stocks it returns would be to bump up the market cap limit up incrementally until you get as many stocks as you desire. Or maybe just double it to $500 million, which would move the universe up out of microcaps, but not too far.

I’m an aaii subscriber, and while I’ve not paid much attention to the Tiny Titans screen, I have used their Shadow Stock model portfolio screen as a fishing ground for small and micro-cap value stocks. I do buy into the premises that in the micro-cap and low end small-cap universe you are more likely to find blatant mis-pricing, due to so much less attention paid to that sector. I’ve had some hits and misses … not been doing it long enough to judge whether my using it to sprinkle in some micro-cap value into my portfolio will pay off.

1 Like

Good catch. Bridgeway Ultra-Small (BRSIX) mutual fund has about 500 holdings with average market cap $170 million, and {TinyTitans} backtests better with the $250 million market cap limit.

{TinyTitans_2022b} results are similar to {TinyTitans_20150401_numerous}. {TinyTitans_2022b} uses:
Market Cap < 250
average daily dollar volume > 10,000
monthly hold, friction 0.4

But friction might be higher with such a low dollar volume. The returns depend on trading skills that keeps trading costs low.

http://gtr1.net/2013/?~TinyTitans_2022b:h21f0.4::PS:gt0:Sale…

3 Likes

With the caveat that nothing in real life works as well as it does in models and back tests, it seems, you could easily make trades of 2,000.00 without much friction.

I just ran the screen at aaii, and 25 stocks came up, so it seems you get enough picks.

How difficult would it be to add them to the weekly picks? I remember reading about them, and he said you have to hold 25. I just reread the section on them, and he recommends holding for a year.

That book is pretty old, so I don’t know if holding for a year still works, although it could because institutions don’t buy stocks like that.

I tuned Tiny Titans and ended up with {TinyTitans_2022c}. To limit turnover, I added htd and no rebalancing to get {TinyTitans_2022c2}.

      Screen         **CAGR** SAWR  GSD  MDD  Sharpe  AT   Depth
TinyTitans_20150401   **25**  11   31   -70   0.91   4.7    70
  TinyTitans_2022b    **25**  10   30   -70   0.93   4.6    63
  TinyTitans_2022c    **27**  13   31   -66   0.95   8.3    10
 **TinyTitans_2022c2    33    15   32   -66   1.07   3.2    10**

http://gtr1.net/2013/?~TinyTitans_20150401_numerous:h21f0.4:…
http://gtr1.net/2013/?~TinyTitans_2022b:h21f0.4::PS:gt0:Sale…
http://gtr1.net/2013/?~TinyTitans_2022c:h21f0.4::styp.a:et10…
http://gtr1.net/2013/?~TinyTitans_2022c2:h21r0f0.4::styp.a:e…

{TinyTitans_2022c2}
step0 to step 7: liquidity criteria
step8: [Total Return % over 204 days] Top 15%
step9: [PS] < 1
step10: US company
step11: [pctMktCap after step7] <= 18
step12: [MktCap] Bottom 40
step13: security type is US with no REITs
step14: [Closing price]/[Highest daily price over 252 days] Top 10

Command Translation
Define {TinyTitans_2022c2}
step0: [Security Type; lag=1 days] == 10,11,18,48
step1: [Mkt Days Since Security Opened; lag=1 days] >= 252
step2: [Average dollar-volume over 63 days; lag=1 days] >= 20000
step3: [Average dollar-volume over 10 days; lag=1 days] >= 20000
step4: [Actual closing Price; share_lag=1 days; quote_lag=1 days] >= 2
step5: [MktCap] > 0
step6: [Sales] >= 50
step7: [PS] > 0
step8: [Total Return % over 204 days; lag=1 days] Top 15%
step9: [PS] < 1
step10: country is US
step11: [pctMktCap] <= 18
step12: [MktCap] Bottom 40
step13: [Security Type; lag=1 days] == 10,11
step14: [PH52i] Top 10; Cash When Fewer
Holding period = 21 mkt days
Hold while passing step11
Frictional loss of 0.4% applied to all sales
Rebalance to equal weight among new positions only
All trades at market close.

15 Likes

{TinyTitans_2022c} is somewhat similar to {LowPS+}. To make a comparison, I changed the liquidity and momentum criteria to match and ended up with {LowPSplus2022a}. The results are very close with both screens around CAGR 33 GSD 32. Adding a $200k daily dollar volume limit, {TinyTitans_2022c} CAGR is 27, and {LowPSplus2022a} CAGR is 27. With a $500k daily dollar volume limit, {TinyTitans_2022c} CAGR is 12, and {LowPSplus2022a} CAGR is 22.

      Screen       **CAGR** SAWR  GSD  MDD  Sharpe  AT   Depth
**TinyTitans_2022c2   33    15   32   -66   1.07   3.2    10**
 **LowPSplus2022a    34    14   31   -61   1.09   7.2    10**
      LowPS+        **29**  10   36   -63   0.89   11     10

https://gtr1.net/2013/?~LowPS+:h21f0.4::styp.a:et0!10!11!12!..
http://gtr1.net/2013/?~LowPSplus2022a:h21r0f0.4::styp.a:et10…

{LowPSplus2022a}
step0 to step 7: liquidity criteria (same as {TinyTitans_2022c})
step8: [Total Return % over 204 days] Top 15%
step9: US company
step10: [Closing price]/[Highest daily price over 252 days] > 0.89
step11: [PS] bottom 10

Command Translation
Define {LowPSplus2022a}
step0: [Security Type; lag=1 days] == 10,11,18,48
step1: [Mkt Days Since Security Opened; lag=1 days] >= 252
step2: [Average dollar-volume over 63 days; lag=1 days] >= 20000
step3: [Average dollar-volume over 10 days; lag=1 days] >= 20000
step4: [Actual closing Price; share_lag=1 days; quote_lag=1 days] >= 2
step5: [MktCap] > 0
step6: [Sales] >= 50
step7: [PS] > 0
step8: [Total Return % over 204 days; lag=1 days] Top 15%
step9: country is US
step10: [PH52i] > 0.89
step11: [PS] Bottom 10; Cash When Fewer
Holding period = 21 mkt days
Frictional loss of 0.4% applied to all sales (closing long/opening short)
Rebalance to equal weight among new positions only
All trades at market close.

 year   SP1500equalWgt  LowPSplus2022a  TinyTitans_2022c2
 1987         5               -1               -18
 1988         20              15                30
 1989         26              42                41
 1990         -12             11                13
 1991         35              104              100
 1992         15              78                77
 1993         15              62                45
 1994         1               7                 25
 1995         28              78                41
 1996         21              44                50
 1997         27              60                46
 1998         6               51                37
 1999         11              29                36
 2000         14              63                67
 2001         10              53                49
 2002         -15             18                15
 2003         44              148               97
 2004         19              76                68
 2005         8               43                30
 2006         16              40                62
 2007         -1              3                 10
 2008         -36             -29              -54
 2009         46              41                87
 2010         26              56                51
 2011         -1              -12               -4
 2012         18              37                39
 2013         39              61                85
 2014         9               40                40
 2015         -5              -6                4
 2016         24              13                52
 2017         14              27                13
 2018         -10             -5                -1
 2019         25              23                52
 2020         14              6                 44
 2021         31              124               48
 2022         -6              -4                -9
 
  avg         13              39                38
  sd          17              38                33
 cagr         12              31                33
correl       0.86            0.79             1.00

{LowPS+} was defined in 2006 in the post https://discussion.fool.com/alternate-si-rankings-09jun06-242163…

9 Likes

Thank you very much Boris.

2 Likes

Another tweak, check every other week rather than once a month and discard the 25% high beta’s.
beta(1,190,1) bottom 75%

http://gtr1.net/2013/?~TinyTitans_2022_2wkLowBeta:h10r0f0.4:…

CAGR 33.7%, SAWR 15.85, GSD 31.9, MDD 66.5, Sharpe 1.08, AT 4.3

6 Likes

{TinyTitans_2022d} is a version using ranks that has similar results. The htd might be easier to manage. No rebalancing reduces turnover slightly and does not affect results much.

Screen             CAGR  SAWR  GSD  MDD  Sharpe  AT
TinyTitans_2022c2   33    15   32   -66   1.07   3.2
TinyTitans_2022d    34    15   32   -64   1.11   2.6

Command Translation
Define {TinyTitans_2022d}
Create [rA]: [Rank by [MktCap] (Asc) at step8]
Create [rB]: [Rank by [PS] (Asc) at step8]
Create [rD]: [Rank by [PH52i] (Desc) at step8]
Create [rC]: [Rank by [Total Return % over 186 days; lag=8 days] (Desc) at step8]
Create [rAll]: [[rA] + [rB] + [rC] + [rD]]

step0: [Security Type; lag=1 days] == 10,11
step1: [Mkt Days Since Security Opened; lag=1 days] >= 252
step2: [Average dollar-volume over 63 days; lag=1 days] >= 20000
step3: [Average dollar-volume over 10 days; lag=1 days] >= 20000
step4: [Actual closing Price] >= 2
step5: [MktCap] >= 25
step6: [Sales] >= 50
step7: [PS] > 0
step8: [rAll] Bottom 100
step9: [rAll] Bottom 10
Holding period = 21 mkt days
Hold while passing step8
Frictional loss of 0.4% applied to all sales
Rebalance to equal weight among new positions only
All trades at market close.

http://gtr1.net/2013/?~TinyTitans_2022d:h21r0f0.4::styp.a:et…

7 Likes

{TinyTitans_2022d} is sensitive to liquidity criteria. Limit orders might be needed to avoid excessive friction.

adv(1,63)>=  CAGR  GSD
     0       34.4  31.1
  10,000     34.3  31.1
  20,000     34.0  31.0
  50,000     33.3  31.6
  100,000    31.9  32.7
  200,000    28.7  34.4
  400,000    21.7  35.1
  800,000    17.9  36.3
 1,600,000   13.7  36.1
3 Likes