1. Assessing the Asymmetric Information
Associated with the Equity Market:
A CART Based Decision Rule Analysis
Owen P. Hall, Jr., P.E., Ph.D.
Pepperdine University
CART Conference
May, 2012
San Diego, CA

2. Presentation Agenda
 Overview
 Problem Statement
 Results Analysis
 Conclusions

3. Problem Statement
Assess the effectiveness of analytics to
detect asymmetric information associated
with the equity market
 Models
• Probabilistic Neural nets
• CART
 Factors
• Classic (e.g., Price Momentum)
• Tobin’s Q
• Entropy

4. Challenge
 In an efficient market, the current
prices of securities represent
unbiased estimates of their true or
fair market value at all times
 This principle suggests that neither
technical analysis nor fundamental
analysis can assist investors in
identifying undervalued or
overvalued stocks
I'd be a bum in the street with a tin cup if the markets were efficient
-- Warren Buffett

5. Classic Factors
 Price Momentum
 Earnings Momentum
 Valuation
 System
 Economic

6. Entropy
 The basic idea is that more volatile
securities have a greater entropy state than
more stable securities
 Two fundamentally different phenomena
exist in which time based securities data
deviate from constancy:
 Exhibit larger standard deviations
 Appear highly irregular
 The standard deviation measures the extent
of deviation from centrality while entropy
delineating the extent of irregularity or

7. Entropy
 Two entropy models
 Approximate entropy (ApEn)
 Sample entropy (SaEn)
 Model inputs
 Time series
 Matching template length (M)
 Matching tolerance level (r)
 Time series length (50 months)

8. Tobin’s Q
Q = Market value / Replacement value
 Reflects the expected current and future
profitability of capital
 Q values less than one identify under
valued equities
 Q values greater than one suggest than
capex will increase share holder wealth
 Q values less than one suggest making
acquisitions is cheaper than capex

9. Tobin’s Q
(US Market)

10. Valueline Timeliness Ranks
(1965 – 2009)
Rank Weekly (%) Yearly (%)
1 15,575 30,778
2 10,727 4,174
3 4,924 252
4 2,846 - 60
5 5,266 -99

11. Database
 2008 (4) – 2010(1) – 6 Quarters
 Sources
 Value Line Investment Survey
 Ford Equity Research
 Mergent Online
 Sample Size (100 ~ 400)
 Target Variable – PGQ (binary- lagged)

12. Two Step Analytic Process
 Screen variables with
neutral nets
 Develop decision rules
using CART
 Holdout Assessment

13. Probabilistic Neural Networks
 An extension to the classical backward
propagation neural net
 Non-parametric
 “Black Box”
 Results often difficult to interpret and
operationalize

14. Neural Nets

15. CART
 Non-parametric
 Interactive effects
 Non-normally distributed variables
 Decision tree logic makes it easier to
apply model outcomes
 Model is extremely robust to the effect of
outliers
 Results easy to interpret and implement

16. CART Tree

17. Neural Net Results
Rank 8-4 9-1 9-2 9-3 9-4 10-1
1 PSS ROA PSS SMO CNE PRM
2 PRM SUE PVA PSS EMO Q
3 PVA PSS SEP ROA SMO ROA
4 ROA SMO PRM EMO VMO VMO
5 SEP EMO SMO SEP PEG EMO
6 VMO SEP Q Q Q SMO
7 EMO PRM VMO PRM SUE PSS
8 SUE PVA PEG EMO PRM PER
9 PEG PEG EMO PEG PSS SEP
10 SMO VMO ROA PVA SEP COM

18. Classification Analysis
(9/4 -> 10/1)
Actual
Predicted 1 0
1 31 15 67% PPV1
0 16 33 67% NPV2
Total 47 48
66% 69%
Sensitivity Specificity
PPV = ratio of the number of winners classified correctly divided by the total number of securities classified as winners.
1
NPV = ratio of the number of losers classified correctly divided by the total number of securities classified as losers.
2

19. Results
(Modified Sharp Ratio)
Case Qtrs./ Quarter Value Going NSI Selling NSI
Sample Line Long Short
Size Ones
1 1/89 9-2 0.289 0.392 38 0.210 53
2 1/91 9-3 0.775 0.853 51 -0.022 37
3 1/88 9-4 1.177 0.771 53 -0.043 40
4 1/93 10-1 0.513 0.553 38 0.485 56
5 1/94 10-2 -0.580 -0.328 46 -0.583 49
6 2/180 9-3 0.775 0.800 23 0.789 65
7 2/179 9-4 1.177 0.598 62 0.749 31
8 2/181 10-1 0.513 0.514 49 0.512 45
9 2/187 10-2 -0.580 -0.498 59 -0.728 36
10 4/361 10-1 0.513 0.613 49 0.418 45
11 4/366 10-2 -0.580 -0.493 70 -0.605 25

20. Conclusions
 Modeling approach generally
performed as well or better than
Valueline 100
 CART results provide an
operational strategy
 Adding transaction costs reduces
model effectiveness
 Portfolio size based on binary
target variable remains
problematical

21. Future Research
 Expand data set from 6 to 12
quarters
 Ternary classification target
 Variable selection optimization
 Add economic factors
 CPI
 UEM
 Explore “super” factors
 Q / ApEn
 PRM / SpEn

22. Thanks for Listening!