Aug

19

A Master, from David Lamb

August 19, 2007 |

 I may be using the wrong term here but what really makes a person a "master" of something? And, what does it take to master something? I had this conversation on the golf course the other day with the other three and we came up with four different ideas. Mine was the least kind.

Does being a master mean that one is at the top of the class in a particular endeavor? Does being a master mean there is no room for improvement, that perfection has been achieved? Webster's states a master (in one of its definitions) is one being an expert at something. Can one fall from being a master, such as the Master's Championship in golf, where he is no longer the master?

Can a professional baseball player who hits .400 be called a master at his trade if he only accomplishes a 40% success ratio? Can a trader ever master the markets, at least for more than a brief moment? Perhaps a master is the person who is the best, or one of the best, at any given moment.

Kim Zussman writes: 

It seems impossible to get more than a small percentage of mostly random market movements right. However even doing that does make a difference, and a critical question a trader should try to answer is whether they add any value over buy and hold. And the question should be asked under different market conditions: Bear, bull, and flat.

There are lots of ways to ask this question, but here is an example using mutual fund manager John Hussman's HSGFX. (I don't have any position, but he is interesting and runs a mutual fund with option hedges, was finance professor, and his research seems thorough, though he's been bearish for a while.) I did regressions of his weekly performance vs. SPY (both w/div) over three periods:

11/00-3/03 Bear
3/03-6/05 Bull
6/05-8/07 Recent

Actually he does pretty well.  First during bear mkt 11/00-3/03:

Regression Analysis: HUSBEAR versus SPYBEAR

The regression equation is
HUSBEAR = 0.00340 - 0.0257 SPYBEAR

Predictor      Coef        SE Coef       T       P
Constant   0.003399  0.001135   3.00  0.003
SPYBEAR  -0.02569    0.03942   -0.65  0.516

S = 0.0124754   R-Sq = 0.4%   R-Sq(adj) = 0.0%

highly signif alpha with flat beta

Here is the same for bull period, 3/00-6/05:

Regression Analysis: HUS BUL versus SPY BUL

The regression equation is
HUS BUL = 0.00116 + 0.469 SPY BUL

Predictor       Coef           SE Coef      T      P
Constant       0.00116    0.00064    1.79  0.076
SPY BUL        0.46866    0.04235  11.07  0.000

S = 0.00674420   R-Sq = 52.2%   R-Sq(adj) = 51.8% 

Alpha is not quite significant, but he must have removed the hedges to get beta to 0.5 and extremely signficant correlation. Finally here he is recently, 6/05 to last week:

Regression Analysis: HUS REC versus SPY REC

The regression equation is
HUS REC = 0.00123 - 0.0652 SPY REC

Predictor       Coef            SE Coef        T       P
Constant     0.0012253  0.0004826   2.54  0.012
SPY REC     -0.06516      0.03213     -2.03  0.045

S = 0.00510472   R-Sq = 3.5%   R-Sq(adj) = 2.7%

Alpha again is highly significant, but now with negative beta.

This suggests he knows how to make money in up and down markets, partly as consistent alpha and partly when to turn up beta. You can see it also in one-sample t-test vs H(0) of return = 0:

One-Sample T: HUSBEAR, SPYBEAR, HUS BUL, SPY BUL, HUS REC, SPY REC
Test of mu = 0 vs. not = 0

Variable      N       Mean     StDev       SE Mean          95% CI                T         P  

HUSBEAR  122   0.00347  0.01244  0.00112  ( 0.00124, 0.00570)    3.08   0.003
SPYBEAR   122  -0.00279  0.02876  0.00260  (-0.00795, 0.00235)  -1.07   0.285
HUS BUL   114   0.00262  0.00971  0.00091  ( 0.00082, 0.00443)    2.89   0.005
SPY BUL    114   0.00313  0.01498  0.00140  ( 0.00035, 0.00591)    2.23   0.027
HUS REC   114   0.00109  0.00517  0.00048  ( 0.00013, 0.00205)    2.25   0.026
SPY REC    114   0.00204  0.01494  0.00140  (-0.00072, 0.00481)   1.46   0.147

There seem to be three ways to beat the market:

1. Significant alpha with insignificant beta
2. Insignificant alpha, but significant positive beta in up-markets (market timing)
3. Combination of 1+2

This kind of analysis doesn't factor in volatility of returns, which is also important. There is less value in matching the market with a weekly standard deviation of 50% than 5%, at least in terms of stomach upset. 

Jeff Sasmore comments: 

WADR, Hussman's technical stuff may look good in a table.

The NAV on HSGFX hasn't moved since mid 2004. He missed almost all of the rally since he's so bearish. Must be related to a prominent editorialist at Barrons. If you want that sort of stability buy a CD. 

Rod Fitzsimmons Frey writes:

"Master" is a term widely used in the hobbyist craft market in order to tout videos, TV shows, etc. Person X is a master woodworker and so you should buy his four-hour DVD about sharpening plane irons.

In the 17th and 18th centuries a woodworker or joiner would serve a four to five year apprenticeship which was really more akin to slavery than employment. Then he could call himself a journeyman.

The next day if he had the money he could open a shop, employ an apprentice (slave), and call himself a master.

So in woodworking, the original sense of "master" was one who, largely due to protections built into the system, could employ slave labor. It has morphed into a marketing term used to hawk wares to people who often would like to substitute money for hard work and skill. I won't insult anybody by spelling out the investment-world parallels. 


Comments

Name

Email

Website

Speak your mind

Archives

Resources & Links

Search