January 31, 2004

The Joys of Speculation

CHRIS PELLERITO at Samizdata has penned an important essay on personal-finance journalism which would be worth your while to read. In his post, Mr Pellerito takes a look at the ever-popular but completely-useless brand of finance journalism which touts "hot" equities and promises to make the reader richer than the Aga Khan. He borrows a term -- financial pornography -- to describe such journalism.

Mr Pellerito writes that he hopes the term gains wider acceptance in the finance community, although we think that would be overdoing things. The finance community already has a word for such articles, and that is "crap." But let's take a closer look at Mr Pellerito's post:

----------

One of the biggest purveyors of financial pornography online is the MSN.com website, and this column doesn't disappoint: Seven Signs a Stock is Ready to Soar. The author purports to explain how to locate 'hot' stocks, those that are about to appreciate rapidly in price, by reviewing some research on what types of conditions most often preceded (notice I did not say caused, and neither did the research) a price increase.

It should not take a Wharton MBA to figure out what is wrong with the premise of the article. The cited research identifies the seven conditions that most often preceded a big run-up in the price of a particular stock, but nowhere does it suggest that these conditions were sufficient (or even necessary) to cause a stock price to take off.

Obviously, all of the conditions that make up the 'CANSLIM' acronym are desirable things for a corporation -- for its management and for its ownership. But that doesn't mean that the stock in question is about to outperform the market. I'm not a hard-and-fast believer in the semi-strong efficient market hypothesis -- I think a few super-stud investors can outperform the market -- but for the average investor reading MSN's Money Insight column, the CANSLIM approach is not going to turn those people into super-stud investors. EMH is still going to apply to those investors; there are just too many other investors who have the same type of information and insights at their fingertips.

---------

We are not personally fans of the "market-timing" approach to investing, although we will allow that a seasoned professional who spends rather a lot of time at it can succeed -- provided he is not whipsawed in a bad market, and provided he doesn't get in too late or out too early with his equities, and provided his trading commissions and tax bill don't eat up his gains, and provided he can do it over and over again over time. That's not to condemn market-timing, but rather to say that it takes a lot of work and requires an investor to be on top of his game.

We also note that on the page which Mr Pellerito cites, the following small module appears:

Using a screen designed to emulate each strategy, the AAII (American Association of Individual Investors) constructs a portfolio at the beginning of each calendar year and tracks each portfolioís performance for the year. Then it builds a new portfolio at the beginning of the following year.

For the years 1998 through 2003, AAIIís CANSLIM portfolios produced annual returns of 38%, 64%, 49%, -10%, 9% and 31%, respectively. CANSLIMís 332% cumulative six-year return ranked sixth of all strategies tested. Thatís not bad, considering that three of those years produced negative market returns, the worst of times for momentum strategies.

Let's leave CANSLIM aside for a moment -- we have no quibble with it, although as we said we're not market-timers -- and look at this test.

Quite frankly, this supposed test is built on some awfully stupid parameters. If an American investor turns over his portfolio on an annual basis, he is going to find that his massive gains are eaten away in tax -- short-term gains are taxed at the same rate as ordinary income. So in the first year, if the investor is in the 28 percent bracket, that 38 percent gain will really only be 27 percent. Do all the calculations -- for simplicity, assuming that the loss year has no tax implication -- and one finds the 332 percent return falls to roughly 284 percent, by our math.

Now, don't get us wrong. That's an incredible return for six years no matter how you slice it. But one can see how much money -- perfectly good, compoundable money -- was eaten by tax. And the trouble is that in the real world, such things aren't as cut-and-dried. They aren't even as cut-and-dried in the CANSLIM system, which the MSN writer finagles with to serve his own ends. We have personally found those seven principles useful, but in terms of applying them to individual equities, one must also do a lot of chart-reading and figure-analyzing; and in the end, guess right. To oversimplify things, as this article does, just isn't very helpful.

As for the efficient market hypothesis, we are not convinced that it holds: for some people can and do outperform the market on a regular basis; just as many people can and do underperform the market on a regular basis.

And the trouble with this hype-based financial journalism is that it can often lead to underperformance. One often sees writers tout Equity X or Company Z as a great investment; but an investor who buys shares because of that could find himself behind the curve from the get-go. Or, the writers go on about some supposed market trend or shift in some indicator, and it prompts unsophisticated investors to get all panicky: leading to lower gains or unnecessary losses.

These are not good things, as one former stockbroker might say.

Posted by Benjamin Kepple at January 31, 2004 01:51 PM | TrackBack
Comments

A very insightful analysis. With regard to the CANSLIM system's track record, I would also point out that many systems can work when back tested, for example the "super Bowl Indicator". But back testing really proves nothing since a certain percentage of all approaches will naturally fall outside the expected rate of return by a statistically significant percentage simply due to the normal distribution of outcomes. The Super Bowl Indicator has worked great when back tested, but I wouldn't want to stake my economic future to it as a method for selecting investments.

Posted by: Swammi in Solon at January 31, 2004 04:53 PM