4th Year a Charm? A Look at Oversimplified Predictions

There’s kind of an odd article in today’s NYT that I wanted to look at: The column "Fundamentally" by Paul Lim usually looks at, well, fundamentals. So I found it a bit odd when today’s run ventured into statistical and quantitative analysis.

Here’s the premise:

"Bull markets often follow a predictable pattern. In the first year of a rally, bulls tend to charge out of the gate: the Standard & Poor’s 500-stock index has posted rip-snorting price gains of 38 percent, on average, in the initial year of bull markets since 1942, according to a recent study by S.& P.

This is typically followed by a more subdued second year, with the S.& P. 500 up around 12 percent, on average. And in the third year, the rally starts to sputter, with average gains of just 3 percent.

This is close to the way the last three years have unfolded – with the S.& P. up more than 26 percent in 2003, 9 percent in 2004 and 3 percent last year."

While there’s nothing in the above that is false, it is a good example of single variable analysis.  The author takes one variable — the history of 4 year bull markets — and extrapolates a pattern from that.

As we noted in Single vs. Multiple Variable Analysis in Market Forecasts, because of the complexity of markets, economy, and investing, the more confirming variables you can add to a model, the higher the liklihood you will be able to reach a valuable deduction.

I never rely on a single element; in fact, if these pages have taught you anything, its that there is no one magic bullet, no single variable, that will determine how the market behaves. (Of course, this truism has not stopped people for looking for a McGuffin).

Further complicating things — each and every element takes place within a context of multiple variables. Interest rates, earnings, inflation, prior comparables, trend, valuation, commodity prices, recent market performance, internals, sentiment, money supply, deficits — these are just some of the variables that change the context of other specific elements.

Compare that kind of analysis with "Earnings are good, so I like stocks" or  in the present case, "Stocks do well in the 4th year of a Bull market."

Rather than relying solely on the historical observation "4th years generally do well," I would introduce the other variables to see how likely a positive outcome is for the coming year. Consider the following elements:

This is 2nd year of a Presidential term;

• The S&P has been positive for 3 consecutive years;

• Is this a Secular or Cyclical market? Are we in a long term trading range, or is there a well defined trend?

• How are stocks valued — are they cheap or expensive?

• Is sentiment excessively Bullish or Bearish?

• How recently the market suffered a 10% correction;

• Are P/Es expanding or contracting? Are we mean reverting, or overshooting?

On a year-over-year basis, how are S&P500 earnings ? Are they terrible, and getting better, or great and getting worse?

• How is the Yield Curve sloped:  Positively, Flat or Inverted?

I can keep going, but you get the idea.

Any single variable will give you an easy prediction, a goiod bumper sticker, but have a low probability of a correlated predictive outcome. The more variables you introduce — up to a point — the more likely your predictive outcome will be. I freely admit to not knowing precisely where the line of introducing too much complexity is — but its likely much higher than 4 or 5 variables.

click for larger table


courtesy of NYT


Bottom line: Avoid the oversimplified single variable predictions. They are money losers . . .


Is the Fourth Year a Charm for the Bull Market?


NYT, January 22, 2006

Print Friendly, PDF & Email

What's been said:

Discussions found on the web:
  1. Hans Suter commented on Jan 22

    “single variable predictions …. are money losers . . . “. Short it ;-) ?

  2. Joe Abraham commented on Jan 22

    Nice web site.

    How in the world do you find time to do it???

  3. Lord commented on Jan 22

    Single variables are often an artifact of data mining.

  4. The Stalwart commented on Jan 25

    Ritholtz on Meaningless Patterns

    I’ve been meaning to link to this Barry Ritholtz post on why it’s meaningless to say something like The fourth year of abull market usually does ___. It touches on all the the themes we like, like false patterns, and

Posted Under