One of the best definitions of a bear market is that all news is bearish. The news of a large drop in February's consumer confidence helped send the markets into a tizzy -- not that the usual spate of profit warnings and bizarre rumblings out of Washington helped matters.

Well, as they say around these parts, money talks and you-know-what walks. There's a major difference between attitude and behavior. Attitude is what you think, while behavior is what you do. Infants observe what their parents do far more than what they say, and this is a good habit to retain. Confidence surveys, no matter how well constructed -- and both the University of Michigan and Conference Board polls are crafted carefully and consistently -- cannot measure behavior.

As we struggle with the multibillion-dollar question of whether or not we are in a recession, let's ask whether confidence should be part of the equation, and more important, does it have anything to do with the course of financial markets? And, while we're at it, let's ask whether growing retail sales are important for financial markets.

Confidence: A Reliable Indicator?

No product peers more deeply and honestly into the soul than the cash register. This suggests that both the level and the growth trend of retail sales, which the Census Bureau reports each month, should be key behavioral yardsticks. As this series is quite noisy even after seasonal adjustment, it's necessary to smooth it. The method chosen here is the percentage change in sales for the current three-month period over those for the current 12-month period. The data are extended back into the late 1970s to incorporate the previous period of economic malaise.

While the two measures don't appear to have a close relationship, two periods stand out and are highlighted. The first is the sharp collapse in both confidence and retail sales in early 1980; these underscored general dissatisfaction with present affairs and led to Reagan's election in the fall. While retail sales rocketed higher in 1981, neither confidence nor the stock market did much. In fact, confidence didn't start to move higher until the stock-market breakout in August 1982.

The second period highlighted is November 1994, which marked the conclusion of a terrible year for both stocks and bonds. The Federal Reserve had embarked on a series of interest-rate hikes in February of that year. Retail sales growth slowed significantly and stayed slow all the way until September 1998. Consumer confidence, however, stabilized and then started to jump higher in the summer of 1996. It hit its high in February 2000, just before the stock market hit its own peak. Retail sales growth peaked simultaneously and dropped both earlier and more rapidly than did consumer confidence.

If You Buy It, They Will Smile

In both the early-1980s and mid-1990s, sustained growth in confidence appeared to follow, not lead, a stock-market rally. Because financial markets are allegedly discounting mechanisms for investors' expectations, let's switch to a second survey measure, consumer expectations, and see how well they correlate with the growth rate in the S&P 500. The same smoothing procedure is used for stock prices as for retail sales.

The relationship here is reasonably tight, but is slightly counterintuitive. The S&P 500 appears to lead consumer expectations far more than the other way around. It looks more like investors are saying, "My portfolio's getting fat, and the country's on the right track," than "Things are going great, and I'm going to buy some stocks." A cumulative wealth effect appears to be operating as well: The big percentage gains in the S&P 500 peaked in mid-1997, but as portfolios continued to swell in absolute dollar terms, expectations grew as well and did not peak until February 2000.

Wall Street's Having a Sale

Stocks, as we all are aware, are much easier to sell when prices are rising. Economists may say this is because they are a "superior good," like perfume or Scotch whisky. It's time to answer the question of whether retail sales are related to stock prices.

The answer here seems to be "not since the last time we had a recession." The two growth rates matched each other closely during the early 1980s, but then diverged until the parallel collapse in late 2000. In other words, a recession is bad for both stocks and the economy, but a strong stock market by itself does not seem to produce a spending boom in the real economy. So much for that vaunted wealth effect: We stop spending when our portfolios shrink -- far more rapidly than we increase spending on the way up.