Be wary of bellwethers’ abilities to predict the future

Bellwethers are predictors. Ohio is a bellwether that predicts who wins a presidential election. Paris and Brooklyn have had their days as fashion bellwethers. Some people use a groundhog in Pennsylvania as a bellwether of more winter, and others have used skirt length as a bellwether of the stock market.

Investors and policymakers use companies as bellwethers. For example, investors paid a lot of attention to Amazon’s quarterly report on Thursday because its sales numbers are a bellwether for e-commerce sales.

{mosads}Analysts use Walmart’s sales numbers as a bellwether for retail spending. Policymakers in turn use retail spending as a bellwether for the strength of the economy.

Investors use more than a company’s sales numbers. They also use a company’s stock price as a bellwether for the stock prices of other companies in the same industry and even for the stock market as a whole.

When Intel gave its quarterly report on Thursday, analysts watched its stock price as a bellwether for the stock prices of other companies in the chip industry. When Caterpillar and 3M gave their reports, analysts watched their stock prices as bellwethers for the stock market as a whole.

All these bellwethers are predictors, but they are imperfect predictors. How can an analyst or policymaker decide how much weight to give bellwethers like Walmart or Caterpillar?

A good place to start is to ask why a bellwether would be a good predictor. What is the logical basis for the bellwether to be good at predicting what you want to predict?

One basis is size. Amazon is big in relation to e-commerce sales: It accounts for about 50 percent of e-commerce sales. If a company accounts for 50 percent of what you are trying to predict, the company will be a good predictor.

Walmart also is big. It sells more than the next three largest U.S. retailers combined, and it is not just big; it also sells across a lot of retail categories, from clothing to toys to cosmetics to food. That makes Walmart more representative of retail sales as a whole than a more focused retailer like Walgreens or Kroger.

A bellwether that is more representative of what you want to predict is more likely to be useful than a bellwether that is less representative.

How useful is it for investors or policymakers to use companies like 3M and Caterpillar as bellwethers not just for an industry, but for the economy as a whole? 3M is large — a Fortune 100 company — and it produces a wide variety of products that serve a wide variety of industries.

It’s not just Scotch Tape and Post-it notes. It sells chemicals, orthodontic products, medical devices, power storage equipment and more. It is a widely-diversified industrial and consumer products company.

Because 3M is large and sells across a wide swath of the economy, you would expect its sales to be correlated with the economy as a whole. It could be a useful bellwether for the economy (and also for other companies whose results are closely linked to the state of the economy).

Caterpillar is another Fortune 100 company, but it is more focused than 3M. It is a heavy industrial company that makes construction and mining equipment, industrial turbines and railroad locomotives. Exports represent 59 percent of its sales.

Its focus may make it a better bellwether for the construction, extraction and infrastructure industries than for the economy as a whole. Its emphasis on exports might make it a useful bellwether for the effects of the trade measures on U.S. exports and exporting companies.

How useful is it for investors to use 3M’s and Caterpillar’ stock prices as bellwethers for the large-cap stock market as a whole? One way to assess their usefulness is to examine how price changes in the two stocks compare with price changes in the S&P 500 index of large-cap stocks.

As of Tuesday, the S&P was up 3.87 percent over the prior 12 months. The price of Caterpillar’s stock was down 15.5 percent and 3M’s stock was down 22.17 percent.

The price moves of the companies’ stock were not good predictors of the moves of the S&P. (A formal, statistical measure of the association between the stock prices of 3M and Caterpillar on the one hand and the S&P on the other, R^2, also is not very high.)

3M and Caterpillar may be useful bellwethers for the health of particular industries or the economy, but less useful for predicting stock prices in the large-cap market as a whole.

That means we should be careful when we use bellwethers. A bellwether could be good at predicting one thing but not another. Don’t bet much on the Pennsylvania groundhog. Don’t overreact to corporate earnings reports or stock price moves.

Whether you are an investor or policymaker, be careful about using what a pundit calls a bellwether unless you understand how well the bellwether actually predicts what you want to predict.

Erik Gordon is a clinical assistant professor at the University of Michigan’s Ross School of Business.