November 6, 2002

The headlines screamed that consumer confidence fell to the lowest level in 9 years.  If I am making corporate investment decisions, should I be cautious?  If I am a loan officer at a bank, should I tighten my criteria?  If I am a retailer depending upon Christmas sales to meet my payments, should I be worried?

To my knowledge, the first consumer confidence indices were developed at the University of Michigan to help the auto industry improve their forecasting for the next model year.  The presumption was that if consumers are nervous, they will postpone the replacement of their vehicles.

Although the Michigan forecasting model still uses confidence in their economic forecasts for auto sales, I understand that substantial adjustments are made to that relationship.  In other words, confidence is not that reliable a forecasting tool.  

Indeed, the headline alone should warn people that confidence measures do not always predict future economic events.    Nine years ago was the early stages of the past economic expansion.  If we get the same future that occurred after that low confidence of almost a decade ago, we should be satisfied. 

Having indicated that confidence is not that reliable a forecasting tool, I must acknowledge that it has its purposes. 

First, consumer confidence is one of the leading indicators.  If other indicators, such as interest rate spreads, the workweek, building permits, and orders are moving in the same direction, then its predictive powers are substantially enhanced. 

Moreover, the current drop almost certainly means the leading indicators will decline for their fifth consecutive month.  Only once has those indicators fallen so long without a recession following close behind. 

Second, confidence is closely aligned to stock values.  In an investment model I was developing some years ago, I wanted to use stock values as an explanatory variable.  Because stock values are hard to predict, I substituted confidence.  To my surprise, stock prices and consumer confidence were almost completely inter-changeable. 

The relationship clearly has changed, but this means that confidence and stock prices are very closely aligned.  Furthermore, stock values and investment decisions are strongly related (not only statistically but also by economic theory).

Before running out to sell your stock portfolio, let me also mention two other points.  Confidence and stock prices are interchangeable, but one does not predict the other.  They change together. 

Second, once all the other relevant variables that anticipate changes in an economy are put into a sound model, confidence adds no further value.  In other words, confidence reflects conditions.  It does not anticipate them. 

As a model builder, I also learned that I could predict confidence using other reliable forecasting variables  better than I could use confidence to forecast conditions. 

So, where does all this lead?  Confidence reflects economic conditions, but needs other conforming variables to predict future conditions.  When confidence is low, the consumer needs enticements to spend.  However, as auto sales firmed with zero financing even as confidence fell in October, we learned that those other enticements could overcome weak confidence. 

Now to an answer for those three questions at the top of this column.  Retailers should be worried about this Christmas.  Economic conditions are making consumers nervous, and that cannot be good for business.

Use promotions and other enticements early and often. 

If I am making investment decisions, I already am nervous about stock prices.  If investors are not pricing my existing capital at a premium, why should I add more capital.  (Lower costs and increased quality are two very good answers).  As stock prices already are providing enough information for my investment decisions, I do not need to double count by also including consumer confidence. 

If I am a loan officer, consumers already are opting out of spending and my loan committee has stock performances to guide them.  Indeed, the consumer reluctance to take on debt (also related to consumer confidence) may require being more aggressive to build the loan portfolio that is most profitable for the bank.  Under no circumstances should I use the drop in confidence to raise my lending hurdles. 

And if I were a central bank about to decide what is needed to guide this economy to maximum long term growth without inflation, such as the Federal Reserve, I would let the confidence information tell me that current conditions are not consistent with that goal.  Something needs to change.   



mbar.jpg (9380 bytes)