|
The Truth About Economic Sentiment Indicators
At any given time, the theoretical value for any stock is simply the present discounted value of its future dividends. As the firm makes money, theoretically, the profits will be paid out to investors.
The investor's job is obviously to determine which companies are going to make a profit and capitalize by investing before the profit is known.
As economic data is made available, investors analyze the data and determine future profit scenarios.
To a macroeconomist, this is generally about as technical as our profession gets when analyzing investor behavior. We tend to focus on the macroeconomic events that drive the market.
What makes no sense to us is why the market reacts so strongly to sentiment indicators.
These variables tend to have no influence on consumption numbers yet investors use them as strong indicators on where the economy is headed. Some the usage of the sentiment indicators can be traced back to media talking heads that use them in their "forecast analyses." But many smart and influential researchers, including Christina Romer, the Chair of the Council of Economic Advisors, have stated that the economic recovery is pinned to the increase in consumer sentiment.
While it's hard to dismiss sentiment indicators as complete fluff when they often push the market on the day they are released, it's important to maintain a sense of understanding that a long position should not be developed based on a better-than-expected sentiment reading.
Consumer Sentiment
The two main consumer sentiment indicators are the University of Michigan Consumer Sentiment Index and the Conference Board Consumer Confidence Indicator. Both data points strive to explain the same thing -- consumer behavior.
Consumer sentiment is extremely volatile. The above graph takes a three-month moving average of the monthly change in each index from January 1979 to August 2009.
One of the biggest problems with using a sentiment indicator is knowing which indicator is the correct assessment of consumer behavior. During the 377 months that are evaluated in the graph, 115 months showed the indicators moving in the opposite direction. How can you trust what the consumer believes when 31% of the time the surveys report conflicting information?
In reality, the consumer sentiment indicators are just a snapshot of the consumer's mood. As you know, advertising can greatly affect what someone is going to buy. Likewise, the way the media presents the news plays an important role in determining consumer sentiment. The above graph presents the two sentiment indicators along with a reading of the number of news events that used the word "recession" in the headline, according to Google.
The sentiment indicators started trending down at the same time that "recession" entered the lexicon toward the middle of 2007. As the Google trend line spiked, the sentiment index declined at its most rapid rate. Likewise, when the word "recession" started being used less frequently in the media, the sentiment indicator rebounded. We can see a similar relationship with the sentiment indicators if we used "economic recovery" or "green shoots" as the search word.
We see a similar relationship when the sentiment indicator is compared to gasoline prices.
In general, sentiment indicators tend to move on media reports, gasoline prices, unemployment, and stock market returns.
Consumer Sentiment and Consumption
The purpose of the sentiment indicators are to explain consumer behavior. It is often implied that consumer sentiment and consumption should go hand-in-hand. Therefore, an increase in consumer sentiment should indicate an increase in consumer consumption.
Unfortunately, consumption and sentiment do not have much in common.
Using a simple statistical regression, the above graph predicts real personal consumption expenditures based upon the consumer sentiment index. The obvious result is that neither sentiment indicator does a good job with predicting real consumption.
It's possible that the reason why neither indicator can predict the total consumption is because the index is highly volatile based upon consumer moods. Instead, it might be best used to predict the change in consumption expenditures. So, if the consumer feels better in this month they will increase their expenditures.
Using the same statistical method, the change in consumption cannot be predicted by the change in sentiment.
Consumption cannot be predicted by just using the sentiment indicator.
What Does Predict Consumption?
The best indicator for predicting consumption will always be income.
As you can see, the prediction for consumption by using real personal disposable income follows very closely with the actual real PCE data.
The relationship should be solid. There are only two options a consumer can do with their income: save or spend.
The difference between the prediction and the actual real PCE is just the difference between the actual savings rate and the median savings rate. When the savings rate jumped in 2008 due to the start of the consumer deleveraging process, the predicted real PCE value overshot the actual.
What It All Means
The media tends to overinflate the usefulness of the consumer sentiment indices, and as a result, the market tends to move sharply on days when the data is released.
To economists this is very odd. The sentiment indices are simply an index that shows how the consumer is reacting to media reports, gasoline prices, unemployment, and stock prices.
For those that are going long on the market, the indices have no statistical relationship with predicting consumption. You should not use a positive movement in sentiment as an indicator that the consumer is beginning to rebound!
If you are looking to predict consumption, the most effective variable to look at is income.
Feeds!