HOW MUCH SHOULD YOU RELY ON SENTIMENT SURVEYS?

By Charles Rotblut, CFA & Vice President with AAII

The AAII Sentiment Survey was the subject of a Wall Street Journal article this morning. The article discussed the survey's popularity, its limitations (we do not conduct a random sampling) and the correlations between the survey and market turning points.

One of the statistics highlighted in the article was the spread between bullish and bearish sentiment (aka, the bull-bear spread). Research firm Birinyi Associates calculates that whenever the bull-bear spread exceeds +30 points, the S&P 500 has tended to fall over the next six months.

This week's survey showed the bull-bear spread at +30.5 percentage points.

Though this sounds bearish, do not hit the sell button. In fact, you should never buy or sell stocks based on any single sentiment indicator, whether it is our survey, the State Street Investor Confidence Index or the Investors Intelligence Advisors Survey. You also should not base your buy and sell decisions solely on one fundamental or technical indicator. Rather, look at a broad array of factors, including your portfolio allocation. Ideally, you want several indicators to provide the same signal.

For example, consider shifts in the economic data, prevailing valuations, business conditions and revisions to earnings estimates. You can overlay this with technical analysis and the results from investor sentiment surveys. Alternatively, you can use charts and sentiment surveys as an impetus to conduct further research. Charts and surveys can often tip you off to something you may have overlooked in the economic and fundamental data.

This is particularly the case when using our survey to gauge market direction based on the level of optimism and pessimism. Extraordinarily high bullish and bearish readings in our sentiment survey have been correlated with market reversals. (Extraordinarily high bullish readings have preceded market corrections and extraordinarily high bearish readings have preceded market rallies.)

The proximity of bullish and bearish sentiment to their historical averages is one of the things I pay attention to in our survey results, in addition to the bull-bear spread. Specifically, I look to see whether bullish and bearish sentiment are within or beyond one standard deviation of their averages. A reading within one standard deviation means the reading is well within the norm we have seen over the life of a survey. A reading beyond one standard deviation means the reading is unusually high or low. A reading beyond two standard deviations is extraordinary and suggests that the pendulum has swung too far in favor of optimism or pessimism.

Putting some numbers to this narrative, bullish sentiment has averaged 39% and bearish sentiment has averaged 30% since 1987. The standard deviations are 11 percentage points for bullish sentiment and 10 percentage points for bearish sentiment. Thus, a reading above 50% for bullish sentiment would be one standard deviation above average and a reading of 41% for bearish sentiment would be one standard deviation above the average. A statistic can also be one or more standard deviations below average. Readings of 28% for bullish sentiment and 20% for bearish sentiment would be one standard deviation below average. (The numbers are rounded. A spreadsheet with the historical data for the survey and the standard deviations can be downloaded from our Investor Sentiment Survey page.)

Just keep in mind that an optimistic or pessimistic attitude expressed in a survey does not necessarily equate with action. An investor can say he is pessimistic about the short-term outlook for stocks, but not actually be selling any of his current holdings. This is why I suggest using our sentiment survey as just one part of your market analysis, as opposed to making it the cornerstone. Rather, look at a variety of indicators.

Finally, consider your long-term financial goals. Sentiment ebbs and flows; your strategy for building and maintaining long-term wealth should not.

We republished an article how the AAII Sentiment Survey can be used to determine when a market reversal may be forthcoming. Though written in 2004, the article's emphasis on paying attention to extreme readings in bullish and bearish sentiment continues to hold true.

THE WEEK AHEAD

Five S&P 500 members will report earnings next week. Best Buy (BBY) is scheduled for Tuesday. Discover Financial Services (DFS), FedEx Corporation (FDX), General Mills (GIS) and Oracle (ORCL) are scheduled for Thursday.

The week's first set of economic data will be released on Tuesday, when the November Producer Price Index (PPI), November retail sales and October business inventories are scheduled. Wednesday features the November Consumer Price Index (CPI), November industrial production and capacity utilization, the December Empire State Manufacturing Survey, and the December National Association of Home Builders (NAHB) housing index. November housing starts and building permits, the December Philadelphia Fed manufacturing survey and third-quarter Current Account data will be published on Thursday. Friday's sole report will be the Conference Board's November Leading Indicators Index.

The Federal Reserve Open Market Committee will hold its last meeting of the year on Tuesday. This will be a one-day meeting and, besides a possible update on asset purchases, no change in Fed policy is expected.

Friday will be a quadruple-witching day, which means both monthly stock options contracts and quarterly futures contracts will expire.

Charles Rotblut, CFA is a Vice President with AAII and editor of the AAII Journal.

Post Footer automatically generated by Add Post Footer Plugin for wordpress.