CREATING YOUR OWN SENTIMENT INDICATOR
NOTE: This article was originally published in the October 2002 issue of Active Trader magazine.
Sentiment is the level of investor fear or greed in the market at any given time, whether it is on a short-, intermediate- or long-term time frame. In practical terms, it is important to be aware of the prevailing market sentiment one time frame higher than the one on which you are trading. For example, if you are trading based on 60-minute price bars, you would want to know if the daily trend is poised to change because of a sentiment extreme. If it is, you would need to adjust your shorter-term tactics accordingly.
A good way to gauge the intermediate- to long-term market mood is to create a composite index derived from the four most popular sentiment surveys, which come from Investor’s Intelligence, Consensus Inc., Market Vane and the American Association of Individual Investors. These reports are available weekly in Barron’s.
The Investor’s Intelligence survey is derived from 130 investment newsletters spanning a wide range of investment methodologies. The survey determines whether these newsletters are bullish, bearish or neutral. The results are expressed as a percentage, so a typical weekly reading would be 44.5% bulls, 35.0% bears and 20.5% correction (those who are basically bullish but expecting a correction – in other words, neutral). The examples in this article ignore the neutral readings.
The Consensus Inc. survey draws upon an extensive mix of brokerage analysts and independent market advisory services. Like the Investor’s Intelligence survey, the Consensus Inc. data covers a broad spectrum of market approaches, including fundamental, technical and cyclical. Consensus Inc. then compiles their data and reports the percentage of their contacts who are bullish (e.g. 30%).
The Market Vane survey monitors the trading recommendations of Commodity Trading Advisors (CTAs), who are registered professional futures traders. Similar to Consensus Inc., they also report a single number representing those who are deemed to be bullish.
Finally, the American Association of Individual Investors takes surveys of their membership, consisting mostly of retail investors and traders. Those surveys are broken down into those who report that they are bullish, bearish or neutral, similar to the Investor’s Intelligence survey.
The first three surveys represents different segments of the professional investment industry; the last survey more accurately reflects public market sentiment. Rather than relying on just one of these sentiment surveys, consulting an indicator that incorporates all of them can provide a more accurate picture of overall market sentiment. The following technique, which creates an index of the four surveys, can be done in just a few minutes each week (after you have set up the initial calculations).
One way to get a feel for market sentiment is to simply look at the degree of each survey’s bullishness and bearishness. A better way to determine current market sentiment is to calculate a stochastic indicator based on the bullish percentage of the surveys. In other words, you can create a stochastic oscillator that uses market sentiment data instead of price data. For those surveys that provide both bullish and bearish percentages, use the bullish ratio (bulls/(bulls + bears)).
The stochastic calculation compares the current reading to the range of readings over a given time period. It can be calculated on any time frame; in this case, we will use a one-year (52-week) interval. The basic stochastic formula is:
(CURRENT – MIN) / (MAX – MIN)
Where:
CURRENT = The most current reading
MIN = The lowest (minimum) reading over the lookback period (e.g., one year)
MAX = The highest (maximum) reading over the lookback period
This means we will subtract a survey's lowest bullish percentage of the past 52 weeks from the most recent bullish percentage, and divide the result by the difference between the highest and lowest bullish percentages of the past 52 weeks.
Stochastic readings range between zero (when the most recent reading is the same as the lowest reading of the lookback period) and 100 (when the most recent reading is the same as the highest reading of the lookback period).
The chart below shows what a spreadsheet might look like if you were computing a stochastic indicator for the Consensus Inc. sentiment data, with the bullish percentage in column C. You would perform the same calculation for the other three sentiment surveys.
To create the Composite Sentiment Index, calculate a straight average of the four stochastic readings. Like the individual stochastic calculations, the resulting indicator oscillates between 0 and 100. Exceptionally low readings mean the composite bullish percentage (or bullish ratio) of the four surveys is low – that is, pessimism is excessive. On the other hand, if the stochastic reading is high, investors are extremely optimistic, which means you should be on the lookout for a possible market downturn.
Now let’s look at how this Composite Sentiment Index could have helped alert you to trend changes and kept you in step with the market over the past few years.
Although investors are often “correct” in the middle of a trend, the herd tends to become too bullish or bearish when a trend extends itself up or down, respectively. However, it is advantageous, especially in trading ranges, to be bullish on the lower end and bearish on the upper end of price moves – something that goes against every fiber of most traders’ beings. After all, haven’t we been taught the best traders buy strength and sell weakness?
That can be true - but only when sentiment is not at an extreme. The chart below shows the Composite Sentiment Index with the S&P 500 index from July 1999 to June 2002. A general rule is that when the index reaches 60, sentiment is becoming quite optimistic. In these conditions, you would want to – at the first sign of price weakness – tighten stop-loss orders on long positions, or even increase shorting activity. Conversely, when the index drops to 40 or lower, sentiment may be too pessimistic and the market might be near a market bottom.
Although this indicator is not perfect (notice the premature buy signals in late 2000 and early 2001), it captured the majority of intermediate and longer S&P turning points over the three-year period.
Despite the arrows indicating buy and sell “signals” in the chart below, you should not rely on the Composite Sentiment Index alone to place trades. A more appropriate approach is to watch the index to know when the market is reaching a sentiment extreme. In such situations you must be extremely cautious about taking trades in the direction of the existing trend because the odds of a countertrend move are high.
There are other ways to track investor and trader sentiment, but this approach is easy to calculate and comprehensive (because it incorporates multiple sentiment measures from different areas of the market). By taking a few minutes each week to update your sentiment data, you should be better able to stay on the right side of the market.
You can see this indicator updated weekly, with in-depth analysis, at www.sentimentrader.com.
- Jason Goepfert
September, 2002