Sentiment Supports the Borrowing of 2017 Gains Argument

Support for the concept of the post-election ("Trump") rally borrowing gains from 2017 can be found in our Sentiment Survey. Optimism was both unusually low (a bullish sign) and unusually high (a bearish sign) in recent weeks. Though seemingly contradictory, the two signals may not be when the post-election rally is taken into account.

I’ll start with the very large jump in optimism that occurred last month. Bullish sentiment surged by 26.3 percentage points between November 2 and November 23. This was the 13th biggest three-week increase in the survey’s 29-year history. It was also the largest three-week increase since 2010. (Bullish sentiment rose by 30.2 percentage points from August 26 to September 16, 2010. For those of you who are curious, the largest three-week increases were 40-point moves in August 1987 and July 2000.)

The most obvious question is: What does the move say about market direction? The average six-month gain for the S&P 500 following the 12 bigger upward moves in bullish sentiment was 0.3%. The large-cap index rose six times and fell six times during those periods. In comparison, the S&P 500 has realized an average six-month gain of 4.3% throughout our survey’s history.

Should the S&P 500 underperform over the next six months, it would reinforce the contrarian nature of unusually high bullish sentiment readings. Whenever optimism has exceeded its typical range, the S&P 500 has realized an average six-month return of 2.7%. Last week’s bullish sentiment reading of 49.9% was above the typical range of 28.0% to 48.7%. (Statistically, we define unusually high readings as being more than one standard deviation above average, and low readings as being more than one standard deviation below average.)

Underperformance is not the same thing as a loss. The S&P 500 has gone on to realize a six-month gain following nearly three out of four of all unusually high bullish sentiment readings. The 27.5% chance of a loss is close to the 26.1% for all six-month periods throughout the survey’s history. So, if history repeats, returns will be lower, but positive. Keep the phrase “if history repeats” in mind when thinking about that last statement. The future can unfold in ways we don’t we expect it to.

If we were to go back to the start of November, a different contrarian signal was being flashed: Bullish sentiment was at 23.6% on November 2, 2016. On average, the S&P 500 has realized a 7.2% six-month gain following an unusually low bullish sentiment reading. Currently, as stated above, any reading below 28.0% is unusually low. We’ve seen optimism come in at unusually low levels many times this year—23 different weeks, to be exact.

The contradictory signals could be reconciled if one were to believe the post-election rally is borrowing from next year’s gains. Between the election and November 23 (the end of last week’s survey period), the S&P 500 rose by 3.0%. Add in an additional 2.7% gain following unusually high readings (such as what occurred last week) and the cumulative return would be both above average for the six-month period starting on November 2 and below average for the six-month period starting on November 23.

This is an explanation based on historical data. It demonstrates how two contrarian signals can both be valid even though they seem contradictory at first glance. It is not a forecast of what will happen even though it lends support to the concept of Mr. Market having borrowed from his 2017 gains. Unusually low or high bullish sentiment readings do not cause stocks to outperform or underperform. Furthermore, the new administration, monetary policy, economic growth, corporate earnings and valuations will all influence sentiment (both individual and institutional) and the direction of stock prices. The S&P 500’s return over the next six months could very well be different than what our survey’s historical data suggests.

For those of you who are interested, the table below contains an updated analysis of the AAII Sentiment Survey. The numbers are calculated without hindsight. Rather, I used the historical data that would have been available on a given week in the past to determine whether bullish, neutral or bearish sentiment was unusually high or low.

When crunching the numbers, I observed a shift in how unusually high levels of neutral sentiment should be viewed. It’s no longer a clear contrarian signal. The S&P 500 has gained 4.3% following unusually high levels of neutral sentiment. This is down from the 7.2% average gain calculated in May 2015. The difference is due to a larger sample size: 111 unusually high readings now versus 71 in 2015. The sample size is still comparatively small relative to the data we have for bullish and bearish sentiment, however.

Table 1. Performance of the AAII Sentiment Survey without Hindsight