Predicting the Stock Market: Contrarianism and the AA II Investor Sentiment Survey

And an Income Orientated US Exchange Traded Fund

Quartz crystal ball image courtesy of Wikipedia

Stock market pundits are in as much demand as astrologers and soothsayers were before the Enlightenment. Last week's post on this blog included a comment from Howard Marks that "There's only one way to describe most investors: trend followers. Superior investors are the exact opposite." Marks is a contrarian. To be contrarian, we need to know what other investors are thinking.

The American Association of Individual Investors (AA II) surveys its members every week to ask them if they are bullish, neutral or bearish on the US stock markets. The results of the AA II Survey are as close as we can get to a consensus opinion of other investors on the future movement of the US stock market. Over the long term, AA II Bulls exceed Bears by 8.5 percentage points. What makes the survey interesting is that AA II members nearly always get it wrong.

Last week's survey revealed a very high bearish sentiment. Bears exceeded Bulls by 35.2%. This is the most negative sentiment since July 2010, when Bears outnumbered Bulls by 36%. Then the S&P 500 recorded a weekly low of 1062 points, which was followed by an uninterrupted bull market. In March 2009, Bears outnumbered Bulls by 51%. March 2009 also coincided with the lowest point in the S&P 500, at 696 points, since October 1996.

As the UK stock market closely follows the US stock market (see the graph below), an American sentiment survey is almost as valid for a contrarian here as there.

(Graph courtesy of Yahoo, click to enlarge)

Can we use the AA II indicator to predict future trends in the stock market, provided that we do exactly the opposite? In January 2013, the CXO Advisory Group, a consultancy, published a study comparing the AA II investor sentiment survey to movements on the S&P 500.  By comparing the net investor sentiment (Bulls less Bears) to prior and subsequent movements in the stock market, they found the following correlations.

In a layman's terms, the above correlations show:

1. AA II sentiment is largely driven by what has happened in the past, especially what happened 4 weeks ago.

2. As anticipated, the stock market has tended to move, looking forward, in the opposite direction of AA II investor sentiment. Bearish sentiment tends to be followed by bullish markets and vice-versa, but the relationship is very, very weak, negligible one week ahead and still tiny 6 months ahead.

3. The study also finds that, since the beginning of the survey in 1987, AA II members have 'learnt' from their errors. They are less prone to error in their sentiment. And the value of their opinion, for a contrarian, is not worth as much as it used to be.

4. However, the study found that very bearish sentiment, when the net sentiment is 16% more bears than bull, is followed, especially after 6 months, by a more bullish market. But even the most extreme Bullish decile does not indicate shorting.

 See the table below (click to enlarge).

The study found that the 6-month return after the 1% most bearish net sentiment readings (14 observations < -35% in 25 years) is about 16%. The study concludes: “evidence indicates that investors may be able to exploit extreme values of AAII net investor sentiment [Bears outnumber Bulls by more than 35%] as contrarian signals, but reliable (extremely bearish sentiment) signals are rare.”

It happens that the latest AA II survey has thrown up exactly one of these rare moments, with a Bear minus Bull reading of 35.2%. Should we jump into equities?

Skeptical investors will note that:

1. The S&P 500 is only 3% off its 90-year high of 1597. This is not a repeat of March 2009, when the S&P 500 was 44% off its all-time high.

2. Statistically, the number of readings when this contrarian indicator (most extreme Bearish sentiment) has worked is very small (only 14 instances to date).

3. The AA II survey, as a contrarian indicator, is useless for predicting bear markets.

4. AA II members are merely reacting to recent past information.

5. No one can or has ever been able to consistently predict future events. Why should they be able to predict stock market movements?

------------------------------------------------------------------------------------------------------------------------

An Income Orientated US Exchange Traded Fund: SPDR S&P US Dividend Aristocrats ETF

The easiest and cheapest way for a UK investor to invest in the US stock market is by way of an Exchange Traded Fund (ETF). The Investors Chronicle (1 November 2012) recommends HSBC S&P 500 ETF (HSPX), which has the lowest cost (TER) of all such funds at 0.09% p.a.

An alternative ETF, the SPDR S&P US Dividend Aristocrats ETF (USDV), offers the investor a narrower selection of stocks based on the S&P High Yield Dividend Aristocrats Index. This is comprised of the 84 stocks of the S&P Composite 1500 Index that have increased dividends every year for at least 20 consecutive years. According to State Street, the USDV fund manager, "these stocks have both capital growth and dividend income characteristics, as opposed to stocks that are pure yield, or pure capital oriented." USDV yields 2.95% and stands on a collective PE ratio of 16. This compares to HSPX's yield of 1.75% and a collective PE ratio of 18.

As USDV was launched in October 2011, its track record is rather short. To date, the ETF has outperformed the S&P 500. (Graph courtesy of Yahoo, click to enlarge)

 

USDV is domiciled in Ireland, which does not impose any withholding taxes, and the fund has distributor status, ensuring that all gains are taxed at the rate for UK Capital Gains and not as UK Income Tax. The ETF is physically backed by securities and, capitalised at over $1 billion, it is both liquid and large enough to reap the benefits of scale. It has an annual cost (TER) of 0.35%.

The wise investor will note that:

1. As USDV is only invested in companies that have a 20-year dividend record, this excludes a very large number of newer technology stocks. Only 3.7% of the funds are in the IT sector compared to 18.2% of the S&P 500. Consumer staples, industrials and utilities are all disproportionally represented in the USDV portfolio.

2. USDV digs deeper than the S&P 500 into smaller cap stocks that are included in the S&P 1500. The minimum market cap for the S&P 500 is $4.5 billion, whereas USDV's cut-off is at $2 billion.

3. USDV is weighted by the dividend yield of each component (instead of the customary market capitalization) up to a maximum of 4% of the total fund. Hence the largest fund component is Pitney Bowes, famous for its postal franking machines, with a market capitalization of only $2.9 billion. Its number one spot is due to its enormous dividend payout of 9.9%.