Mark Hulbert’s Stock Newsletter Sentiment Index
Posted in Individual Gurus, Sentiment Indicators
April 16, 2010
A reader suggested that we review the stock market commentary of Mark Hulbert, editor of the Hulbert Financial Digest, which tracks the recommendations of a wide range of investing newsletters. He is also a regular columnist at MarketWatch. Because Mark Hulbert uses his Hulbert Stock Newsletter Sentiment Index (HSNSI) as a principal quantitative tool in formulating his market outlook, we evaluate the usefulness of that index in predicting stock market returns rather than his qualitative commentary. HSNSI “reflects the average recommended stock market exposure among a subset of short-term market timers tracked by the Hulbert Financial Digest.” Mark Hulbert presents HSNSI as a contrarian signal for future stock returns; when HSNSI is high (low), he views the outlook for stocks as generally bearish (bullish). Using a sample of 224 values of HSNSI over the period 7/22/02-4/15/10 (generated by searching MarketWatch.com for “HSNSI” and its predecessor “HSSI”) and contemporaneous daily closes of the S&P 500 Index, we find that:
The average value of HSNSI for the sample is 23.4%, with standard deviation 26.1% (high volatility). The high for the sample is 67.8% (on 11/20/06), and the low is -36.1% (on 9/20/08). Assuming normal variable distributions, the sample is large enough to test predictive power with reasonable reliability for short-term stock returns (a few trading days or weeks), but is not large enough to test HSNSI reliably as a long-term stock market indicator. The sample may be biased due to selection and clustering, because Mark Hulbert may be more likely to cite HSNSI, and cite it frequently, when either HSNSI or stock market behavior is extreme.
The following chart superimposes the HSNSI sample on a plot of the S&P 500 Index. The blue dashed line is the average HSNSI for the sample. As noted, HSNSI has high variability. Visual inspection suggests that HSNSI, like many other sentiment indicators, tends to be high (low) during market advances (declines).
To add precision to that observation, we examine how HSNSI relates to future S&P 500 Index returns.
The following scatter plot relates the return on the S&P 500 Index over the next five trading days to HSNSI. The Pearson correlation for the two series is -0.10, indicating a slight tendency for the stock market to be relatively weak (strong) when HSNSI is high (low). The R-squared statistic is 0.01, indicating that variation in HSNSI explains 1% of stock market returns over the next five days.
Using future return intervals of 21 and 63 trading days yields R-squared statistics of 0.00 and 0.00, respectively, for the relationship between HSNSI and future stock market returns.
Note that the sample may have some selection/clustering biases at this point. Clustering leads to some overlap of five-day test intervals, effectively reducing sample size and perhaps overweighting extreme values of HSNSI. What happens if we alleviate clustering?
To eliminate measurement interval overlap and alleviate clustering bias, we winnow sample points so that no two surviving points lie within the same return interval.
The next chart depicts the relationship between the return on the S&P 500 Index over the next five trading days to HSNSI for a winnowed sample (171 surviving sample points), ordered from lowest to highest HSNSI. Ordering helps determine whether some ranges of HSNSI might be more useful than others. Note that the horizontal axis is not time-sequential. The Pearson correlation between the two series is -0.18, again indicating some tendency for the stock market to be relatively weak (strong) when HSNSI is high (low). The R-squared statistic indicates that HSNSI explains about 4% of stock market returns over the next week.
About half the explanatory power of HSNSI in this case comes from the large peak at the left side (from Fall 2008).
Using future return intervals of 21 and 63 trading days on appropriately winnowed samples yields R-squared statistics of 0.02 (with negative correlation) and 0.00, respectively, for the relationship between HSNSI and future stock market returns. Winnowed sample sizes are 72 and 28, respectively.
For a different perspective, we evaluate future stock market returns by winnowed tercile of HSNSI values.
The final chart summarizes average S&P 500 Index returns over the next five, 21 and 63 trading days by tercile (ordered third) of the winnowed HSNSI distributions for the entire sample period. Results generally support a belief that high (low) values of HSNSI indicate relatively high (low) future stock market returns.
Note that the terciles for the 21-day and 63-day future returns are very small, 24 and 9, respectively. One or two new observations could change results substantially.
See “Purifying Stock Market Sentiment Indicators” for a summary of a study finding that poll-based sentiment indicators, including HSNSI, are proxies for past stock market behavior and that they add little or no value to price-based indicators.
HSNSI is, in fact, more strongly related to past than future stock returns. Pearson correlations between HSNSI and S&P 500 Index returns over the past five, 21 and 63 trading days (unwinnowed) are 0.38, 0.60 and 0.69, respectively. When the stock market advances (declines), HSNSI tends to rise (fall). For returns over the future five, 21 and 63 trading days (unwinnowed), correlations are -0.10, 0.01 and -0.01, respectively. In other words, it is much easier to predict the future value of HSNSI based on past stock returns than it is to predict future stock returns based on the value of HSNSI.
In summary, evidence from simple tests indicate that, while results are mixed, the Hulbert Stock Newsletter Sentiment Index may have some power to predict future stock market returns.
See below for Mark Hulbert’s comments on a prior iteration of this analysis.
See Guru Grades for a snapshot of the accuracies of various experts in forecasting the direction of the U.S. stock market, including links to detailed individual evaluations. We do not include Mark Hulbert in the snapshot because the above review is fundamentally different from the reviews of guru commentaries.
On 10/31/08, Mark Hulbert wrote:
Thanks for your rigorous analysis of the Hulbert Stock Newsletter Sentiment Index. I am sorry you had such limited data with which to conduct your analysis, which — as you conceded — greatly reduces the potential power of any test results. We have found–using our data set back to 1985, and correcting for various statistical biases, including the ones you mention–markedly greater statistical significance than you found from your more limited sample.
Response:
The sample used derives from the nearly [now over] 200 mentions of the Hulbert Stock Newsletter Sentiment Index in MarketWatch columns over the past 6+ [now 7+] years. It may be that there is some skew in the published data, or that the predictive power of the index varies over time (and has diminished). If you have made public your study of the index, we will be glad to cite it. Or, if you make public the complete dataset, we will be glad to run analyses on the complete set, as well as subsets to test robustness over time.
