August 24th, 2014
by Elliott Morss, Morss Global Finance
I have always found Paul Samuelson's Ph.D. thesis on how quickly and well the stock market adjusts for new information quite convincing. So it would surprise me to find many economic journalists who outperform the market. Wall Street will pay a lot more for investment gurus than they can ever make in journalism.
So my hypothesis is that economic journalists will not outperform the market. Seeking Alpha (SA) is one of the best sites for information on financial investments. And like many financial sites, it has numerous writers recommending investments - mostly stocks and ETFs. But unlike other sites, SA has just courageously introduced a new feature that allows you to judge just how good its writers are - it tracks how well the stocks they recommend have actually done. I say "courageous" because I am certain a number of its writers will not be pleased with readers having a way to judge their performance. So in what follows, I examine the performance of their most popular "Long Idea" writers.
SA gives three periods for performance data: 3 months, 6 months, and one year. In my analysis, I use the longest period for which data on individual stock recommendations are available. Many of SA journalists recommend stocks numerous times. In my analysis, I only include the first recommendation. Some authors recommend both purchases and sales (shorts). I exclude any stocks that authors recommend for purchase and sale. My argument is that inasmuch as these are SA's "long idea" writers, they should not be recommending both purchases and sales in the same year (in some instances writers recommend purchases and sales in consecutive months). As a benchmark for these recommendations, I use the S&P 500. Since my data came from SA on August 15th, my benchmarks are May 15th, February 15th, and August 15th 2013.
Data on SA's "most read authors in last 90 days is given in Table 1 with IAEResearch with the most readers and Bret Jensen ranking 5th. Note the large number of recommendations actually made by Quoth the Raven and Bill Maurer versus how many were actually used. This is because repeat recommendations and buy/sell recommendations for the same stock were eliminated.
Table 1. Background Data on Authors
Table 2 provides "performance data" on the five authors - how well the stocks they recommended people to buy did over the three periods. IAER and Quoth the Raven have done quite well over the last year while Brett Jensen at least matched the S&P.
Table 2. - Performance on Buy Recommendations
It is important to read Table 2 with the number of observations behind each estimate in hand (Table 3). The large sample size behind Jensen's one year performance estimate provides support for my contention that picking stocks is a pretty random exercise: that if you pick enough, your performance will reflect market performance. On the other hand, Maurer's exceptional performance for 6 months has to be tempered by the fact it was based on only two observations.
Table 3. - Observations Behind Table 2 Buy Estimates
Data on how the authors did on their sell recommendations is quite interesting. These data are presented on the three authors who recommended short sales. Keep in mind that with a short sale, you only make money if the price falls. And to "beat the market" for one year performance, the percent fall would have to be greater than 17.7%. In fact, the short recommendations of all three authors increased in price over the one year period, with Maurer and Jensen's recommendation gaining more than the benchmark!
Table 4. - Performance on Sell (Short) Recommendations
Again, these results, while amusing, should be read in conjunction with sample sizes given in Table 5. And in fairness to the authors, it has to be said that the sample sizes are too small to take seriously.
Table 5. - Observations Behind Table 4 Sell Estimates
Can anything really be concluded from the data presented above on the question on whether most stock market recommendations are random (no better than average stock market performance in long run) or something special? Not really. While the one year buy recommendations of IAER and Quoth the Raven are quite impressive, the time periods are too short to say they will beat the market on a regular basis.
But suppose they are unique. If so, some financial house will "computerize" their selection methods. They will then buy enough so stocks are again nothing more than a random selection.
Are there real gurus who have stood the proverbial "test of time"? I think so. My current favorite is John Reese of Validea. His "Hot List" performance since 2003 is given in Table 6.
Table 6. - Validea Hot List Performance
 Paul A Samuelson, "Proof That Properly Anticipated Prices Fluctuate Randomly", Industrial Management Review, 6:2, 1965 (Spring) and Burton G. Malkiel, "A Random Walk Down Wall Street", W.W Norton & Co., Inc., 2007.