|Executive Summary | Portfolio | Guru Analysis | Watch List|
|Executive Summary||October 12, 2012|
The economic news has been fairly positive over the past couple weeks, with improving data coming out of the employment market, manufacturing sector, and service sector.
The job market improvement wasn't overwhelming, but it was solid. The Labor Department said 114,000 jobs were added during September, and it also revised the August increase significantly higher, to 142,000 (though nearly a third of those were government jobs). That was enough to drop the headline unemployment rate to 7.8% -- the first time it has fallen below 8% since January 2009. And the decline could not be attributed to people giving up looking for a job, as the number of people not in the labor force actually declined by more than 200,000. Still, the "U-6" rate, which includes the headline number as well as discouraged workers and those working part-time because they can't find full-time work, stayed the same, at a very elevated 14.7%.
The manufacturing sector, meanwhile, which had contracted ever so slightly from June through August after a three-year expansion, rebounded in September, according to the Institute for Supply Management. ISM's manufacturing index moved back into expansion territory, and the survey's sub-indices for employment and new orders both made nice jumps.
ISM's service sector index showed that the service sector also expanded in September, and did so at a faster pace than it did in any month since March, a great sign. The sector has now expanded for 33 straight months.
Recent housing market data, meanwhile, indicates the recovery there is continuing. New home sales in August were about the same level they were in July, which was 24% above the year-ago level, according to the Commerce Department. And, sale prices climbed to the highest level in five years. The National Association of Realtors said pending home sales fell 2.6% in August, but they were 9.6% above year-ago levels.
Consumer data was a bit of a concern, however. Real disposable personal income fell 0.3% in August, according to a new government report. Real personal consumption expenditures rose 0.1% for the month, which led to the personal savings rate -- that is, personal savings as a percentage of disposable personal income -- falling to 3.7% in August. Still, while a concern, that savings rate is still well above levels we saw heading into the Great Recession.
Lingering in the background of all this is the U.S. "fiscal cliff" -- the combination of expiring tax cuts and unemployment benefits and automatic budget cuts scheduled to go into effect in 2013 if Congress doesn't get its act together soon. The International Monetary Fund issued a report this month saying that inaction on the cliff situation could, "at the extreme ... result in a fiscal withdrawal of more than 4 percent of GDP in 2013". Given that the IMF is projecting growth of just over 2% for the U.S. next year, that would tip us back into recession. The message is clear: Policymakers need to act, and act soon.
Since our last newsletter, the S&P 500 returned -1.0%, while the Hot List returned 0.3%. So far in 2012, the portfolio has returned 15.6% vs. 13.9% for the S&P. Since its inception in July 2003, the Hot List is far outpacing the index, having gained 161.4% vs. the S&P's 43.2% gain.
Man vs. Machine
As we move deeper into Fall, we're getting to a time of year that is particularly rife with predictions. Pundits and pollsters tell us who will win the Presidential and Congressional elections, and what it will mean for the country, economy, and stock market. Meteorologists start to tell us what kind of winter we can expect. And talking heads tell us who they think will emerge victorious from Major League Baseball's playoffs, the NFL's weekly battles, and the NBA's upcoming season.
In the end, most of them will be wrong. At least, that's what history and the data tells us. Consider Philip Tetlock's research, which he detailed in his 2005 book Expert Political Judgment: How Good Is It? How Can We Know? Tetlock, a professor at the University of Pennsylvania, conducted a seven-year study in which both supposed experts and nonexperts were asked to predict an array of political and economic events. The study was the largest of its kind ever done, looking at more than 80,000 predictions. What he found was that "experts" could explain only about 20% of the total variability in outcomes in their predictions. The results were pretty consistent regardless of the expert's educational background, duration of experience in the pertinent field, and ability to access to classified materials. They also showed that, generally, the more famous the alleged expert, the less accurate he or she tended to be!
That may seem shocking, given the exalted status -- and plentiful airtime -- many so-called experts are given in the media. Whether the story involves politics, economics, sports, or some other field, "experts" are constantly being asked to tell readers and viewers what is going to happen in the future. And, according to Tetlock's research, usually about 4 out of 5 times they end up being wrong.
That data is striking, but my own experience with Validea has supported the general findings. Initially, I created the company as a service that tracked the recommendations of various pundits who were featured in the media. Generally, the performance of these "experts" wasn't great. One of the most striking lessons was that there was no consistency from one period to the next in terms of which pundits performed well. Someone might have a good quarter, but they wouldn't continue it over the longer term.
What Tetlock found -- similar to what I also found in my experience in the stock market -- is that computer models are better predictors of future outcomes than most people. Relatively crude algorithms that Tetlock examined could explain 25 to 30 percent of outcomes, while more sophisticated algorithms could explain 47 percent -- more than twice as much as "expert" human forecasters.
Experts did tend to beat non-experts, for the record. In a 2010 paper in the journal Critical Review in which he revisited the issue of experts' predictive abilities, Tetlock said his work "should be roughly as annoying to elitists as to populists: Experts do not know nearly as much as they think they do, and they work hard to cover up mistakes, but they do still at least perform better than the general public." (He also spoke of the "model-of-man-trumps-man effect" -- the tendency for prediction models constructed using experts' strategies to outperform those same experts themselves, which I found particularly interesting given the nature of my guru-inspired models. He says decades of psychological research document that effect.)
James O'Shaughnessy, the growth-value guru featured in this issue's Guru Spotlight, also gives some excellent insights into the issue of human predictive ability in What Works on Wall Street. In the book, O'Shaughnessy cites additional studies that all found that human prognosticators couldn't match statistical-actuLato forecasting models. In one study, for example, an actuLato model did better in predicting whether certain high school students would be successful in college than did admissions officers at many colleges. In another, a researcher named Jack Sawyer reviewed 45 different studies that compared human and actuLato predictive ability. "In none [of the 45] was the clinical, intuitive method -- the one favored by most people -- found to be superior," O'Shaughnessy writes. "What's more, Sawyer included instances in which the human judges had more information than the model and were given the results of the quantitative models before being asked for a prediction. The actuLato models still beat the human judges! "
So, how can most humans so consistently fail to beat computer models when it comes to predictive powers? Part of it goes back to what we looked at in the last issue of the Hot List: People are emotional creatures, and emotions lead to inconsistency in how we assess problems. O'Shaughnessy explained it with great eloquence: "Models beat human forecasters because they reliably and consistently apply the same criteria time after time. Models never vary. They are always consistent. They are never moody, never fight with their spouse, are never hung over from a night on the town, and never get bored. They don't favor vivid, interesting stories over reams of statistical data. They never take anything personally. They don't have egos. They're not out to prove anything."
Another possible reason experts fail to beat models involves something a bit more disconcerting: the notion that many experts don't actually care whether their predictions are right. In a critique of Tetlock's work that was also published in 2010 in Critical Review, for example, University of Warwick Professor Steve Fuller contended that many of the experts quoted by the media may be "pundits whose predictions are themselves meant to be interventions that reinforce or subvert existing political tendencies. Here the media may not be so much reporting politics as participating in it, however inadvertently."
Models don't have such agendas -- though that doesn't mean they are perfect. They get it wrong plenty of times. But here's the thing: In the stock market, they don't need to be perfect. If you are right on a little over half of your picks, you should make some nice money. The Hot List, for example, has an accuracy rate of 56% since its mid-2003 inception (meaning it has made money on about 56% of its picks). And that has been enough to generate annualized returns of nearly 11% vs. just 4.0% for the broader market. In fact, my best-performing individual 10-stock guru-based portfolio -- the one I base on the writings of Motley Fool co-creators Tom and David Gardner -- has an accuracy rate of just 50.5%. Yet it has posted annualized returns of more than 13% -- its winners have won bigger than its losers have lost.
To be sure, there are experts who are legitimately experts -- Tetlock's study showed that to be true. When it comes to stock-picking, however, we aren't so naive or overconfident that we think we are among the few who have such abilities. Fortunately, some of those true experts -- investors like Peter Lynch, O'Shaughnessy, Joel Greenblatt, and Benjamin Graham -- have been kind enough to leave us a blueprint for how to produce strong returns over the long haul. Our job, we believe, is not to outthink them or re-invent the exceptional wheels they've already invented; it is to implement their proven approaches in a disciplined, unfailing manner. So far, through its nine-plus years, that has worked quite well for the Hot List, and I think it's still the best way to post strong returns as we move forward.
Guru Spotlight: James O'Shaughnessy
To say that James O'Shaughnessy has written the book on quantitative investing strategies might be an exaggeration -- but not much of one. Over the years, O'Shaughnessy has compiled an anthology of research on the historical performance of various stock selection strategies rivaling that of just about anyone. He first published his findings back in 1996, in the first edition of his bestselling What Works on Wall Street, using Standard & Poor's Compustat database to back-test a myriad of quantitative approaches. He has continued to periodically update his findings since then, and today he also serves as a money manager and the manager of several Canadian mutual funds.
In addition to finding out how certain strategies had performed in terms of returns over the long term, O'Shaughnessy's study also allowed him to find out how risky or volatile each strategy he examined was. So after looking at all sorts of different approaches, he was thus able to find the one that produced the best risk-adjusted returns -- what he called his "United Cornerstone" strategy.
The United Cornerstone approach, the basis for my O'Shaughnessy-based Guru Strategy, is actually a combination of two separate models that O'Shaughnessy tested, one growth-focused and one value-focused. His growth method -- "Cornerstone Growth" -- produced better returns than his "Cornerstone Value" approach, and was a little more risky. The Cornerstone Value strategy, meanwhile, produced returns that were a bit lower, but with less volatility. Together, they formed an exceptional one-two punch, averaging a compound return of 17.1 percent from 1954 through 1996, easily beating the S&P 500s 11.5 percent compound return during that time while maintaining relatively low levels of risk.
That 5.6 percent spread is enormous when compounded over 42 years: If you'd invested $10,000 using the United Cornerstone approach on the first day of the period covered by O'Shaughnessy's study, you'd have had almost $7.6 million by the end of 1996 -- more than $6.6 million more than you'd have ended up with if you'd invested $10,000 in the S&P for the same period! That seems powerful evidence that stock prices do not -- as efficient market believers suggest -- move in a "random walk," but instead, as O'Shaughnessy writes, with a "purposeful stride."
O'Shaughnessy has revamped his strategies over the years, most recently in his new, updated version of What Works on Wall Street (which I reviewed in the Nov. 25, 2011 edition of the Hot List). I've chosen not to tinker with my O'Shaughnessy-based models, however, given the exceptional performance they've had over the long term. The models discussed here are thus based on O'Shaughnessy's 1996 version of What Works on Wall Street.
So, let's start with the value stock strategy. O'Shaughnessy's Cornerstone Value approach targeted "market leaders" -- large, well-known firms with sales well above those of the average company -- because he found that these firms' stocks are considerably less volatile than the broader market. He believed that all investors-even the youngest of the bunch -- should hold some value stocks.
To find these firms, O'Shaughnessy required stocks to have a market cap greater than $1 billion, a number of shares outstanding greater than the market mean, and trailing 12-month sales that were at least 1.5 times the market mean.
Size and market position weren't enough to make a value stock attractive for O'Shaughnessy, however. Another key factor that was a great predictor of a stock's future, he found, was cash flow. My O'Shaughnessy-based value model calls for companies to have cash flows per share greater than the market average.
O'Shaughnessy found that, when it came to market leaders, another criterion was even more important than cash flow: dividend yield. He found that high dividend yields were an excellent predictor of success for large, well-known stocks (though not for smaller stocks); large market-leaders with high dividends tended to outperform during bull markets, and didn't fall as far as other stocks during bear markets. The Cornerstone Value model takes all of the stocks that pass the four aforementioned criteria (market cap, shares outstanding, sales, and cash flow) and ranks them according to dividend yield. The 50 stocks with the highest dividend yields get final approval.
The Cornerstone Growth approach, meanwhile, isn't strictly a growth approach. That's because one of the interesting things O'Shaughnessy found in his back-testing was that all of the successful strategies he studied -- even growth approaches -- included at least one value-based criterion. The value component of his Cornerstone Growth strategy was the price/sales ratio, a variable that O'Shaughnessy found -- much to the surprise of Wall Street -- was the single best indication of a stock's value, and predictor of its future.
The Cornerstone Growth model allows for smaller stocks, using a market cap minimum of $150 million, and requires stocks to have price/sales ratios below 1.5. To avoid outright dogs, the strategy also looks at a company's last five years of earnings, requiring that its earnings per share have increased each year since the first year of that period.
The final criterion of this approach is relative strength, the measure of how a stock has performed compared to all other stocks over the past year. A key part of why the growth stock model works so well, according to O'Shaughnessy, is the combination of high relative strengths and low P/S ratios. By targeting stocks with high relative strengths, you're looking for companies that the market is embracing. But by also making sure that a firm has a low P/S ratio, you're ensuring that you're not getting in too late on these popular stocks, after they've become too expensive. "This strategy will never buy a Netscape or Genentech or Polaroid at 165 times earnings," O'Shaughnessy wrote, referring to some of history's well-known momentum-driven, overpriced stocks. "It forces you to buy stocks just when the market realizes the companies have been overlooked."
To apply the RS criterion, the Cornerstone Growth model takes all the stocks that pass the three growth criteria I mentioned (market cap, earnings persistence, P/S ratio) and ranks them by RS. The top 50 stocks then get final approval.
The Growth/Value Investor model I base on O'Shaughnessy's two-pronged approach has been a one of my best performers since its inception back in 2003, with my 10-stock O'Shaughnessy-based portfolio gaining 146.7% (10.3% annualized) since inception, while the S&P 500 has gained just 44.1% (4.0% annualized; figures through Oct. 9).
The O'Shaughnessy-based portfolio will pick stocks using both the growth and value methods I described above. It picks whatever the best-rated stocks are at the time, regardless of growth/value distinction, meaning the portions of the portfolio made up of growth and value stocks can vary over time. For a long time, the portfolio has been leaning strongly to the growth side. It's shifted back toward an even balance at times in 2012, but currently seven of the ten holdings are growth picks. Here's a look at the portfolio's holdings:
BRLI -- Bio-Reference Laboratories, Inc. (Growth)
TJX -- The TJX Companies, Inc. (Growth)
AZN -- AstraZeneca PLC (Value)
CHL -- China Mobile Ltd. (Value)
FRED -- Fred's Inc. (Growth)
STO -- Statoil ASA (Value)
DUF -- Duff & Phelps Corp. (Growth)
RUE -- Rue21 Inc. (Growth)
ECPG -- Encore Capital Group, Inc. (Growth)
LKQ -- LKQ Corporation (Growth)
Not Just Numbers
O'Shaughnessy is a pure quant, but you should be aware that some of his most critical lessons are less about specific criteria and numbers than they are about the general mindset an investor must have. Perhaps more than anything else, O'Shaughnessy believes in picking a good strategy and sticking with it -- no matter what. In What Works on Wall Street, he writes that in order to beat the market, it is crucial that you stay disciplined: "[C]onsistently, patiently, and slavishly stick with a strategy, even when it's performing poorly relative to other methods."
The reason involved human emotions, which cause many investors to bail on a good approach and jump onto hot stocks or strategies that are often overhyped and overpriced. "We are a bundle of inconsistencies," he writes, "and while that may make us interesting, it plays havoc with our ability to invest our money successfully. Disciplined implementation of active strategies is the key to performance." Wise words, whether you follow O'Shaughnessy's approach or another proven method
News about Validea Hot List Stocks
Nu Skin Enterprises Inc. (NUS): Nu Skin shares have jumped about 12% since our last newsletter (as of Thursday evening). There was no clear catalyst for the jump, but it seems the gains may be a rebound that followed an overreaction to claims that Nu Skin was operating illegally in China. Several counterpoints have been written refuting those claims in the past two weeks, which may have helped push shares higher.
The Next Issue
In two weeks, we will publish another issue of the Hot List, at which time we will rebalance the portfolio. If you have any questions, please feel free to contact us at email@example.com.
The names of individuals (i.e., the 'gurus') appearing in this report are for identification purposes of his methodology only, as derived by Validea.com from published sources, and are not intended to suggest or imply any affiliation with or endorsement or even agreement with this report personally by such gurus, or any knowledge or approval by such persons of the content of this report. All trademarks, service marks and tradenames appearing in this report are the property of their respective owners, and are likewise used for identification purposes only.
Validea is not registered as a securities broker-dealer or investment advisor either with the U.S. Securities and Exchange Commission or with any state securities regulatory authority. Validea is not responsible for trades executed by users of this site based on the information included herein. The information presented on this website does not represent a recommendation to buy or sell stocks or any financial instrument nor is it intended as an endorsement of any security or investment. The information on this website is generic by nature and is not personalized to the specific situation of any individual. The user therefore bears complete responsibility for their own investment research and should seek the advice of a qualified investment professional prior to making any investment decisions.
Performance results are based on model portfolios and do not reflect actual trading. Actual performance will vary based on a variety of factors, including market conditions and trading costs. Past performance is not necessarily indicative of future results. Individual stocks mentioned throughout this web site may be holdings in the managed portfolios of Validea Capital Management, a separate asset management firm founded by Validea.com founder John Reese. Validea Capital Management, which is a separate legal entity and an SEC registered investment advisory firm, uses, in part, the strategies on the web site to select stocks for its clients.