How do you choose stocks: via fundamentals or technical analysis? You’re wrong!
What’s your style: growth or value? You’re wrong!
Who do you like: Warren Buffett or Jim Cramer? You’re wrong!
No, I’m not just trying to be cute. Well, maybe I am trying to be a little witty, but the point here is important: As soon as you constrain yourself by saying “I do it this way” and more importantly, by saying or implying “I don’t do that sort of thing,” you start to hurt yourself. And the more dogmatic you are, the more committed you are to one approach to the exclusion of others, the more you’re likely to limit your investment returns.
Those of you who’ve seen my work in the past know I often use something called a QVGM ranking system, which I built for stockscreen123.com, to select stocks. Typically, I create a screen and then choose the top 15 or so passing stocks using the QVGM model as the basis for sorting. QVGM stands for Quality – Growth – Value - Momentum.
This model is, actually, an equally weighted combination of four separate stockscreen123 ranking systems, each based on well-known well-respected approaches to investing. There is more than one way to implement a given style (e.g., there are many different ways to define value) but if you look at each of these four style-based models (which you can do using the links in the prior paragraph, which take you to stockscreen123 pages that are visible to non-subscribers as well as subscribers), you’ll see that these approaches are reasonable expressions of their respective styles.
My goal today is to demonstrate that each of these styles, on their own, is inferior, substantially inferior, to a combination that simultaneously uses all of the styles.
The Quality Style
Of all the investing philosophies, this is the one that’s least likely to be referred to by name, but it is among the most revered by many who write and read how-to investing books. It’s also an important element of what such gurus as Ben Graham and Warren Buffett do, and is an important underpinning to Morningstar’s approach.
Return on capital lies at the core of this style based on the notion that a company that can turn $100 of capital into $20 of profit is superior to a company that can generate only $5 from the same pool of capital. Return on capital is, actually, a combination of margin and turnover (volume), and the ranking system also includes factors relating to both. Financial strength is also relevant. A company that produces $20 of profit based on $100 of capital all of which consists of equity is typically favored over the company that produces $20 of profit from capital consisting of $10 of equity and $90 of debt.
You’d think companies that fare well under such criteria ought to be worthy investments. To test that, I created a simple screen on stockscreen123.com that eliminates ADRs, stocks that trade OTC, stocks classified as Financial Services (Miscellaneous), many of which are closed-end funds, stocks priced below $5 and stocks whose 60-day average daily volume is less than 15,000 shares. Since I like to stay small, I also limit consideration to stocks with market capitalizations below $500 million. (Objective analysis tends to be more effective for smaller stocks, a topic I’ll cover in more detail another time. But for those who really prefer bigger companies, I’ll later present a list from that area as well.) From among the stocks that pass my broad micro-cap screen, I select the 15 that have the highest Quality ranks. Figure 1 shows how such a strategy would have fared against the Russell 2000 over the past five years (assuming the list is refreshed every four weeks).
(Click charts to expand)
This strategy didn’t bankrupt us. But considering all the admiration heaped upon high-quality companies, I think we should have hoped for a lot more than this, which for the most part matched the iShares Russell 2000 ETF (NYSEARCA:IWM), sometimes the portfolio won, sometimes the ETF won, but on the whole, neither approach had a meaningful edge.
This raises for us a dichotomy between the ivory tower and the real world. We know quality is important, but when we get into the market trenches, we don’t really think that much about it. Consider how often (never?) you hear CNBC leading with a story about how good a company’s return on equity looked in the latest quarter. Is it wrong-headed? Yes, probably. But you can’t fight the world. (How long can the market stay wrong? A lot longer than you can stay solvent.) And besides, let’s not judge the thundering herd too harshly. After all, don’t you prefer dessert to vegetables! (Quality metrics are the market’s vegetables. Other things comprise the desserts.)
The Value Style
I almost think I should stand up now and salute. For many, this is it! I couldn’t even begin to count how many articles and comments I’ve seen on Seeking Alpha wherein the writer proudly announces that he or she is a value investor and suggests, sometimes subtly and sometimes not so subtly, that those who aren’t paying attention to value are . . . well, you know.
Value adherents have a lot going for them. For starters, there’s common sense (it’s better to pay $5 for a stock that’s worth $10 than to pay $15). For another, there’s some incredible pedigree (besides being Quality devotees, Buffett, Graham, and yes, Morningstar focus heavily on value). And it even lies at the heart of academic theory (the key thread uniting all valuation approaches is that one way or the other, they purport to determine an appropriate price for a stock).
With so much passion and intellectual ammunition backing value, let’s see how a purely value-based strategy (picking the top 15 based on the Value rank) performs.
Yikes. What the $^&!@ happed? Gerstein must be an idiot who doesn’t know what value ratios to use. What, exactly, is in that Value ranking system? Well let’s see. We have trailing 12 month P/E, P/E based on estimated earnings, the PEG ratio, price/sales, price/cash flow, price/book. Obviously, there are other metrics. We can work with free cash flow, enterprise value, tangible, book value, etc. I’ve used them all over the years and I have to tell you that you’re not going to move the model much by tweaking. But if it will make value-oriented readers feel better, I will tell you that value did outperform the other pure styles by a lot in backtests that started at the earliest available date, 3/31/01 (such tests captured a very strong pro-value early-2000s interval). But the point to which I’m leading (that a combination style outperform all the pure approaches) would hold up even with 3/31/01 starting dates and I chose to focus on the last five years because I want you to see how things would have looked under what was undoubtedly a more challenging period.
The Growth Style
Who doesn’t love growth? Actually, that is what stocks are all about. Bonds have yields and these are often higher even than those of stocks designated as income plays. The only reason any rational person would ever consider equities is growth. For some, this can refer to a growing dividend stream. For most, it means a rising share price. Growing earnings and sales are needed and expected either way.
The Growth ranking system is based on what most would expect such a model to be based on. Sales and EPS growth rates over a variety of time periods figure prominently with a greater emphasis on EPS growth. Also considered is acceleration, situations where recent growth rates are above longer-term rates.
On paper, this, too, sounds appealing. In fact, many who are new to stock screening often start by creating growth screens. So let’s see if the strategy actually delivers. Figure 3 shows backtest results for a strategy that is identical to the one we looked at above except that here, the top 15 are selected on the basis of the Growth ranking system.
Now that might seem shocking. Above, I referred to quality metrics as akin to stock-market vegetables. Growth, on the other hand, might be seen by many as dessert. We love it. We crave it. We want it. So why don’t we trade on it?
Actually, those ever-present and often-irritating lawyers might have the answer. “Past performance does not assure future results.” Surely you’ve heard that before. All of the growth rates we can look at come from the past. Still, it’s natural to assume the future will resemble the past, and that is often reasonable because change is often evolutionary rather than revolutionary. But Table 3 suggests instances of abrupt revolutionary shifts can occur often enough to sandbag a purely growth-oriented investing strategy.
The Momentum Style
Now, let’s turn to a widely hated style, one that’s often associated with Jim Cramer and other supposed Wall Street demons (actually, the Cramer association is sketchy at best since he is quite adept at and pays a lot of attention to fundamentals, but the popular image is out there). Essentially, momentum means buying stocks that are hot, assuming they’ll stay hot.
This is exactly what blew up many brokerage accounts a decade ago, and there are many in the media who are only too happy to help people wipe out their accounts again and again. When I worked in dot-com, I constantly struggled and argued and annoyed the heck out of people as I strove to keep those stupid but much-beloved (by those who courted ad-supported page views and users who didn’t know better) “What’s Hot” lists as far away as possible from the highly visible sections of the sites.
Actually, though, not all momentum styles are alike and this method can be quite useful to those who approach the topic thoughtfully. The Momentum system I created is a bit of a blend. It looks for stocks with generally decent relative strength, but tries to catch them during what are presumed to be temporary pullbacks.
Let’s see how this works.
OK. At least we feel vindicated by the fact that Momentum underperformed the other styles! Phew. One aspect of the world makes sense.
The QVGM Combo Strategy
What happens if we combine all four strategies, as is done when I use the QVGM ranking system. Let’s see.
Voila. We have something that works!
There’s nothing fancy going on here. We’ve already considered Q, V, G and M separately. For those who have access to stockscreen123, all I did was use the dropdown menu to change from one ranking system to another before clicking on the button that cues the backtest to run. All other aspects of the screen remained constant.
Table 1 shows the micro-cap stocks that currently make the QVGM strategy.
Notice how the ranks are distributed. (Don’t try to re-create the QVGM score by presuming that the Q, V, G, and M scores are simply averaged. The weighting occurs with the raw scores, not the percentiles you see here.) Notice how the numbers run. None of these stocks had style-specific ranks high enough to have made it onto the list had we focused on a single style. But all the stocks are generally pretty good in all styles. None are dogs under any of the philosophies. So instead of getting into situations that rate extremely well in any particular style, we’re favoring stocks that rate generally well across the board; we’re favoring generalists over specialists.
The superior backtest results we saw for this approach are not at all coincidental. Although quants tend to whine louder and louder as the number of factors rises, there are several reasons why it actually makes sense to be a broad-based generalist.
No style is hot all the time: Sometimes growth is hot. Sometimes value is hot. You’ve heard this many times before so there’s no need to beat it to death now.
We don’t help ourselves when we pre-judge how a stock can demonstrate its merit: The stock market is an incredibly varied place, so much so that it almost seems a grammatical error to say “the stock market.” We really should use the plural: “the stock markets.” In truth, there are many completely different stock markets that just so happen to occupy the same space at the same time. One market is for in-and-out traders. Another is for the Warren Buffett fans. Another is for William O’Neil and his adherents. Etc., etc. etc. This is an important way to recognize when you’ve “arrived” as a bona fide investor. Newbies are quick to say “this” is the way. The pros know better (including the penultimate pro, Warren Buffett, whose actions often vary quite starkly from the expectations of those who write about him). You don’t help yourself when you refuse to tolerate merit that exists based on a set of criteria that happens to not be your favorite.
Think not just about your preferred style, but also about the style of the person who might buy your shares when you take a profit: When all is said and done, the stock market is about supply and demand. Let’s say you’re a value investor and you’re looking at two terrific value stocks. Value Stock “A” has stupendous value metrics, but not much else going for it. Any value investor in his or her right mind would love to own it, but others could care less. Value Stock “B” has pretty good value metrics, but can’t really measure up to those of “A.” Yet there are other things happening at “B” and the growth crowd has an interest as do those momentum folks who you despise with all your heart and soul. Bearing in mind the forces of supply and demand, and that neither stock can produce a good return for you unless new sources of demand for the shares materialize, which stock do you think offers better prospects: “A” which must look entirely to the value crowd for potential future buyers, or “B”, whose price can be pushed up by new demand from a wider assortment of potential buyers? Here’s another way to distinguish between stock market players and pretenders. For the players, it’s “show me the money.” When they want to take a profit, they just do it. They don’t feel a need to admire, respect, or even like the philosophy of the person who buys the shares they sell. So they prefer stocks that will be favored by as broad a range of investors as possible.
Extremes are rarely, if ever, sustainable: From a common sense point of view, this should be obvious. In fact, it extends well beyond investment analysis. (Check Chapter 9 of the “Tao Teh Ching,” which can be easily Googled!) A company can post an EPS growth rate of 70%, but is it likely to be sustainable? Probably not. It’s important that we avoid losing sight of this when we look at data. It’s tempting to take the “best” or “top” stocks in a list, but before doing that, we have to ask ourselves if we’re likely to wind up looking at unsustainable extremes. Investment is not about the data upon which a top-pick list is based. It’s about the future. So sustainability is the key consideration. When we work with quantitative rankings, the fewer the number of factors we use, the more exposed we are to unsustainable extremes. We control this by broadening our view. For example, it’s easy for a company with one stupendous and potentially unsustainable growth rate to rank well under a simple growth-oriented system, but it’s very hard for such a company to bubble up under an approach like QVGM unless it has a lot of other things going for it.
We need to hedge against data oddities: This is a huge and unfortunately for many quants ill-considered issue. (The formal name for this state of affairs is the “mis-specified model”). Strategies are typically based on unspoken assumptions that the data actually means what it seems to mean. This is often false. The well-known “value trap” is related to this (a P/E that looks low but doesn’t really mean much because earnings are about to fall off a cliff), but there are any worse scenarios, such as a P/E that looks low because EPS was temporarily boosted by a gain on an asset sale. Non-recurring items that temporarily enhance or depress earnings are legion, sometimes resulting from unusual transactions (e.g. an asset sales) and other times resulting from a corporate strategy (such as an acquisition or divestiture). In an ideal world, we’d be able to identify all such things and adjust for them, as the classic Graham & Dodd text tells us to do. But when working with modern data bases, this is often not feasible. So anybody who works with quantitative models must accept the fact that results will periodically and not as rarely as we wish be influenced by factors that cause companies to meet the letter of the law but falter in its spirit. Using large numbers of factors representing a variety of styles helps us hedge against the impact of such oddities.
As promised above, Table 2 is a list of high QVGM stocks drawn from those with market capitalizations between $5 billion and $50 billion. As was the case with the micros, the QVGM approach outperforms each style individually although as expected the relative superiority was much narrower than with the micro-caps. Interestingly, though, in this larger-cap group, the value-only approach was an especially poor performer apparently because it captured so many financial companies. The QVGM strategy wound up presenting value investors with a much more appealing list consisting of stocks that had pretty good, but not necessarily eye-catching valuation metrics combined with more appeal in terms of other styles.
I know there’s much unsaid here about the issue of large-cap versus small, but that topic really warrants a separate article unto itself. The role of Momentum is another potent topic also deserves more in-depth treatment and is on my to-do list. But for now, my goal was to illustrate the benefit of stylistic flexibility and show you (through the rank score breakdowns) the kinds of stocks such an approach can be expected to uncover.
In terms of the specific stock ideas listed here, those in Table 1 are more to my liking because of my preference for smaller companies. Actually, when it comes to real money (and aside from income stocks and ETFs), I like to go even smaller than that, to the under $3-stock price neighborhood, and that’s why the disclosure below says “No positions.” But I do use QVGM to narrow my list. And even before I got involved with StockScreen123 and developed QVGM, I had long preached the importance of stylistic flexibility (during my time at multextinvestor.com and reuters.com, at Money Show type events and in two books). Hopefully, you now understand why I’m so passionate about this topic and might be willing to broaden your own horizons a bit.
Disclosure: I have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.