Over time, investors have developed hundreds if not thousands of systems to use as guides in trying to answer the two most basic questions of portfolio construction – what to own, and when to own it. Some of these are rather complex or downright arcane, while others are so simple they can almost seem silly. In 23 years in this business, I have come across many systems which seem like they have worked at various times, and may even be worthwhile to try and implement, but more often than not, it is easy to come up with reasons to pass because the case is either not compelling enough or there are practical limitations to implementation, such as trading costs or scalability.
There are two basic principles, however, which for me have cut through the clutter and have left an enduring impression on me. And I would argue that both are fairly uncontroversial in the world of investing.
The first is that you want to try and buy things when they are cheap and avoid them when they are not. Pretty obvious in principle, but deceptively hard to follow in practice for many investors.
The second is captured in the saying “the trend is your friend.” Specifically, for whatever reasons it may be in place, once a trend is established, it is more likely to continue into the near future than not. Of course that will not always be true, but in this business it is about stacking the odds in your favor, and logic would dictate that you want to own things while they are in an uptrend, price-wise, and avoid then when they are not. Again, kind of a no-brainer. If there are many (a majority of) components and sectors of a market in uptrends, we can call this “technically strong.”
We can look at these two simple principles in conjunction with one another to provide us with a sensible way to evaluate risk and make portfolio management decisions in a logical and dispassionate manner.
The Matrix is a tool I use to visually evaluate the present risk and reward relationship in the US stock market. It is presented below:
It is a simple 10-by-10 grid, where the horizontal or x-axis is valuation, and the vertical or y-axis is technical strength. I will describe how we evaluate the market along these two variables and apply the observation to our approach to investing in stocks.
The Standard and Poor’s 500 Index (S&P 500) generally consists of the 500 largest publicly-traded companies in the US. The actual number of components will vary slightly around this number on account of mergers and reorganizations but will always be pretty close to 500 – and while there are actually 7x that number of publicly-traded companies in America, the top 500 typically comprise more than 80 percent of the total value of US stocks at any point in time. As such, looking at what is going on with these 500 stocks is a good place to start.
It is also important to understand that the S&P 500 (and the Russell and MSCI indices for that matter) are all market-cap weighted. The practical consequence of this is that large companies, based on market value, have an outsized influence on the index. For example, as I write this, Apple Inc. comprises 3.66% of the S&P 500 Index. If all of the stocks in the index were weighted equally, that would be 0.20% each. At 3.66%, Apple has more than 18x the influence on the index that Micron Technology, which at #120, does carry a 0.20% weight. This implies that approximately 380 companies in the index each have even less influence than that. The point here is that these cap-weighted indices tend to be more than a little “top-heavy” and it can be very helpful to look beyond the changes in the index value itself to try and figure out what is really going on.
Although the S&P 500 has origins back to the 1920’s, it was expanded to 500 holdings in 1957, so we have data on this measure of US equities going back 60 years we can look at.
In terms of valuation, we look at four measures, and compare them to past observations. By developing a composite of these four numbers we make sure that we are not placing too much emphasis on any one of them:
- Price-to-Earnings Ratio for the Index, as weighted by market-cap
- Median Price-to-Earnings Ratio of the index components
- Price-to-Sales Ratio for the Index, as weighted by market-cap
- Median Price-to-Sales Ratio of the index components
After evaluating all four and comparing them to their histories, we can place the market into “decile” along the cheap/expensive axis. If the market is “cheap”, meaning that a composite of these statistics produced lower than average values, historically-speaking, then it would be in a lower decile and placed somewhere on the left hand side of The Matrix. If it is “expensive” (higher than average values) then it would land in a high decile on the right-hand side.
It should be said at this juncture that merely determining the market is cheap compared to its own history is a very poor timing tool. Just because something is cheap, that doesn’t mean that it can’t, or won’t become a lot cheaper before the overall trend reverses. Likewise, an expensive market can get a whole lot more expensive. As John Maynard Keynes once quipped, “The markets can remain irrational a lot longer than you can remain solvent.” In fact, the frustration which this tries to express has been a hallmark of human nature for centuries. Isaac Newton lost a whole lot of money in the South Sea Bubble of 1720. In the wake of this, he famously said “I can calculate the motion of heavenly bodies but not the madness of people.”
That is the bad news. In the short run, markets can do just about anything, no matter how disconnected from reality or what a prudent investor might consider reasonable in term of providing a fair trade-off between risks and anticipated returns informed by economic principles and history.
The good news is that, in fact, these valuation metrics have a long history of reverting to their respective means, and as such, are very valuable in terms of evaluating prospective returns over a longer time frame, such as 10 years or longer. I have seen valuation models using various measures of these two metrics (price-to-sales and price-to-earnings) which have better than a 90% coefficient of correlation to future 12-year returns. If we are trying to invest for something a decade or more into the future (virtually everyone), then failing to take full account of this data and the predictive value of it is extremely foolish. This is why the horizontal axis of The Matrix is labeled “Cheap (Good Long-Term)” and “Expensive (Bad Long-Term)” By “Long-Term” we are talking 10-12 years or more.
So, how does the vertical axis (Technical Strength) fit into this? It informs us as to what we should anticipate, or at least the most likely outcome, over the shorter run.
The methodology here is similar, except that here we are simply asking 10 yes-or-no questions about the S&P 500 Index and its components:
- Six questions about what has happened to the index value itself over the previous 50, 125 and 200 days
- Four questions which survey the 500 components and give them equal weight in the answer
The number of “yes” answers we get determines which decile the market is in from a technical perspective and where to place the market into The Matrix along the vertical axis.
Ideally, you want to see a situation where you end up in the Northwest quadrant (potentially good for both short and long-term), and if the overall market doesn’t get you there, then try to find markets that do (a sector, or overseas?). Alternatively, you clearly want to avoid markets when this exercise puts you into the Southeast quadrant (technically weak and expensive).
Where are we today? I will leave that for a future post in the near future – stay tuned!