Why I Worry So Much, Part 2

Includes: DIA, QQQ, SPY
by: Alhambra Investment Partners

In Part 1 last week, I reviewed the two most important indicators of future economic growth - workforce growth and productivity growth. Both of these metrics are trending negatively with the workforce shrinking and productivity growing at rates less than 1/3 of what we experienced in the late 1990s. It is my view that changing these trends will require major changes in monetary, fiscal and regulatory policy. While it is certainly possible that the necessary changes happen soon and these negative trends reverse, I am more than a little skeptical. I certainly haven't seen any indication yet that Washington is about to experience a surge in humility, honesty and common sense. The obvious conclusion is that if these trends persist, economic growth is going to be hard to come by and that bears directly on corporate earnings and therefore stock prices.

In this week's installment, I want to review some valuation metrics and put the current market in perspective for long term investors. I believe that the most important factor for long term investing success is avoiding excessive losses that cause so much angst you abandon your investing plan (and if you don't have a plan, well …). To accomplish that an investor must know how the current market stacks up against history. If you are invested in US stocks and you don't know the potential downside it is impossible to align your portfolio with your risk tolerance. Of course, we can't say exactly how much the market will change in the future - up or down - but at least we should know what history says about current valuations.

If we are to review the current valuation and estimate the risk of loss in the US stock market, we must first choose a valuation method. There has been recently a lively debate about what valuation techniques are useful and while I don't want to replicate that here, I do think it is important to draw a distinction between techniques that are intended to provide risk information versus those that are intended to provide market timing advice. I will write today about three techniques that are intended to provide information regarding risk for the long term investor. They are not in any way useful for timing tops and bottoms in the stock market over the short term (defined as anything less than 5 years). Valuation tells you absolutely nothing about the direction of the stock market over a short time horizon. Valuation tells you about risk and nothing more. There are techniques that can assist the trader concerned with short term market movements, mostly related in some way to sentiment, but the key word there is "assist". To my knowledge there are no consistent, reliable techniques for predicting short term market movements.

So with that as our framework, let's see where the current market stacks up against history and see if we can estimate the risk of staying fully invested in the US stock market.

Shiller P/E

The Shiller P/E or Cyclically Adjusted Price to Earnings Ratio (CAPE) is a valuation technique developed by Robert Shiller of Yale University. The CAPE is the current price of the S&P 500 divided by the average of 10 years of earnings, adjusted for inflation. This method was originally suggested in the bible of value investing, Graham and Dodd's Security Analysis. Using CAPE smooths out the fluctuations in profit margins across the business cycle and it has proven through the years to be a good measure of risk and future returns. Other measures of P/E, such as trailing 12 month or expected earnings provide little guidance to the long term investor. Trailing earnings can be unusually depressed (such as at the bottom of the current bull cycle in 2009) and forward estimates are notoriously inaccurate, especially at turning points.

The current CAPE is 25.5 and the mean is 16.5. The all time high was 44.2 (at the peak of the internet mania) and the all time low was 4.8 in 1920. If we merely assume that the Shiller P/E will return to its mean value, it would mean a loss of roughly 35%. Of course, there are any number of ways that might happen. The S&P 500 could produce a long period of low returns while the 10 year average of earnings rises to reduce the ratio or the market price could drop to produce the same effect. History favors the latter scenario but there is no way to know in advance how the adjustment will occur.

So how do we use this information? I believe, for purposes of risk control, one should assume the ratio adjusts via a price drop and then see the effect on one's portfolio. For a moderate investor in a 60/40 portfolio, a 35% drop in the stock portion of the portfolio, assuming no change in the value of the bond portion, would mean a loss of 21% for the entire portfolio. So if that isn't a drawdown you can endure, your stock allocation probably needs to be adjusted down. Of course, for most investors the process is more complicated since the stock portion probably isn't fully invested in the S&P 500 but the likelihood is that a drop of that magnitude in the US market would be felt globally. Theoretically, with foreign markets at lower valuations, the drop in the non US portion of the portfolio should be less, but that is theory. The reality is that a drop of that size would probably be associated with a US recession and foreign markets, which depend to a large degree on US demand, might perform even worse.

One should also consider that this analysis only assumes a return to the long term mean. Bear markets, like bull markets, tend to overshoot. In 2009 the Shiller P/E bottomed around 13 and previous bear markets have bottomed at even lower values. One should also consider that the mean has been rising over time and prior to the great bull market that started in 1982, the mean was around 14.5. So a greater fall is certainly within the realm of possibility. Lastly, it would seem prudent to consider the bond portion of your portfolio. The role of bonds in a portfolio is to reduce risk and as investors have stretched for yield over the last few years, I fear that isn't true for most. If you have junk in the bond portion of your portfolio, it is unlikely that your bonds will reduce the overall drawdown.

Q Ratio

The Q Ratio was developed by Nobel Laureate James Tobin and is the current price of the market divided by the replacement cost of all its companies. Obviously, replacement cost is not something that is easy to determine and is subjective to say the least but using data from the Financial Accounts of the United States from the Federal Reserve produces a current value of 1.05. The all time peak was 1.63, also at the peak of the internet bubble, and the all time low was 0.28 right at the bottom of the last secular bear market in 1982. The long term mean is 0.68 so as with the CAPE, Tobin's Q ratio shows a market overvalued versus historical values. A drop to the mean would be, again, about a 35% drop in the US stock market producing a similar portfolio drawdown for the moderate risk investor.

Market Cap to GDP

This indicator has been described by Warren Buffet as "the best single measure of where valuations stand at any given moment". Using data from the St. Louis Federal Reserve's FRED system, this indicator has a current value of about 1.25 (the data is quarterly so there is some estimation involved). The all time high was over 1.5 at the 2000 peak and the all time low was around 0.3 in the early 80s. The mean is 0.65 so this measure indicates a greater overvaluation than the other two metrics and implies a drop of 48% to the mean. For the moderate investor a drop of this magnitude would shave almost 29% off the total portfolio value.

These are just three valuation techniques but they are ones with a fairly long history and have proven useful in the past for evaluating risk. They are not the only methods we value and there are others that show the US market to be overvalued, some by more and some by slightly less. There are no reliable valuation methods of which I am aware that show the current market undervalued versus historical experience. As I said at the beginning, these are not short term market timing indicators. It seems the debate about valuations has devolved into an all or none type argument where overvaluation means sell all stocks now but we don't view things that way. Overvaluation to us indicates heightened risk and we feel obligated as fiduciaries to take that into account and reduce our client's exposure to equities to a level commensurate with their risk tolerance.

I do not know how the stock market will perform in the near term and neither does anyone else. Markets are nothing more than a collection of individuals and in the short term are moved by the collective beliefs and emotions of those individuals. I know of no way to anticipate when those beliefs and emotions will change. Successful long term investors view the market through the lens of history and adjust their portfolios based on the most likely long term outcomes. Right now that means reducing exposure to stocks that are overvalued versus history.

So why do I worry so much? I worry because I don't think most people understand the level of risk they are taking now and believe they will be able to adjust in the heat of the moment after today's high valuations start to adjust lower. All the available research says different, that most investors will buy high and panic near a bottom and sell low. That seems to be human nature, a survival instinct that works against you as an investor. Rather than allow that instinct to destroy their investing plan, long term investors should adjust their portfolios to acknowledge the risks that exist today.