When not to buy stocks

You don’t want to buy US stocks right now. If you own them, be prepared for a deep downturn or long period of very low returns, or more likely both.

To value an investment is to assess its likely returns over a holding period. Very few in the financial profession take a statistically valid approach to valuation. In fact, the industry promotes ignorance among professionals and the public.

Statistical validity requires lots of data, both in sample size and in time. If someone smokes a pack a day, it doesn’t mean they will die this year. Better if we say that out of fifty such smokers of a certain age two are likely to die this year, compared to one non-smoker. And don’t be shocked if all the smokers live and three non-smokers die. But track 10,000 people over 10 years and you can lay a big bet, provided you don’t get too precise.

It works the same with securities. Don’t get too confident that your meticulously selected picks will beat the market in any given year. Even Warren Buffett has lagged the S&P for a third of his career, at one point by as much as 50%. Like card-counting, his method only proves its worth when the law of large numbers is in effect.

To get truly precise you need to step out in scale and in time. There are a number of reliable valuation methods, such as Robert Shiller’s CAPE, Tobin’s Q, or Market Cap / GDP. They have some key differences from Wall Street nonsense like forward operating multiples:

  1. They look at an entire stock market or broad index
  2. They use input data that is either not volatile (sales or balance sheets) or smoothed over time (a 10-year moving average of earnings)
  3. They look forward far enough to make short-term fluctuations irrelevant (10+ years)

Right now they are all saying the same thing: this market is nuts.

The master of market valuation is John Hussman, a professor turned fund manager who has honed long-term S&P 500 forecasting to a true science. In his latest weekly comment he includes two graphics that should dissuade anyone from owning US equities. The first compares valuations (in blue, upside-down scale) to actual subsequent 12-year returns (in red, normal scale):

That is a shockingly tight fit. To say that the blue line is a prediction is not to do it justice – smooth away the volatility and those are your future returns, sure as your age in 12 years is your current age plus 12. If you bought in the late 70s or early 80s, your 12-year annualized returns were in the teens (meaning you made several times your money). If you bought in 1999 or 2000, your return by 2012 was only 1-2% (meaning you were up about 20%).

From now until 2029 the annualized returns of the S&P 500 will be between zero and three percent, including dividends. Actually, the dividends are about all you’ll get, since the current yield is 2%. For reference, a new 10-year Treasury note if bought today will give you a guaranteed 2.4%. These numbers are before inflation, by the way – factor that in and you could easily be in negative territory.

Markets don’t move for very long in smooth, upward-sloping curves. As surely as today’s buyer is locking in low returns, he is also committing himself to one hell of a ride if he hopes to get his 2%. The following graph (from the same letter, and inspired by a request from my buddy Aaron Brask), shows the range of outcomes over a 4-year horizon from a given valuation level:

The dots on the left are the results of good times to buy: a tolerable downside of 20-30% and an upside where doubles are common. Those dates are from bleak times like the Carter years and the bottoms in 2003 and 2009. The dots on the right are the results of buying when everything looked rosy, as in the dot-com, housing, and current bubble years. 40% losses are common, and gains are 20% at best.

You can see the last two bubbles and crashes below. Just seeing this 25-year chart of the S&P 500 should be enough to give you pause.

What we’ve got today is a lousy, lopsided bet. We’re risking very large near-term losses for the possibility of small temporary gains and the certainty of minuscule long-term returns.

Thank goodness buying and holding US stocks isn’t the only option anymore. With the range of low-cost ETFs available, you can diversify geographically and into different asset classes. By owning a largely uncorrelated mix of fairly-valued foreign stocks, corporate bonds, government bonds, commodities, and real estate, and perhaps adding a momentum factor, your account becomes far more stable and likely to deliver good long-term returns.

It’s a big world out there, and I’ll have more to say about global asset allocation in future posts.


PS – ignore this message if you’re either 1) too young to have much invested, in which case I’d be rooting for lower valuations, or 2) too rich to care and would rather not pay the taxes.

Leave a Reply

Your email address will not be published. Required fields are marked *