Over-Expectations

It's been a while since my last blog. Spring, little league baseball and too many hours spent pondering the state of the world have distracted me from my scheduled semi-occasional written musings. Not that I haven't been thinkin' -- I just haven't been writin'.

I've spent many of my waking hours trying to figure out what to make of all this. The economy doesn't seem to be in free-fall any longer. It is, as best I can tell, far from a recovery. But still, we have reached somewhat of a steady-state. Consumption, income and employment are all below previous highs. I expect they might stay that way for a while. Stocks have recovered strongly from the lows reached in early March.

As money managers, we have gone from a six-month state of constantly reacting to rapid changes (I made more trades from Sept 08 to April 09 than in the previous six years combined) to now having a little breathing room to make some forward-looking decisions.

Most of my head-scratching the past six weeks havs been about stocks. What should we expect from stocks? After staring at data series to the point of blindness, I have some thoughts about expectations. Namely -- people expect too much from stocks. We are prone to anchoring our expectations in the good years (double-digit returns) and forget to factor in the so-so years (losses). I still come across articles by various financial planning pundits such as Suze Orman that counsel us to expect high returns over any given reasonably long holding period. Such advice is usually accompanied by a cherry-picked starting and ending point for the analysis. The starting point is invariably a period of low stock valuation, and the ending point is one of normal or high valuation. Of course returns were high with such assumptions. But, if you bought stocks in 1966, you have had a lousy 43 years. Bonds would have done better.

But really -- what can we expect with some degree of certainty, or at least rationality? Your stock portfolio grows faster than inflation due to three and only three sources:

- Growth in Real* Earnings per Share
- Dividends
- P/E* Expansion

(*"Real" means after-inflation; "P/E" means Price divided by Earnings, using our method of adjusting for inflation, and averaging Earnings over 10 years to smooth out accounting shenanigans.)

The earnings per share of a large basket of public companies grows at a real rate about 1-2% less than the growth rate of GDP. You would think that these companies would grow their collective earnings at about the same rate that the whole economy grows. They do actually come close, but that's on an aggregate basis. Your portfolio of stocks only goes up if the earnings rise per share -- and big companies relentlessly dilute their shareholders with new share creation. Since 1945, real EPS grew at 1.7%; real GDP grew at 3.3%. Over longer periods, the gap is worse.

So, let's assume that real EPS grows at an optimistic 1% slower than GDP. In a near-perfect world, GDP grows at about 3 to 3.5%. Alas, the world is not perfect in the U.S. Working age population growth is slowing -- and that growth rate is an arithmetic addition to GDP growth. If you drop 1% off the population growth rate, you eventually reduce GDP growth by the same 1%. GDP's other arithmetic input is productivity growth (yes, those are the only two long-run inputs to GDP growth), and productivity growth is likely to be dampened as we work off our debt. So, we've got a double-layer wet blanket thrown over the economy.

I would be surprised to see real GDP grow by more than about 2.5% per annum over the next decade. That would mean real EPS growth in the 1-1.5% range. (And I don't mean growth from the temporary lows of today -- I mean growth from the line we were on before the boom and bust. Growth back up to that line is merely a recovery -- it's not really moving us forward.)

Add the current 2% dividend rate to that EPS growth figure and you get an expected return on stocks of 3 to 3.5% after inflation. And, in fact, that's the return we've achieved since the last time the market was priced like it is today. (See the blog below for our discussion of the price of the market.) This is the "baked in" return on stocks -- it will grind away in our favor regardless of what the market does year to year.

Now -- what about P/E expansion? Yes, every rise in the P/E ratio is additional returns to stock owners. If the P/E rises from 17 to 22 over a three year period, that's a bonus of 8.8% per year added to our 3-3.5% expectation. Good times indeed!

But, should you expect the P/E to expand from its historic levels in the mid-to-high teens? If it is low to start with, sure. But, if it is at or above reasonable long-term levels, no. Realize what you are expecting to happen if you expect to sell your stocks to somebody else at a higher P/E in the future. You bought stocks expecting a return of 3-3.5%, and you are hoping someone else will come along expecting something lower and buy your stocks from you. Remember, the 3-3.5% expected return is baked in and doesn't change with the market. The more someone pays for stocks, the more they are eroding their ability to actually capture that 3-3.5%.

P/E expansions above trend are only sustained by the speculative hope that someone will come along and pay more. They are not sustained by fundamentals.

We expect stocks to pay 3-3.5% real return over the coming 7-10 years -- we are not going to hang our retirement hopes on P/E expansion. That is not a terrible return, but it does give us pause in going back to full stock allocations. It's because stocks are not a take-it-or-leave-it choice. They are an either-or choice, and we have other assets to choose from that offer similar or higher expected returns.

More on those choices in the next installment...