How Cheap is the Market? (corrected)

I have run a few models suggesting decent long-term returns, and might chose to share those results at some point, but in the meanwhile, I thought some perspectives on valuation and expected long-term returns in the stock market.

Much of the gyrations on Wall Street deal with predicting what's happening now or next year, whereas the true driver of long-term returns is valuation and long-term earnings growth. The problem is Wall Street values the market on a multiple of recent earnings, which are highly volatile. Take a look at the chart below:


The blue thin line and thick trendline are the trailing 12 month (TTM) earnings for companies that make up the S&P 500 index. The problem with the traditional way of valuing the market using a multiplier of TTM earnings (or worse, some future year's earnings) is that it indicates that the market is worth 46% less now that it was a few years months ago. Indeed, that's consistent with the sell-off we see in the market, but for a long-term investor, is a whole bunch of bull!!

The pink line and red trendline represent a 10-year moving average of the TTM earnings. Essentially you are smoothing historical earnings to make long-term decisions while smoothing cyclical variability. Voila, it's fairly smooth. For a long-term investor, the market value hasn't changed in the last few years months - only the price.

So what does it all mean? The next chart is the S&P 500 price history, and the trend of 12, 16 and 25 times the 10-year earnings average. Think of this as a low, median and high value. I have been working on charts since 1910, but I need to verify a few things, so I'll only present the last almost 15 years.


Notice that the big crash in 2000-01 did not make the market cheap. It just took the market from ridiculously overvalued to expensive. And that explains why we have had mediocre returns since that time. In contrast, valuations today appear to present a much rosier picture in the long-term.

That "in the long-term" is the key piece! Short-term, prices could ... heck, almost certainly will fall more. As I write this, the markets are in a 3-day rally, but I expect any rally will fizzle. How low could markets go? I don't know that there is a clear answer in the data.

I would normally be tempted to answer down to the level of 14x 10-year earnings. That would suggest about 630 - quite a haircut more! But given the low level of interest rates and poor outlook for sovereign debt, I'd imagine that level seems low. I intuitively think 16x earnings is a bit high for a market bottom, and my guess is that the market will probably end up bouncing between 750-800 before a gradual rebound.

But we don't know. And at present levels, we can reasonably expect a decent return. How much? Earnings have historically grown about 6% a year. The dividend yield is about 3.2%. If we assume no multiplier change, that's still over 9% a year - a lot better than treasuries or CDs!