Permanent Value

Week in Review 8/26/13

Bruce Doole
August 26th, 2013

Economist Darrell Spence discusses his outlook for the U.S. economy and financial markets

The past few days have raised uncertainty about the direction of U.S. markets. What we want to do today is move away from the short-term volatility of the markets and look at what we consider to be the fundamentals — the building blocks that drive asset returns over long periods of time — and see if they can tell us anything about how markets are positioned and what they are likely to return, not just over the next quarter or the next year, but over the long term, for the next decade or so. We are going to do this for three major asset classes: government bonds, corporate bonds and U.S. equities.

Starting with government bonds, creating a forecast for a 10-year return on a 10-year bond is fairly straightforward. History suggests that the likely return over a 10-year period is roughly equal to the coupon at the beginning of that 10-year period. For U.S. Treasuries, up until recently, that coupon was about 2.2%. So that would be my 10-year return forecast for U.S. Treasuries. For U.S. corporate bonds, I have 3.5%. They yield a little bit more, but they should because they’re a riskier asset class than Treasury bonds.

Building a 10-year forecast for the Standard & Poor’s 500 Composite Index is a little bit more complex because there are more moving parts that determine equity market returns over time. There are earnings growth, valuation and dividend yield. So we’re going to take a look at each of those components individually to try to come up with an assumption about what the next 10 years or so might look like.

Nominal GDP as the constraint on earnings over long periods

Starting with earnings growth, I have an assumption that earnings grow 5% a year on average over the next 10 years, which might seem a little bit low. But there is a good reason for assuming 5%, and that is because over long periods of time there is a constraint on how fast earnings can grow in the United States and that constraint is the growth of the economy or nominal GDP. Nominal GDP and earnings have a close correlation over time. Obviously, earnings are a lot more volatile than growth in the overall economy but, at the end of the day, earnings and the size of the economy tend to grow at the same rate.

We also need to make an assumption about an appropriate valuation for the market. My assumption right now is 17.4 times earnings. The way I came up with that number is by looking at the historical relationship between the price-to-earnings ratio on the S&P 500 and the 10-year Treasury bond yield.

Not surprisingly, when you’ve had very high levels of bond yields, you’ve tended to have very low P/E ratios, because that’s the way the discounting mechanism works. As bond yields have come down, P/E ratios have tended to expand, but that’s only true to a point. History suggests that when you’ve had periods of very low bond yields, you’ve actually also had fairly low P/E ratios. This is because very low bond yields tend to be consistent with periods of economic duress or deflationary risk.


Source: American Funds from Capital Group