Monday, June 18, 2012

Valuation Matters.... Equity vs Bonds Edition

I've shown that valuation matters numerous times over the years when it comes to long-term equity returns (see here, here, and here for some of my favorite examples). The below post uses the same concept in that it compares valuation (i.e. yields) with forward returns, but in this version we compare the relative performance of equities vs. bonds.

The first chart shows the factor that serves as our starting point for valuation... earnings yield of the S&P composite (i.e. the inverse of the P/E ratio) and the yield of the ten year Treasury bond going back 100 years. What we see is a relationship between the two starting about 50 years ago that was non-existent the previous 50 (and the recent divergence that is the widest in almost 40 years).

But the lack of a relationship from 1912-1962 doesn't mean it the relationship wasn't always important. The next chart outlines the forward ten year return differential (annualized) for each starting point against the starting excess yield (the equity earnings yield less the bond yield). Interesting to note that we can easily see the unwarranted excess return that equities saw over bonds starting in the 1980's (i.e. the equity bubble), that was given back over the past ten or so years.

To summarize the above, the next chart outlines the forward ten year return differential (annualized) for each starting point by "bucket" (note that at current valuations we just made it into the 5-7.5% bucket, hence the yellow highlight). The takeaway is that starting yield differentials matter... a lot. To be more specific, the current 5-7.5% bucket means that for every period over the past 100 years when the yield differential was between 5-7.5%, the average annualized ten year forward return differential was a bit more than 8% (8% over the current 1.5% ten year would be 9.5% absolute returns for equities).

While I refuse to state that returns will be anywhere near this 9.5%, by almost all measures stocks appear cheap on a relative basis to Treasury bonds. Unless earnings collapse back to a "normal" percent of the overall economic pie abruptly (definitely possible, but in my view not likely) or the economic pie contracts abruptly, stocks are going to outperform bonds over the next ten years.


  1. How about bonds dropping 10% in the next 10 years and stocks dropping only 0.5% ? Would still give you 9.5% outperformance over bonds, but would you be happy with it?

  2. Not mathematically possible to have negative nominal returns for a Treasury yielding 1.5% unless there is a default (insanely small possibility). The yield is what you get in nominal terms. See here:

    In real terms, that could (and will likely) be true, though negative equity returns from this starting point is very unlikely. I do agree that equities pose a real risk over the short run, but much less concerned over the longer run.