Subject: Re: OT- Guy Spier on Podcast/ limited BRK discussion
If things remain much as they have been since 1995 in terms of growth of S&P 500 smoothed earnings and valuation levels, I'd expect something like inflation + 2.8%/year real total return from the S&P 500 in the next 7 years.
...
It may also be worth noting that, in the 25 years since 1998, both earnings growth and the P/E multiple are much higher than the historical average, so the condition that those 2 crucial factors both remain at their last-25-year levels is possibly a bit optimistic. If you used earnings growth rates and multiples from the previous 25 years, what would the annual return be, then? (I'm guessing it's negative')

Absolutely true.

Whenever I do this exercise it usually sounds like a permabear saying the same old same old: the market is overvalued, look out below.
So what I say in such a post is generally the most optimistic/bullish possible view of the data, to forestall the inevitable "it's different this time, you just don't get it" ripostes.
I quote the average valuation levels since 1995 (the expensive era), not the average since 1980 or 1960 or 1930. I keep tabs on all of those, too.

Valuations have been extraordinarily high in the last ~25 years compared to deeper history. Three or four separate bubbles are included the sample, depending on how you look at things. So thinking that this range is the "normal" range is certainly a risk. Yet I don't have a great reason to assert that valuation levels before 1995 have even more to say about the likely future levels...who knows?

There has been a long run trend of rising valuations, which means that the returns in any "average" period were in fact getting a tail wind that can not be extrapolated. However, at least this specific problem doesn't exist in the time frame I looked at, as there is no up-trend since 1995. The old tech bubble was long enough ago that it cancels out the recent high multiples.

The bigger problem, which you also note, is the rising profitability. Net profit margins are very much higher than used to be the case, a mix of lower share of GNP going to labour and (perhaps) lower share to tax. Those trends also can not be extrapolated. Really, my analysis should be done a second time with price-to-sales metrics rather than price-to-earnings metrics, since earnings ultimately can't rise faster than sales over the long run. The two separate results would give a broader insight into what constitutes "normal".

For those interested in the change in profitability: it always varies with the business cycle, but it seems there was also a big change in the range in the US.
Net corporate margins in the US averaged 6.12%/year from 1951 to 2003. Standard deviation 1.01%.
Net corporate margins in the US averaged 9.94%/year from 2004 to 2012. Standard deviation 1.05%, pretty much unchanged.
We don't know how long this will last, but for the last ~20 years each dollar of sales has been worth 62% more to shareholders than it used to be in the prior ~50 years.
We do know that profitability won't keep rising forever, as profits can't exceed sales. Businesses will forever have some expenses. It would make sense to assume that future real profit growth will be close to future real GDP growth, not equal to the recent rate of profit growth.

Jim