Subject: OT: S&P 500
I have no idea whether broad market valuation levels will be high or low in the years to come. I wish I did.
Consequently, I have almost no idea what future returns from the S&P 500 will be, since changes in valuation multiples dominate the result.
But, I do keep a close eye on what the current valuation levels are like, and what they were in the past. No matter what conclusions you draw, it is a remarkable data series.
FWIW, valuation levels now, based on the multiple of smoothed real earnings, are pretty much exactly the same as seen on 2001-03-23
Forward returns from then, with dividends and after inflation:
1 year Inflation + 1.1% (valuation multiple contracted by -3.1%)
2 years Inflation -11.7%/year (valuation multiple contracted by -27.9%)
3 years Inflation -1.1%/year (valuation multiple contracted by -13.5%)
4 years Inflation + 0.7%/year (valuation multiple contracted by -12.8%)
5 years Inflation + 2.1%/year (valuation multiple contracted by -12.0%)
6 years Inflation + 3.2%/year (valuation multiple contracted by -11.6%)
7 years Inflation + 1.4%/year (valuation multiple contracted by -26.1%)
8 years Inflation -5.1%/year (valuation multiple contracted by -58.3%), note the end date : )
10 years Inflation + 0.8%/year (valuation multiple contracted by -31.7%)
12.5 years Inflation + 3.0%/year (valuation multiple contracted by -21.3%)
15 years Inflation + 4.0%/year (valuation multiple contracted by -18.6%)
20 years Inflation + 6.3%/year (valuation multiple EXPANDED by +16.5%)
On the other hand, one might see a stretch of continuing higher valuations.
Today's valuation level is also the same as it was 2016-12-16
Forward returns from that date:
1 year Inflation + 18.2% (valuation multiple expanded by 13.9%)
2 years Inflation + 7.0%/year (valuation multiple expanded by 5.3%)
3 years Inflation + 11.7%/year (valuation multiple expanded by 19.6%)
4 years Inflation + 12.8%/year (valuation multiple expanded by 29.4%)
5 years Inflation + 14.7%/year (valuation multiple expanded by 46.9%)
6 years Inflation + 7.4%/year (valuation multiple expanded by 5.6%)
The start dates above weren't hand picked to prove any point: they're about the only start dates in the last 25-or-so years with the same valuation level as now.
The market was cheaper than today for the whole of the stretch 2001-08-31 through 2018-07-27
That is, in any week during that 17 year stretch, you'd get more cyclically adjusted net earnings for your investing dollar than you would get today.
I'm not predicting a market crash--as I mentioned, I have no idea what future valuation multiples might be.
Though I suppose there is a simple subtext that valuation multiples are currently pretty high compared to history.
The current valuation level is more expensive than "the average in the last N years" for any integer of N in the range 7-106 inclusive.
Even if valuation levels remain unchanged, forward returns will necessarily be lower than the historical average because you're getting fewer earnings and dividends per dollar invested.
To get historically typical returns from here will require ongoing expansion of the valuation multiples.
Note, I don't assume anything about future earnings as a fraction of sales or GDP.
My smoothed earnings data series is just that: recent real earnings prior to a given date, with a smoothing function.
I use four smoothing methods which give similar results, and average the four results. One is the traditional E10 used in CAPE analyses.
Net profit margins have certainly gone up a lot in recent years, but that has now been going on so long that it mostly defines the new normal as far as my smoothing functions are concerned.
Jim