From Research Affiliates:
“What is often lost in the conversation of the right level of CAPE is an appreciation for expectations of return in the absence of any mean reversion. Real EPS trend growth since 1871 has been 1.5%. From all-time peak earnings, as a share of GDP, dare we expect more growth than this? With a dividend yield of 2.0%, this gives us a real return (yield plus growth) of 3.5%, if valuation levels 10 years hence are exactly where they are today. From current valuation levels, dare we expect PE expansion? If not, the maximum likely return over the next decade is 3.5%. Mean reversion to a value of 23 would deliver a scant return of 30 bps a year, whereas reversion to the historical average CAPE ratio of 16.6 would result in a loss of −2.8% a year; both scenarios are net of inflation, but include the positive impact of dividends. Returning to the median valuation level since 1990 would take us to a near-zero real return.
To earn an annualized 5% real return over the next 10 years with a 2% dividend yield, we’ll need 3% real share-price growth. If half the 3% growth comes from earnings growth (matching the trend growth rate since 1871) and half from PE expansion, the CAPE ratio a decade from now will need to be 37. When considering how to invest, we should ask ourselves: Are we comfortable with these heroic assumptions? Or do we want to invest in markets where sensible returns can be expected, based on sensible assumptions about future growth and mean reversion?”