In FT, Gavyn Davies has a piece in which he makes the case for equities for the long-term, from here on. He has a chart on the current after-tax profit as a % of GDP in the US. It is at the highest. Now, he considers the argument made by James Montier about the mean-reversion in profit margins and dismisses it in the light of the so-called age of globalisation of labour (or, factors of production, more broadly?), despite the fact that mean-reversion in margins is very strong. First, it is arguable that globalisation has more to go. It is probably slightly past its peak now.
Second, that was not the only point that James Montier made in his note. I think he also wrote – and correctly – that the reason why we have a strong private sector in the US is now precisely because the government has chosen to bail it out. Huge government debt and deficit and private sector cash mountain are two sides of the same coin.
So, two questions arise : (1) What happens when the government has to start to tighten the belt, say in 2013? Will a recession ensue then, if it has not already started in 2012? (2) What are the distributional consequences of government largesse for the private sector? After all, profit margins are high because the nominal wage growth for non-supervisory workers is 1.6% – well below the rate of inflation. Does it really augur well for healthy and sustainable demand growth in the US?
Two more points about the misleading comparison of equity returns vs. bond yields. One nuanced point is made by John Hussman:
Among the problems with these typical approaches is that stocks are not 3-month or 10-year instruments, but have a duration that is essentially the inverse of the dividend yield (so at present, the duration of stocks is roughly 50 years, compared with a 10-year Treasury, which has a duration closer to 7 years). So the appropriate “risk free” return in these estimates should really be either a Treasury yield of equivalent maturity – none which are available, or at least an estimate of the average expected short-term risk free rate expected over the same horizon. Needless to say, estimates of the equity risk premium get a false benefit if you use today’s unusually suppressed, short-duration risk-free rates.
The larger difficulty is the estimate of the prospective return on stocks. If you want to use a 10-year Treasury yield as a benchmark, you would also want to use a 10-year projected return for the S&P 500. On that note, and using reliable valuation methods (see above), the difference between the expected 10-year total return on stocks and the 10-year Treasury yield is presently less than 2% (nominal). [Link to his full Weekly Market Comment of 26th March 2012]
Of course, ‘see above’ is possible in this blog since he is referring to his weekly comment (link provided above). There, he shares with us information on the model that forecasts S&P 500 stock returns at 4% per annum, going forward. [GMO in his monthly asset-class return forecasts