jueves, 21 de septiembre de 2017

jueves, septiembre 21, 2017

Buttonwood

Analysts struggle to make accurate long-term market forecasts

Historically high valuations for equities complicate the task  
 
That is pretty clear with government bonds. Anyone buying a bond with a yield of 2% and holding it until maturity can expect, at best, that level of return (before inflation) and no more. (There is a small chance the government might default.) With equities, the calculations are not quite so hard-and-fast. Nevertheless, it is a good rule-of-thumb that buying shares with a low dividend yield, or on a high multiple of profits, is likely to lead to lower-than-normal returns.
So a sensible approach to long-term investing would assess the potential returns from asset classes, given their valuations and the fundamentals, and allocate assets accordingly. That is what GMO, a fund-management company, has been trying to do for decades. It has made some common-sense assumptions about the fundamental drivers of returns and then assumed that valuations would return to average levels over a seven-year period.
 
In one sense, this process has been a success. The assets that GMO thought would perform well have offered relatively high returns; the assets it thought would perform badly have offered low ones (see top chart). But if the ranking has been correct, the level of return has not been. Assets that GMO thought would yield a negative return of -10% to -8%, for instance, have in fact suffered average losses of only -2.8%.
 
GMO’s forecasts have been pretty accurate for asset classes such as emerging-market bonds and international (non-American) shares; annual returns have been within 1.5 percentage points of its forecasts. But for American equities, GMO was too gloomy, underestimating returns by around four percentage points a year.
 
The reason for this error is pretty clear. Equity valuations have not returned to the mean, as GMO thought they would, but have stayed consistently above their historical levels. GMO was fairly accurate in its forecast for dividend growth, but its erroneous estimation of valuation accounted for all the forecast error.
 
There are two possible conclusions. One is that GMO is simply wrong about mean reversion.

Equities have moved to a new, higher valuation level. This sounds uncomfortably like the famous quote from Irving Fisher, an economist, before the 1929 crash, that stocks had reached a “permanently high plateau”. But there is some justification for a valuation shift: American profits have been high, relative to GDP, for a long period of time. This may be a result of monopoly power in some industries, or perhaps of the reduced bargaining power of workers in an age of globalisation.

A more obvious argument is that, with yields on cash and government bonds so low, investors are willing to pay a high price for equities because they represent their only hope for decent returns. But given the low level of dividend yields and the sluggish rate of economic growth, profits will have to keep rising as a proportion of GDP to allow high equity returns to continue.
 
That seems unlikely.
 
Either there will be a political reaction—governments will clamp down on firms in response to public unrest—or, more prosaically, tighter labour markets mean that wage growth will start to erode profits.
 
Either way, it is understandable that GMO does not want to give up on the idea of mean reversion just yet. Its latest forecasts are pretty downbeat (see bottom chart). The real returns from most asset classes are expected to be negative; only emerging-market equities offer a decent return. Investors who disbelieve those forecasts are in essence betting that things will be “different this time”. That is certainly possible but it requires a lot of faith.

0 comments:

Publicar un comentario