The population is ageing but that is just one influence on asset markets
By Buttonwood
MOST people are aware that the western world is ageing; that we are living longer, the baby boomers are now retiring, and, thanks to low fertility rates, the supply of future workers will be restricted. All this seems likely to have an effect on financial markets; the classic "lifestyle" model predicts that people will borrow in their 20s and 30s, save a lot in their 40s and 50s and start running down their savings in their 60s. This should affect the demand for financial assets, and thus their price.
That was the conclusion of the 2005 Barclays equity-gilt study which used population predictions to assume that bond yields would rise by around five percentage points per decade, making them around 9% by now and 14% by 2025. In 2010, the same study updated its estimates and forecast that 20-year yields would be 10% by 2020; by 2011, the target for 2020 was just 7%. Both of these forecasts might still come true, but the projections indicated in the graphs suggested 10-year yields of 5% or so by now.
So what went wrong? There has been no great demographic surprise. Instead other factors have weighed on bond yields; the aftermath of the debt crisis, quantitative easing, the euro zone's problems, falling inflation and so on. But there is a broader problem that touches on last week's column on the fallibility of financial studies. The 2011 study had a model that perfectly fitted past data, and so must be assumed to have a great forecasting record. But when we construct models, we know what happened in the past; it guides our judgments. But future changes are, by definition, impossible to fit into the model; it is a bit like the man who fell off a skyscaper, remarking as he passed the 23rd floor "All right so far".
Barclays should certainly be commended for giving the exercise a try; its equity-gilt study is one the more thougthful pieces of analysis that appears each year. Your blogger first recalls attending its launch in 1993 when Michael Hughes (a charming chap with a splendid military moustache) forecast that the FTSE 100 would reach 7,000 by the year 2000. In retrospect, one can either be impressed that he was just 70 points out (the index closed at 6,930 on the last trading day of 1999) or struck by the index's failure to reach that benchmark 22 years later.
This year's study has once again returned to demography and once again sees a turning point; the global savings glut is disappearing. China, a key supplier of savings, is aging too. That should mean higher real rates; a 1 percentage point share in the elderly's population share is associated with a 1.15 percentage point rise in the real rate. (Of course, this is backtesting again.) It should also mean lower equity valuations.
This time round, there is no bond yield forecast but there is the statement that
Demographics are thus likely to generate a strong secular headwind for asset prices in the coming decades
That seems broadly right although the past 10 years have shown us that other factors can matter a lot more. Indeed the very low level of the Treasury bond yields and the fairly high level of US equity valuations tells us much the same thing. But these are long-term factors; they can't tell us about the next 6-12 months.
