When I worked as a sell side strategist, literally back in the last century, I was once advised by my then head of sales that there was no point in writing anything any later in December than the Varsity Rugby match, on the basis that ’nobody would be in the office to read it’.
Unfortunately nobody explained this to the head of research, so we duly toiled on, producing vast tomes of economic and market predictions for the year ahead, working to crazy deadlines, liaising with city printers until the final delivery of hundreds of boxes of hard copy Almanacs on around Christmas eve. The sales desk meanwhile were busy entertaining the very clients these notes were intended for, organising Christmas lunches, quizzes and trips to the Varsity Match. I know who got the better end of that deal.
The Varsity Match was usually around December 12th, meaning an alternative 12 days of Christmas and I have often thought that the Christmas rituals of that song were reminiscent of the attempts by strategists and economists to woo their clients. “On the first day of Christmas, my broker gave to me…a forecast of US GDP” and so on.
Now that I suppose I am that client, I find that, while sadly there are no Christmas broker lunches allowed anymore and hard copy publications are all but gone, there is nevertheless the ritual of ‘The Forecast’, and without being too harsh on my broker friends, much of the forecasting ritual can in my personal view be as irrelevant as the proverbial ‘partridge in a pear tree’.
The process tends to start with the big number, GDP, something I have often described as a lot of interesting small numbers aggregated up into one great big useless number. Certainly as an equity investor I can’t find any good stock or sector thematic from the prediction that US GDP growth will be 2.3% or 3.2, or indeed (from last year’s big ‘call’) that Chinese GDP growth would be 6.4% or 4.6%.
Discussing whether Chinese retail sales would be 11% or 12% might add some value, not so much for the accuracy of the prediction as from the acknowledgment that the mechanics of the equation, GDP = Consumption + Investment + Government spending + Exports minus Imports mean that the sub-components can and do move in different directions and that these – and the sub components of these sub components are where we need to look for insight.
The brokers with the highest GDP prediction traditionally were also the most bullish on equity markets and usually favoured cyclical stocks, while the most bearish tended to be bond houses who always saw the safety in bonds and an imminent crash in equity markets, often described in subtle language, like ‘Armageddon’. I used to regard this with some cynicism, but increasingly I realise that this is just what behavioural finance refers to as confirmation bias; the equity bulls selected out the positive news flow and the bond bulls the negative.
On the second day of Christmas, the GDP forecast is then usually linked to an inflation prediction, broadly reflecting the notion of an ‘output gap’. If GDP as predicted is above what is regarded as trend, then the assumption is that there will be less spare capacity in the economy (the output gap will close) and that prices will rise and vice versa.
This is of course to simplify enormously, but it does expose the key assumptions behind the modelling, that we have an understanding of what the trend growth in an economy actually is, how much spare capacity exists within that economy and how much of our predicted demand increase will affect that.
Far more important however, and similar to the first point on the demand prediction is the fact that a Consumer Price Index is an aggregate of lots and lots of useful pieces of information about pricing power, aggregated up into one big largely useless inflation number.
I think that this big figure can be largely useless from an equity perspective, but it is key for a bond investor, because the most important forecast for them is not what the economy will do, but what the monetary authorities, the Federal Reserve (Fed), will do. Thus on the third day of Christmas we tend to get the predictions of what the policy makers will do to interest rates.
Back in the late 1980s and early 1990s we had a pretty good notion of the models that the Fed were using to set rates. Essentially a variation on the notion of an ‘output gap’ outlined above, stronger GDP tends to lead to tighter monetary policy.
This is the logic behind the market obsession throughout the year with the US non-farm payrolls. This high frequency data is deemed a proxy for changes in the output gap and thus a need for tighter monetary policy, hence short term market responses to this data release – despite the fact that the data is a) lagging anything we already know from the bottom up and b) frequently heavily revised.