How far can we rely on commodity prices to predict future? SHARE WATCH Andrew Miller.
It is easy to forget that it was only a few months ago that still-plunging commodity prices were among the perceived harbingers of an impending global recession.
With this in mind, should the recent bounce in commodities be taken as a bullish signal for the global economy? Can changes in commodity prices tell us something useful about the prospects for global demand? Even the most reliable lead indicators must always be carefully placed in context amidst an ever-changing economic landscape. With commodity prices, such attempts to contextualise are complicated by the fact that supply can change as well as demand. The dramatic plunge in oil prices seen since the middle of 2014, in spite of fairly consistent trends in measured demand, is an example of this.
To this end, we conducted a regression of annual changes in commodity prices against annual global GDP growth and changes in the Fed's dollar index of major currencies. From this admittedly simple exercise, we can see that these two variables can only account for 70% of the total variation in commodity prices from 1992 to the present day, indicating that other unobserved factors such as supply effects account for a significant proportion of the variation.
Of course, isolating the influence of changes in supply on the price of a basket of commodities as diverse as the Bloomberg Commodities Index is fraught with difficulty given incomplete datasets. Instead, we chose to focus on the demand side of the equation.
To do this, we regressed weekly percentage changes in the Bloomberg Commodity Index against percentage changes in a number of market indicators potentially correlated with global economic demand, namely the Fed's dollar index of major currencies, the MSCI World Index, and the slope of the US Treasury yield curve (10-year Treasury yield minus two-year Treasury yield).
The premise here is that equity prices, long-term interest rates, and the dollar are likely to respond to investors' perceptions of global demand, and not so much to changes in commodity supply.
For example, when a change in the price of the Bloomberg Commodities Index is accompanied by a simultaneous change in the slope of the US Treasury yield curve, we assume that both are responding primarily to a common global demand factor. While it is debatable whether these market indicators actually reflect global aggregate demand in reality, it seems reasonable to assume they can provide some (albeit imperfect) information on the global economy.
We can then use the information from these historical relationships to see how much of the drop in commodity prices since summer 2014 we would have expected to observe, given the market's weakening perception of global economic conditions proxied by these variables. Roughly 40% of the decline in commodity prices in the summer of 2014 could be attributed to falling global demand. However, the rest of the fall is probably due to increased supply in the marketplace or variables unrelated to demand, highlighting the risks of leaning too heavily on commodity prices as an indicator of wider global economic health.
In summary, any piece of macroeconomic data must be interpreted in the context of existing economic conditions which may differ signifi-cantly from past episodes. Commodity prices are not exempt from this.
In fact, commodity prices are not necessarily a lead indicator for global growth since they are by nature simultaneously determined by the fundamental forces of both supply and demand. The commodity price plunge of 2014-2015 is a clear example of this.
Investors hoping to get a more accurate picture of global growth prospects should lean on other data, despite their infrequency or lag.
Andrew Miller, Barclays
|Printer friendly Cite/link Email Feedback|
|Publication:||The Journal (Newcastle, England)|
|Date:||May 9, 2016|
|Previous Article:||Players' know-how grows their success; TUC with Neil Foster SPORT.|
|Next Article:||Airline's half-year losses expected to total PS15m; FORECAST ROGER BAIRD firstname.lastname@example.org.|