My table shows correlations since January 1996 between annual price changes in some FTSE sectors and the annual change in manufacturing output (which is a measure of general variations in the economy). This shows that supposedly 'cyclical' sectors such as industrials and support services are no better correlated with the state of the economy than banks or general financials, and are less well correlated with it than the FTSE All-Share index. For example, a one percentage point annual change in manufacturing output has been associated, on average, with a 3.9 per cent return on banking stocks but only a 2.9 per cent return on support services.
| Sensitivities to manufacturing output | ||
|---|---|---|
| Correlation | Response to 1% change in output | |
| All-Share index | 0.62 | 2.9 |
| General finance | 0.59 | 4.6 |
| Chemicals | 0.55 | 4.0 |
| Industrials | 0.53 | 3.4 |
| Banks | 0.50 | 3.9 |
| Support services | 0.47 | 2.9 |
| Mining | 0.25 | 4.2 |
| Based on annual changes since 1996 |
This suggests that, if you're looking for a 'cyclical' play, you're better off with a tracker fund than with many industrial stocks.
There are three reasons for this.
First, sectors – and even more so individual stocks – are subject to idiosyncratic risk. They can do well in bad times and badly in good. Such idiosyncratic moves naturally reduce correlations between shares and any systematic factor, such as the state of the economy. A classic study of the 1990s recession by Paul Gregg of the University of Bath and the late Paul Geroski found that recessions hit an unpredictable minority of companies very hard, while many actually thrived. It's reasonable to suppose that a similar thing is true of upturns.
Secondly, investors sometimes see economic upturns coming, and so buy 'cyclicals' in advance. This reduces the contemporaneous correlation between them and macroeconomic activity. (They rarely see recession coming, but that's another can of worms.)
However, even if investors do see upturns coming, there's one thing they tend not to anticipate – that such upturns will increase their appetite for risk; people are bad at predicting their future tastes, as they tend to overestimate the extent to which they'll resemble their current ones. And this gives us a third reason for the pattern we see in my table. Economic upturns – and remember that UK manufacturing output is highly correlated with world production – increase investors' appetite for risk. This leads them to buy shares generally, and especially higher-beta ones such as banks and other financials. It's for this reason that the sectors least correlated with manufacturing output tend to be defensive ones such as tobacco, food producers and food retailers.
The message here is simple. If you expect the economy to strengthen over the next few months, perhaps you shouldn't rearrange your equity holdings very much. Even if you're not holding many 'cyclicals', you might well have more cyclical exposure than you might think.
Another thing: I hate the words 'cyclical' and 'cycle'. They imply that there is a regular, predictable periodicity to economic fluctuations. This is false, and dangerously so.
