Join our community of smart investors

Shares as harbingers

Currently, it is an especially important issue because many sensible people reckon global equity markets have collectively lost their marbles. The pace at which markets have recovered from the pit of despair in mid-March – especially in the US – pays little attention to what is happening in the real economy, they say.

The International Monetary Fund summed up this feeling pretty well late last month. “Investors are apparently betting on continued and unprecedented support by support central banks,” it said, adding that “markets appear to be expecting a quick V-shaped rebound in activity.”

The aim here is not to discuss the shape of a recovery, but to search for the signs that markets are any good at spotting when economies are about to fall into or recover from recession.

It is not in doubt that they should be able to do so in theory. Buy shares in a company and you buy its future not its past. So an estimate of that company’s future cash flows will drive its value today. Extend that logic to a whole equity market and you must conclude that its present value is the best guess of all its future cash flows discounted for uncertainty.

Since all quoted companies contribute such a large share of the real economy, this also means that changes in equity market values will lead changes in the real economy. Correspondingly, changes in the economy will lag equities.

This idea could be tested in several ways. For example, there is a notion that subtle shifts in employment presage economic downturn, so measuring stock market levels against changes in employment might be interesting. But the obvious comparison is between the stock market and the main measure of an economy’s output, its gross domestic product (GDP). True, GDP is a guesstimated measure subject to substantial revision and increasingly criticised for its inability to measure what we might label ‘non-monetary utility’. But it is probably the least poor comparator.

The table shows the findings for comparing GDP in both the UK and the US with equity markets in those two countries. Absence of data means the exercise for the US goes far back – to 1889 – but only an annual basis. For the UK, quarterly data for GDP was introduced in 1955, which is the start-point for the comparison with a spliced index mostly comprising the FTSE All-Share. The full data set is available on an Excel spreadsheet via the link below.

Since both economic output and stock market values in the developed world have risen much more than they have fallen, it seems futile searching the data for signs that equity market have led to economy upwards (such signs would be ubiquitous). Better to seek out the opposite, and less frequent, indicator – that the markets anticipate economic slowdown.

Who leads, who follows
 US GDPUS shares UK GDPUK shares
Number of years130130Number of quarters256256
Average change (%)3.56.5Average change (%)0.62.1
Variation around average (%)5.117.7Variation around average (%)1.010.0
Maximum gain (%)18.948.7Maximum gain (%)4.990.7
Biggest fall (%)-12.9-48.1Biggest fall (%)-2.8-28.1
No. yr-on-yr falls2743No. qtr-on-qtr falls4992
GDP falls yr after  US shares fall8naGDP falls qtr after UK shares fall23na
GDP falls same yr as US shares19naGDP falls same qtr as UK shares21na
GDP falls yr before US shares fall13naGDP falls qtr before UK shares fall 18na
Sources:; Bank of England: US Bureau of Economic Analysis

So we asked three questions of the data. First, how often does GDP fall after share prices have fallen (indicating the market does a good discounting job)? Second, how often have GDP and share prices fallen together (indicating no causal pattern)? Third, how often has GDP fallen before share prices (indicating equity markets do a poor job)?

For the market-discounting notion to hold good, we want to see many more years in the US and quarters in the UK when GDP falls after share prices have already fallen. However, the data does not give us that. From 130 years of US data, there have only been eight instances where a fall in equity markets was followed by lower GDP next year. These instances were outnumbered by the 13 times that share prices fell the year after GDP had already fallen. In 19 years both indicators fell together.

The UK data does a slightly better job supporting the discounting notion. From a bigger sample of 256 quarters, 23 times UK shares fell quarter on quarter, with GDP falling in the next quarter. In comparison, there were 18 instances of GDP falling first.

Yet the difference between the UK and the US data is not marked, which prompts the thought – perhaps even, the conclusion – that the timing of changes between equity markets and GDP is random. When one leads, it has no more significance than when the other leads. Perhaps that is what we should expect – the complexity of both stock market and economic machines is such that a neat correlation between the two is unrealistic.

But it is good to see the intuition backed by some fact. It also helps support the idea that today’s equity markets have not necessarily subsided into madness by being hopeful in the face of government-induced recessions the world over.