The price-earnings ratio on non-financial shares is now just 9.9, some 40 per cent below its post-1988 average. This poses the question: why are shares so cheap? The answers are not as obvious as you might think.
The obvious answers is that there are huge risks facing the market, such as the danger of a continued recession - UK GDP could fall in Q2 - or of an escalation in the euro area's debt crisis.
These, though, aren't sufficient to explain why shares are cheap. Gilts, and some other government bonds, do well during euro area debt crises or recessions. This means that equity risk arising from these sources can be diversified by holding gilts, some overseas bonds (and perhaps gold too). Such risks cannot therefore be a good explanation for why investors are avoiding shares.
Put it this way: in the last five years, the correlation between monthly returns on gilts and the All-share index has been negative (-0.24). In the 1990s it was positive and often very significantly so. Shares should therefore be more highly valued now than they were then, because their losses are easier to diversify way. But they are not. Something else, therefore, must explain why they are cheap.
One possibility is that the problem is not risk but growth. A good indicator of just how cheap shares are by historic standards is the gap between the dividend yield and the yields on longer-dated index-linked gilts. This is now over 3.5 percentage points, which is nearly as high as it was during the worst of the 2008-09 crisis. You can interpret this gap as being equal to the difference between the equity risk premium and expected long-term growth in dividends. The gap might have increased, therefore, not because risk has increased, but because growth expectations have fallen.
It could be that the crisis has permanently reduced our growth rate. One study by IMF economists has found that "economic contractions are not followed by offsetting fast recoveries. Trend output lost is not regained, on average….The output costs of political and financial crises are permanent on average." And Carmen Reinhart and Ken Rogoff, in their history of financial crises, This Time is Different, found that, after the Great Depression in the 1930s, it took ten years for real GDP per person to return to its pre-crisis level.
One fact corroborates these fears - that labour productivity has stopped growing. Output per worker-hour is lower now than it was in 2007; in normal times it should have risen more than 10 per cent. This matters because productivity growth should be, in the long-run, the main determinant of growth in GDP and dividends. If our productivity slowdown indicates genuine supply-side problems - rather than temporary labour hoarding in the face of weak demand - then we should expect slower growth in future. This would justify a higher dividend yield.
Slower productivity growth matters in another way. It can cause stagflation, the nasty combination of slower growth and higher inflation. This is simply because slow productivity growth, unless fully offset by lower wage growth (which brings its own problems) means higher growth in unit wage costs, to which firms respond either by trying to raise prices or reducing output. A recent paper by Norbert Berthold and Klaus Grundler, two economists at the University of Wurzburg shows that stagflation due to productivity slowdowns is still a threat.
And this matters for shares. Stagflation is very different from recession. In recessions, bond prices rise as investors seek a safe haven. But in stagflations, their prices fall as investors worry about inflation. Whereas recession is a diversifiable risk - as gilts insure us against it - stagflation is not. Now, we don't need a return of1970s-style double-digit inflation to do damage here. Even quite a small rise in inflation would be a shock to gilts. The small danger of this would thus justify a higher equity risk premium and thus a higher dividend yield.
In fact, though, stagflation is only one of a set of dangers which we can think of collectively as correlation risk. Although gilts and equities have recently been negatively correlated, they need not remain so. If the correlation turns positive, shares would become riskier in the important sense that it would be harder to insure against their losses. The danger of this happening is another justification for a high risk premium now.
As for why the correlation might become positive, there are at least three possibilities.
One would simply be if or when the US's super-loose monetary policy ends or when investors anticipate and end to quantitative easing in the UK. As central banks stop buying bonds, their prices would fall. And shares might take more fright from the ending of cheap money than they draw comfort from the better economic conditions that cause that ending.
A second possibility would be if or when Chinese export-led growth slows, or growth becomes more domestically-oriented. For years, the counterpart of China's trade surplus has been its buying of US Treasury bonds. When the surplus declines, so too might this buying. This would be bad for western government bonds generally.
There's a third, more dramatic possibility - that the US might suffer its own debt crisis. The Congressional Budget Office has warned that government borrowing might enter an "explosive path". If investors fear even the small danger of this, bond prices would fall. The accompanying uncertainty and the fear of emergency austerity measures would hit shares too.
Of course, there is no need at all for the US to suffer a full-blown Greek-style crisis, simply because the Federal Reserve can print as much money as necessary to buy Treasuries. For our purposes, though, this is little comfort. It merely converts the debt crisis into an inflation threat, which would hurt bonds and equities together.
You might think the risks of long-term economic stagnation, chronic stagflation or a US debt crisis are small. They are. But the small possibility of total disaster is sufficient to keep share prices low – especially because many small possibilities amount to quite a large possibility.
And let's be clear. If the market is rational – which is an if! – it is risks such as these, and not just the more obvious ones, which really justify why shares seem cheap.
MORE FROM CHRIS DILLOW...
Chris blogs at http://stumblingandmumbling.typepad.com