The falls are widely blamed upon uncertainty about when the Fed will start to withdraw, or 'taper', its quantitative easing. This raises two problems.
One is that we cannot know when or by how much the Fed will do this. This is because its decision will depend upon what happens to unemployment and inflation, but these are the unpredictable outcomes of complex emergent processes. They are the product of millions of decisions that haven't been taken yet, and which are taken for multifarious reasons that no single mind can anticipate.
Secondly, we don't know how shares would respond to tapering when it happens. This isn't just because it's hard to say what is already discounted. It's because the effect of tapering is ambiguous. Clearly, the withdrawal of monetary stimulus would hurt shares. But the circumstances in which the stimulus is withdrawn - a stronger economy - would benefit them. We can't say in advance what the net effect would be.
There's a lot, then, that we cannot know. But we do know three things.
First, we know the distribution of returns. This tells us that the falls we've seen aren't especially freakish. In the three weeks after it peaked on 22 May, the FTSE 100 fell 8.2 per cent. With annualised volatility of 20 per cent - the standard deviation of returns since 1900 - this is a 1.7 standard deviation event, which is a roughly 4.5 per cent chance. This might seem an outside chance. But over 52 weeks, it's very likely that we'd see a fall of this magnitude at some time.
Second, we know that the odds are stacked against equities at this time of year; the market does worse in the summer months than in the winter.
Third, we know that correlations between assets change over time. They are not hard, fixed, facts but rather the variable products of macroeconomic conditions. The fact that shares and bonds have fallen together might be unusual by recent standards, but it shouldn't have been wholly unexpected. Correlation risk matters, which is why investors should hold some cash even at nugatory interest rates.
Herein, though, lies a problem. What we know about the future seems only vague, whereas forecasts of when the Fed will taper seem more precise and vivid. And it's tempting to rely more upon detailed information than fuzzy ideas. This, though, is a mistake. Apparently precise forecasts offer only the illusion of knowledge and illusory comfort. It's better to have a rough knowledge of things that are true than a precise knowledge of things that aren't.
