So far this month, the FTSE 100 has risen 2.7%. There's a link between this fact and Emily Thornberry's resignation for that tweet.

The Footsie's gain is consistent with one of the strongest tendencies in finance - for Halloween to be a buy signal. This has an important and under-appreciated implication. It implies that, underneath their smart suits and brass name-plates, professional investors are in fact just like our pagan ancestors. They get gloomy around Samhain thus pushing share prices too low, and cheerful around May Day, which sends prices too high: the counterpart of "buy on Halloween" is "sell in May."

This, of course, is not the only evidence that the so-called experts aren't as smart as they think they are. It's well-known that most fund managers who try to pick stocks under-perform (pdf) the market. And new research shows that, in their personal investments, professional investors do no better than amateur ones. Insofar as professionals do make money, it is because they benefit from social contacts and seeing order flow, rather than because of other expertise. One reason why bankers tried to rig Libor and the FX markets is that it's so damned difficult to make money honestly.

In this, of course, finance professionals aren't wholly different from other so-called experts. As Philip Tetlock and Paul Meehl (pdf) showed, political scientists and clinicians aren't so hot either.

There is, however, one group of people who do have insight into the future - consumers. Sydney Ludvigson and Martin Lettau have shown that aggregate consumer spending can predict (pdf) US equity returns - a finding corroborated by Bank of England research in the UK.

The success of consumers, in contrast to the indifference of experts, is consistent with the diversity trumps ability theorem: in conditions of bounded knowledge and rationality (or in other words, the real world), the dispersed and diverse knowledge of laymen can do better than experts.

Which brings me to Ms Thornberry. What made her tweet so embarrassing is that it highlights the managerialist Labour party's awkward relationship with the white working class. Its attitude has often been one of contempt for it - the subtext of that tweet - interspersed with a claim to "listen to its concerns" (if only on immigration.).

There is, however, a third way here. It lies in that diversity trumps ability theorem. This tells us that, under some conditions, the working class has dispersed, fragmentary, tacit knowledge that managerialists do not possess.

I say this is a third way because it requires two changes in the current relationship between Labour's leadership and the working class.

On the one hand, it means less pandering to beliefs that are reified nonsense. As Gavin Jackson shows, attitudes to immigration are more positive among people who have hard knowledge of immigrants living in their areas. It is the local experts on immigration that should inform policy, not the pub gobshite rehashing Daily Mail drivel.

On the other hand, though, it requires Labour to devise institutions which harness the dispersed, diverse knowledge that workers have - that is, to create worker democracy in both public and private sectors. This means Labour must respect the working class more and the managerialist pseudo-elite less.

The chances of either of these changes occurring is, however, slim. This is because the managerialist Labour party isn't merely out of touch with workers, but out of touch with economic reality too.