Good piece by Edward Chancellor in the current FT applauding — and taking semi-issue with — some recent papers (here and here) by Andrew Odlyzko. In short, Odlyzko argues bubbles are hard to detect, with the mega-railroad bubbles of the 19th venture eventually producing a decent return. Chancellor says we’re giving ourselves too easy a task if we punt on bubble detection and assume that such misallocations will somehow save the day, eventually.
Andrew Odlyzko of the University of Minnesota has produced a couple of fascinating new papers* on contemporary forecasting during the British railway booms of the 1830s and 1840s. Railways were first established in Britain in the 1820s. But it was only in the middle of the following decade that a great expansion took place. As railway shares soared, planned new investment in the railway network rose to around 8 per cent of GDP, or more than three times higher than expenditure on US fibre optics during the internet boom, according to Mr Odlyzko.
…Mr Odlyzko sees the eventual success of the 1830s mania as an “antidote to claims that bubbles are easy to detect or that all large and quick jumps in asset valuations are irrational”. This goes too far. It is true the money invested in railways at that time eventually produced a decent return. However, by 1836 railway shares were two standard deviations above their historic trend (GMO’s definition of a “bubble”). They later collapsed. This experience was repeated in 1845. The two Victorian railway manias support the argument that bubbles can be identified before they burst and that they should be avoided at all costs.
The extreme over-estimation of the potential market size for railways in the 1840s undermines the notion that financial markets digest all available information. Investors fall for “mythical numbers”, says Mr Odlyzko, who a decade ago pointed out the falsity of the claim that internet traffic was doubling every three months. In similar fashion, investors today accept uncritically the magical vision of a billion Chinese urban consumers.
The railway manias remind us of the pitfalls of forecasting models, however basic. In addition, there is the problem that prominent forecasters are generally in the pay of promoters. The Victorian traffic-takers turned out to be no more independent than Wall Street analysts of the internet age.
Yet investors need not despair. They should brace themselves against the relentless propaganda of promoters and their brokers. And bear in mind Mr Odlyzko’s wise advice. You don’t need any special skills or complicated models to detect a bubble. All that is required is “common sense, an ability to do simple arithmetic, and knowledge of a few basic facts about the economy".
To some degree Odlyzko and Chancellor are talkng past one another, of course. The former is concerned with the long-run effects of investment enthusiasms around new technlogies, while the latter is more concerned with the fate of the companies supplying the capital and making the investments. Just because something works out for society and the economy doesn’t mean it can’t crush the companies involved and cost investors oodles of money. At the same time, there is something to be said for animal spirits in providing the vomitous effluxes of dumb money that makes epochal socioeconomic change possible.