Free Lunch: Problems with finance

Δημοσιεύθηκε: 8 Μαΐου 2015 - 10:56

It's not exactly economics, but we can't resist digging into the UK election numbers. From most of the coverage, would you realise Labour actually increased its vote share, and by more than the Tories? Scroll down to the numbers section for details.

The uselessness of finance

The Economist's Buttonwood has written a long and probing essay on the problems with finance, which links financiers' record of causing the crisis with economists' failure to understand it ahead of time. Because academic economics did not get how financial markets worked - or even cared sufficiently - financiers relied on models that missed important bits of reality.

Buttonwood rightly spends a lot of time discussing the brave new world of behavioural economics, which rejects the unrealistic assumptions of perfect foresight and unwavering rationality built into conventional economic models. But the essay makes a little too much of the idea that this particular failure is main problem with finance theory and that adopting more behavioural models is the way to fix it.

The most interesting thing about some of the other recent research papers Buttonwood cites, which show (too late for this crisis but hopefully not for the next) why finance can wreck economies, is that they do so without resorting to behavioural economics. Among these is a Bank of International Settlements article that Free Lunch discussed when it was published. Its authors documented that a larger finance sector holds back growth rather than boost it. The reasons have little to do with behavioural tics and all to do with standard economic reasoning about risk management: finance favours activities that can pledge physical collateral - think construction - and those activities have notoriously low productivity growth.

Or take the infamous cop-out by Goldman Sachs chief finance officer David Viniar about "25-standard deviation moves, several days in a row", which Buttonwood also mentions. In the models Viniar and his colleagues used, the standard deviation (a measure of how things move randomly from day to day) is narrow enough that a 25-fold could virtually never occur. But that just means the models were wrong. Is that because they forgot that people are irrational? Not really. The always illuminating Andrew Haldane recently showed that taking proper account of the complexity with which financial markets and economies interact would lead one to expect "fatter probability tails" - wider standard deviations and so more frequent large swings - than simpler models.

All credit to Buttonwood's call for incorporating more psychological realism in financial models. But it would be an improvement just to do conventional modelling properly for a start.

Kicking Europe's bank habit

Apart from the general problems with financial theory, there are specific problems with financial practice. One of the biggest is in Europe: the region's extraordinarily high reliance on bank lending to finance economic activity. The poor state of alternative financial channels (such as equity and bonds) is a reason both why governments felt obliged to bail banks out (wrongly, as it brought several countries to their knees) and why the eurozone's credit crunch lasted more than six years and is only just letting up. The chart belows shows that the proportion of eurozone banks tightening their lending criteria for business customers outweighed those easing them until last year.

There has been a lot of change in European finance since the crisis, the biggest achievement being the commitment to progress towards a banking union. A new essay by Nicolas Veron is a readable, in-depth guide to the economics and politics of banking union. In Veron's analysis, banking union is a truly radical step: he argues that it made it possible for the European Central Bank to stabilise sovereign debt markets by promising to buy distressed bonds in a crisis. He also credits it with shifting the eurozone from a politics of bailouts to one of bail-ins, thereby creating a truly unified market where banks are no longer tied to the fortunes of the governments within whose jurisdictions they find themselves. (You might think Greece is the right place to demonstrate that this government-bank death spiral has been broken. Instead the ECB seems determined to chain the fate of Greece's banking system tightly to that of its state.)

But too much remains the same, in particular the dominance of bank loans over other forms of finance. EU leaders have belatedly recognised with their recent plans to create more harmonised capital markets. Until a "capital markets union" is in place, an important short-term step would be to package more loans, especially those to small and medium enterprises, into marketable securities. A blog post from the IMF makes the incontrovertible case for such securitisation and explains how it could be done. It points out that only 5 per cent of SME loans in the biggest eurozone economies are currently securitised, although the default rate on such securities is exceedingly low at 0.1 per cent.