Monday 25 April 2011

Performativity, corporate governance, responsible investment etc

I've spent a couple of hours this weekend reading a few of the pieces in Do Economists Make Markets? which I bought last year but have only recently got around to looking at. It's broadly a collection of papers looking at the idea of perfomativity in respect of economics - how economic techniques not only capture market behaviour, but also influence it. The biggest claim in this area is in respect of option pricing where the proposition is that that economic models used to price options made the market behave the way the model predicted. Interesting stuff indeed.

It's less clear to me that there are necessarily lots of other instances where something has occurred that backs the claims of performativity as well as the option pricing example. Also I can't shake the feeling that perhaps there's actually not much new here. So some of what follows sounds a bit negative, but this is largely because I'm still thinking this stuff through

First I'll go back to home turf. There are quite a few studies that claim a correlation between pro-shareholder governance, or reforms to achieve it, and financial performance. I read something quite recently that suggested that there is an immediate uptick in share price at US firms where resolutions are passed introducing a pro-shareholder reform (I think the example was resolutions seeking to get rid of poison pills). It would be interesting to look at these kinds of studies over the long term because it may be that there is now a premium for pro-shareholder governance where there wasn't before. This might in turn suggest that the increasing adoption of a view amongst investors that corporate governance does matter to shareholders has led to prices reacting to governance-related information as if it does matter. But (making all the assumptions required above!) is this performativity? The option pricing example is much more detailed, and there was, of course, a specific formula that became accepted that particularly influenced the market. But in principle is there a difference?

Similarly, in the dot-com bubble there were some arguments deployed along the lines that 'traditional' models of valuing companies no longer applied, and historical PE ratios were no longer relevant. Of course we now know that some of those using these arguments to put very rosy valuations on companies (or IPOs) were essentially fraudsters. But there were plenty of true believers out there too, and for a while TMT stocks behaved as if they had a point. Now one might counter that the TMT bubble was short-lived so not really an example of an analytical approach changing the behaviour of a market. But in return a) you might still get away with arguing it as a weak form of performativity b) some of the TMT bubble era predictions about the impact of technology may yet come true (or at least have a bigger influence) and c) in the options pricing example there have been periods when markets haven't behaved like the model. Most importantly, if we can get away with a) then I'm not sure what is really new.

Having said all that there is something in here that rings true for the wrong reasons, which is probably best illustrated by a quick example. A few years ago I listened to an SRI fund manager give a presentation about how they constructed portfolios and the returns they generated as a result. I asked if they had looked at whether their approach had an impact in the other direction - were their views significant enough or shared widely enough that they could have a noticeable effect on prices. The answer was, as you might expect, no - they hadn't even looked at it - but in a way I wouldn't have found it surprising if they said yes.

What I mean by that is that it strikes me that many people in the responsible investment world are actually (whether fully aware of it or not) trying to make this happen. There is a view that current financial markets and the prices that emerge in them do not accurately capture everything that matters. That's why we see a huge emphasis on the disclosure of more information on social and environmental issues, and the development of 'enhanced' analysis that incorporates it into valuations. Rest assured this effort would not be expended if many of the people involved did not believe that, at the end of the process, markets would behave differently. ie they don't only think that prices would be more 'accurate' or whatever, they think they could be quite different, and markets might behave quite differently too. (Someone put it to me once that the objective seems to be to put up the cost of capital for 'irresponsible' companies.) At the moment a major effect on prices seems a long way off. But you can see that in an area like climate change, where there's a bit more of an overlap between people trying to amend the system to bring about change and those simply wanting to ensure the right information is available to assist price formation, that it could happen.

It's a funny turn of events. Performativity is quite an enjoyable idea to knock about because it suggests that economics can cross the boundary between describing and creating behaviour. And, if true, that's something that can potentially be used to whack neoliberals over the head with - you (at least in part) create the behaviours and relationships you say are already there and that you claim you are merely making explicit. (Hey, you could even say it's that favourite of conservatives, an unintended consequence). But at the same time we make this attack 'progressive' (sorry) types are trying to develop new analytical processes and models precisely because they expect it to change market behaviour.

PS. Albert Hirschman (yes, him again) pointed out that there's a less talked about mirror image of the unintended consequence - the unrealised expectation. In other words what is expected to result from a given reform does not occur. Sometimes I wonder whether the results of much of the activity undertaken in responsible investment might end up under this label.

No comments: