Well again this depend on when you think the collusion takes place – maybe we have a situation where collusion holds more strongly when “demand is high” (ala Green and Porter), which would imply that the current situation is an example of tacit collusion.

Of course, this would imply that the structure of the market has changed substantially from 2003-05, when price wars appeared to occur during periods of high demand.

Matt Nolan :@steve
However, my point was that Westpac said it and other banks were setting price wrt an arbitrary price target above marginal cost – that sounds like tacit collusion.

Actually if you read the paper again it says just the opposite: through the worst of the credit crisis, the average cost was below the marginal cost. If there was ever a time for banks to collude and raise their prices based on whichever pricing model suited them best, that would have been it – and they didn’t. If anything we’ve disproved collusion. 🙂

Indeed, price can be above marginal cost in a cournot market – although in that case it is because of market power.

However, my point was that Westpac said it and other banks were setting price wrt an arbitrary price target above marginal cost – that sounds like tacit collusion.

Now I don’t actually believe that story – it was just able to be made given how it was framed in the document.

Econometrics is the only way to fight a recession – as it gives you something sufficiently time consuming to do until the recession naturally runs out of puff 🙂

On the collusion, it would have to be tacit – I just didn’t think that a discussion of short-run marginal costs was appropriate for discussing fixed mortgage rates.

Hi Michael,

Fundamentally I DO NOT believe that the banks are colluding. Even if there was some incentive to, I thought that the purpose of Kiwibank (and the significant government investment involved) was to make this type of collusion different. With Kiwibank not offering significantly lower rates I am not convinced in collusion at all.

However, I just found it a bit unusual that the document discussed short-run marginal costs, when expected long-run marginal costs should be the primary determinant of rates. The main purpose of this post was to illustrate why I thought that discussion was a bit inappropriate – I wasn’t really aiming to say there was collusion 😛

*visions of Matt dressed in tights and a big E on his chest, applying augmented Dicky Fuller to Godzilla*

Not to defend Westpac but marginal cost pricing for banks is a lot simpler to do than average cost given the myriad of funding sources. It is in fact an industry standard practice both here and overseas so if it is collusion then it is tacit.

Our point was that if banks were strictly using MC pricing then we would have seen even less passthrough from the OCR to lending rates in late 2008, when funding was at its most expensive (remember that back in the early days of the easing cycle the RBNZ didn’t seem all that confident that OCR cuts would be passed on). Instead, due to AC pricing (perhaps should be smoothed pricing?), we saw a high degree of passthrough at first and less so in recent months.

Michael Gordon
Markets Economist, Westpac

Just to be clear here, P is a function of time and so P=AC at time t won’t imply zero profits given that P is different in other time periods.