Spencer on LVR policy

Last week, Deputy Governor Grant Spencer gave what I consider the best explanation of the RBNZ's policy restricting the proportion of high loan-to-value ratio loans in a bank's new loan portfolio.

A lot of the Bank's prior talk on the issue has seemed incoherent. For instance, RBNZ has no mandate to implement policy to protect people against the possibility that their investments decrease in value, so talk of helping people by preventing them from buying assets that might reduce in value is a bit of a non-starter. Further, such arguments would require an assumption not only that the RBNZ knows more about the future housing price path than do property investors but also that they're unable credibly to communicate such information to prospective buyers.

But Spencer's address here makes what's likely the best case for LVR policy. Here's my interpretation of it. Again, this is not what Spencer says. Towards the end, it is nothing like what he says. I can even imagine him disagreeing with it. But it is the lens through which I interpret his statement.

We have a near-vertical supply curve in housing, mostly due to regulatory failures. That's looking to ease (he's optimistic) but it'll take time. Meanwhile, net migration numbers suggest there will continue to be substantial increases in housing demand. A demand curve that's shifting out coming up a near-vertical supply curve in the short run means continued substantial price increases. Where that feeds into rental prices, we get CPI inflation. And RBNZ has to keep CPI below 3. Further, at least some current demand is fuelled by that mortgages are currently really cheap. As soon as international interest rates start going up, anybody who took out a high LVR loan expecting low interest rates to continue forever is going to be in real trouble; resulting defaults could strain the broader system.

Everything's pretty second-best. First-best would involve getting rid of the supply bottlenecks, but that's just not going to happen. Even if you somehow got local government to stop the stupid, we'd face some very very real constraints in workers and kit available to build houses - it would take time (and price increases) to import more capacity to build houses. So LVR can mop up some of the effects of global loose credit conditions and avoid asset bubbles in housing so there's less down on the downside.

In short, we something like an Austrian-style malinvestment problem, though I'm never going to use either of those words in an RBNZ paper, and it's a hack version of Austrian stuff anyway - how the heck is a consumer good like housing a "lower order of production"? Anyway, loose credit globally makes NZ interest rates look pretty attractive, so lots of folks are willing to back NZ mortgages. And Kiwis are jumping into housing with that cheap money, expecting that even if interest rates go up, housing prices never go down, so things will work out. But when we do sort out the supply problem, housing prices could go down. And a sufficiently large disruption to international credit markets could also have them go down rather more sharply. Those have macroeconomic consequences. Inflation targeting isn't enough where we can get blow-ups in asset prices as well. Weren't any of you watching Greenspan? Sure, inflation rates looked fine in the States in the 2000s, but that's because it was all going into asset bubbles. The Austrians talked ad nauseum about this in the 2000s and nobody listened. Well, we're just not going to let that happen with NZ housing prices this go-round.

CPI targeting is great and we're going to stick with it. But we'd be pretty dumb not to be watching for asset price inflation at the same time. Sure, it'll feed into CPI via rents eventually, but why not nip it before it does that? Especially where there's some resulting bank balance sheet risk, and some risk that wealth effects after a housing crash would kill consumption and induce a recession. Our target isn't just CPI - we also have to watch for unnecessary instability in output. And asset prices are *in* the inflation part of the target! Go read the darned thing. 

LVR isn't much - it shaves a bit off the amount by which we'll need to haul up interest rates next year. But it could set back house price inflation for enough quarters that the supply response gets going. It could never have really long term effects. But it doesn't have to where the government seems to be getting serious about getting supply moving.

On this view, the systemic risk angle was only ever part of the rationale. If it were only systemic risk, something like enforced capital/equity buffers would make more sense as they'd be more neutral across risk categories. But where they're specifically worried about housing asset bubbles, and where they think they can identify those bubbles, we get LVR.

I still think it's a bit presumptuous to think that we can really tell that we're in a housing bubble. It's at least as plausible to me that the gap between rents and housing prices in Auckland is fueled not by any kind of bubble but rather by rational investor expectations of continued demand growth: they've capitalised the expected future increased in rental prices.

Look at the graph Donal put up on Tuesday:

The hand-drawn X is the updated net migration figure. Auckland rents are gonna go up. 

But I'm far less perplexed and angered by LVR under the interpretation above than I had been.

Update: on a related note, whether Treasury or the Government were right on the merits of Kiwisaver deposit subsidies and Welcome Home loans depends on the actual reason for LVR. If LVR were only about systemic risk, then the government's policy would have helped poorer cohorts meet the equity threshold (reducing risk), albeit at the cost of fuelling greater house price inflation. If LVR were meant also to be targeting asset price inflation, then the government was definitely wrong in putting in these measures. This suggests Treasury thought LVR wasn't just hitting systemic risk but was rather also aimed at housing costs generally. I think it's a bad idea to provide subsidies to homebuyers full stop, and doubly so when we're faced with a near-vertical supply curve. It just pushes up housing prices. But the ties to LVR noted in the article suggest something about the purpose of LVR.