Wednesday, 17 October 2012

More on Exchange Rates

Eric posted on Monday about the article in which he was extensively quoted. I am mostly in agreement, with the article and Eric’s quoted comments, but there are a couple of places where I take issue.

First, the article takes as a given that the New Zealand dollar is overvalued. Now, this mantra is so commonly stated that one can hardly criticise a journalist for taking it as a received fact. Indeed, the article is able to reference a fairly high authority on this:
And the International Monetary Fund says the kiwi is about 15 per cent overvalued. It has said that repeatedly, at least as far back as May 2010.
Now staffers at the IMF are not stupid, and there is probably a well-defined question to which “overvalued by 15%” is the answer, but assessing that answer would require knowledge of what the underlying question is. The first thing I tell my 2nd-year micro students each year, that to start a piece of economic analysis with a price is to abdicate one’s responsibility as an economist. It is akin to a psychologist explaining some strange observed piece of human behaviour by saying “that is what they wanted to do”.

Let’s revisit some exchange rate basics. We have a clean floating currency. That means that the price of the currency is set at whatever level enables all who want to buy and all who want to sell at that price to do so. At first pass, that sounds like a perfectly valued exchange rate. Furthermore, the complaints about the kiwi being “overvalued” have been consistently heard for about a decade at least, so it would be hard to say that it is the result of a speculative bubble with speculators consistently entering on the demand but not the supply side.

So why is the dollar at the level that it is? One possible reason in recent years is the high demand for milk solids from China creating a boom for New Zealand dairy exports. Yes, this is not so good for manufacturing exporters, but the technical term for such a change is “the terms of trade going in our favour”. Alternatively, we can point to investment opportunities in New Zealand consistently being higher than the available pool of saving, coupled with a large pool of savings from China being available to prevent interest rates rising to choke off that investment. Again, a big increase in supply of foreign capital when NZ is a net buyer counts as a favourable terms of trade shock.

Pretty much the only way I can make sense of the “NZ dollar is overvalued” mantra is as an extension of the argument that New Zealanders are not saving enough, in the same sense that we don’t eat enough fruit and vegetables, drink too much alcohol, don’t attend enough classical music concerts, and have too much sex during rugby world cups.

But, and this is my second issue with the original article,  Eric is coming perilously close to this view when he is quoted as saying,
Fundamentally, New Zealand has a high exchange rate because we're an attractive place for foreign investors to put their money....Our relative lack of domestic savings in things other than housing means that the returns on other kinds of investment here are relatively high.
Hmmm. The fundamental problem with housing in New Zealand is land-use regulations reducing the extent to which savings can be directed into the creation of new housing stock. As I have noted before, the purchasing of existing houses is neither investment nor saving in aggregate, and so cannot explain the net demand for foreign capital.


  1. Well said. Given we have a floating exchange rate I have never understood what an overvalued exchange rate is. My guess would be that those who argue for the exchange rate being too high work within a mercantilist framework for which exporting is good, importing bad and so anything that upsets exporters is bad. Even if you can give a sensible interpretation to an overvalued exchange rate it is at best a signal that something is wrong somewhere else in the economy. Trying to set the exchange rate at the "right" level just treats a symptom not a cause.

  2. I don't think the NZ Dollar is over or under valued; the price is what it is.

    Let's split housing into two components.

    First, there's the flow of housing services for which you pay a present discounted value when you buy a house. There, I'll agree entirely: housing is neither investment nor savings in aggregate. Because this is more expensive with the supply constraints, income effects make us poorer and we save less, making the domestic pool of savings smaller.

    Second, the expectation that future population growth outstrips growth in housing supply means that people are willing to pay more for the house than is justified by the current flow of housing services: the expected resale value of the house in the future is higher than the current price after netting out maintenance and other costs over time. That bids up the current price of housing until the rate of return on asset price appreciation matches that in other investment markets. In equilibrium, this then can't matter; where supply constraints effectively become tighter over time and those were not anticipated, the move to equilibrium should pull money from other forms of investment.

  3. While considering my response I checked the most authoritative exchange rate source: The Big Mac Index. The most recent one ( has the NZD about 10% undervalued relative to the USD. This also doesn't account for the superior quality of the ingredients in a NZ Big Mac. That's it, case closed, the NZD isn't overvalued.

    Seriously though, with the state of the economies in other developed nations, surely NZ looks like a relatively attractive location to invest for foreigners. If this is the case then it should push the value of our dollar up through no indigenous action. If intervention can really be justified then it should be in the form of policy that would be a good idea regardless of the state of the global economy (such as saner land use rules).

  4. Perhaps of more concern are real exchange rate measures which suggest New Zealand's unit labour costs have risen faster than our trading partners (presumably largely due to relatively slower productivity growth in NZ than elsewhere). This loss of competitiveness must make it harder for our exporters to compete offshore.

  5. @GE.

    Unit labour costs will rise faster than our partners if the difference in nominal-wage growth minus productivity growth between us and our partners is greater than the nominal appreciation of the NZ dollar with respect to those partners. There is no reason to single out productivity growth as the causal determinant of that. More to the point, relative increases in unit labour costs are the mechanism by which resources are redirected in response to changes in the terms of trade driven by agricultural commodity prices or changes in the gap between domestic investment and domestic savings.

  6. Let's keep the stock of housing fixed, and assume an ongoing expected exogenous increase in the demand for housing. Will this lead to an ongoing reduction in the flow of savings. It might, but only if it leads to an ongoing increase in the flow of consumption. It might, but this is the part of the story I would find difficult to assert with any confidence. The story requires that the people selling their houses only because the speculators have bid up the price of houses increase their consumption because of the increase in their wealth, but there is no offsetting reduction in consumption by those who now have to pay more for the flow of housing services. I have a hard time seeing this.

  7. Fixed stock plus rising population means an increase in the intensity of housing use, which means a higher per unit cost of housing services. Assume no investors for now, only owner-residents, but that you might sell half your house to another family and then share the space. There'll be weirdness from that housing isn't perfectly granular, so somebody switching from a single family home to sharing a house with another family will see a discrete cut in total budget devoted to housing services, followed by a ratcheting back up of those costs as implicit rent increases. If we have housing as continuous rather than discrete, the housing share of income would ratchet up over time, with less then left for other forms of investment or savings.