Fixed and floating, monetary policy, and structural imbalances
Over at the Rates Blog Andrew Coleman has written a piece on fixed mortgage rates and the monetary policy transition mechanism.
His initial point is that the ability of the Reserve Bank to smooth economic activity has been constrained by the fact that so many New Zealander’s hand fixed rate mortgages, compared to Australia where the majority of mortgage holders were on floating rates.
While his point holds weight, given that New Zealand is a net debtor country and higher average interest rates imply that the “increase in income” associated with a fall in the OCR was smaller.
However, there are three comments I would need clarification on before deriding the effectiveness of NZ monetary easing.
- My impression was that the target of monetary policy was regarding relative prices. We want to make “additional” consumption more attractive now, and any “additional” savings less so. The fact that fixed mortgage rates are fixed doesn’t directly change this – a change in the OCR will be just as appropriate at the margin. The income loss is unfortunate, but is the result of the risk-return trade-off choice made by individuals.
- The net debt position implies that the negative income effect needs to be offset by a positive income impact somewhere else. As the long-term fixed rates were with banks (not international investors, who were generally borrowed off based on a shorter maturity) does this imply that bank profitability is increasing? If so this would only be a negative income shock for us insofar as dividend payments would head over to Aussie right?
- It is more important to understand “why” NZ felt it was appropriate to fix itself to high mortgage rates to buy over-priced and overly large houses.
Moving past his initial point he does go on to cover the second issue nicely:
Since 1993, New Zealand 90 day interest rates have been at least one percent higher than five year interest rates some 37 percent of the time – and in the last five years a staggering 70 percent of the time. Other central banks seem fortunate enough to control inflation without needing to do this. Since 1993, Australian 90 day bank bill rates have been one percent higher than five year government bond rates only 5 percent of the time, and such a margin is scarcely ever seen in the United States.
It is unclear why such high short term rates have been necessary to counter inflation in New Zealand for such a long period of time, for inflation outcomes have not been noticeably different to those experienced in other developed countries. It is possible that inflation is easier to control in other countries because the tax systems are different, or because their economies have different wage-price dynamics.
Further research on these topics is urgently needed to better understand why New Zealanders face some of the highest interest rates in the world.
That is the kicker. What are the structural imbalances in the economy that implied we required a relatively high OCR.
There’s no mystery as to why borrowers prefer fixed over floating rates: they’re consistently cheaper. The tricky part is why they’re cheaper.
On your second bullet point: banks may borrow short and lend long, but they hedge the duration mismatch using interest rate swaps, leaving them borrowing and lending long – so changes in short-term rates don’t affect their margins. What puzzles me, though, is that the counterparties to these swaps – who are largely offshore – have been content to receive a fixed rate that was much lower than the floating rate that they paid. So they’re happy to allieviate NZ of a major risk and take a big loss in the bargain. Why is that?
Finally, I think Andrew overplays the angle that the RBA has been able to get cash in homeowners’ hands much quicker than the RBNZ. In practice, the majority of Aussies keep their repayments the same when interest rates fall, so that they pay down the principal faster. So borrowers’ behaviour in Australia frustrates the central bank’s efforts to provide a quick fix almost as much as it does in NZ.
And as for his comment that “It is probably no coincidence that New Zealand is now experiencing its fifth quarter of recession, while output in Australia has scarcely declined” – since three of those five occured before the RBA started lowering rates, I can only assume he’s being facetious.
But in US, fixed mortgage rates is considered to have more benefits than floating mortgage rates.
Thank you for covering this, Matt
To respond on your first point, it seems likely that in most countries the focus should be on the relative price effects that stem from interest rate changes. But New Zealand is not most countries because of the net debt position. Consequently, it is necessary to analyse both the relative price effects and the income effects – after all, monetary economics concerns all the ways that the price, quantity, and terms and conditions of nominal debt contracts affects the economy, not just the ways on the margin. One would expect that a 3 percentage point cut in mortgage rates that delivers a 3 percent of GDP (after tax) net income boost to households would have significant effects on consumption: it is a massive amount of money available for spending or saving. (Certainly it would be surprising if an increase of that order did not lead to a decline in consumption, though as always it depends on the fraction of borrowers who are credit constrained.)
Jeremy Stein’s 1995 QJE model of the housing market suggests another way that the average mortgage rate may be important. He observes the housing market is dominated by households that sell one house and buy another, and that recessions are times when the housing upgrade market collapses. When a household has a fixed term mortgage, and the interest rate falls a lot, it becomes psychologically and financially difficult for a household to move (normally upgrade) because the financial loss breaking a mortgage is large. This may contribute to a decline in housing volumes. Note that this doesn’t occur when interest rates fall in the US because almost all of their (long term 30 year) fixed rate mortgages come with an option giving the household the option to break the mortgage and refinance if it wishes – something that took place on an incredible scale in the US between 2002 and 2004. (Of course, now that many US households have low rate 30 year mortgages, and 30 year mortgage rates now are higher than 30 year mortgage rates then, there is a disincentive to move, for the mortgage is generally not transferable to another property so moving means a higher mortgage rate.)
To Miguel; yes it is somewhat gratuitous holding up the success of the RBAas an example; but the RBA has had a mighty good run in the last 17 years, acieving very similar inflation rates to NZ without recessions, banking collapses, and with real interest rates consistently lower than those in New Zealand. In any case, I like being gratuitious about the RBA. It is hard to avoid noticing that if New Zealanders were able to have Australian level interest rates over the last decade (maybe a big “if”, but maybe not), New Zealanders would be better off by over $300 each every year: quite a large sum, really, particularly if the additional stabilisation benefits we get in practice from having an independent monetary policy are quite small.
I don’t think a 17-year good run precludes the RBA simply being lucky: Australia had something to sell that the developing world wanted to buy, which allowed the RBA to let the exchange rate do their job for them (and even then only barely). But that’s a debate for another day; my point was that the prevalence of floating rates in Australia can’t explain why they avoided recession in the first half of 2008, since the RBA were actually raising rates through that period.
I’m not clear what your “$300 a year” refers to, is that how much borrowers would be better off? If so, I have two questions:
(1) How much worse off would savers have been?
(2) What would borrowers have done with the extra $300? The cynic in me says they would have just leveraged up even more against the same assets and ended up paying an extra $300 a year in interest anyway.
Should have established a fiscal money policy, combined with dumping economic system.
Hello Miguel
The $300 plus is a back of the envelope calculation based on net debt of $170billion or approximately 100% GDP, a population of 4.3 million, and a 0.75 percent nominal interest rate differential with Australia (more on floating rates, less of 5 year rates; from 2000-2009, NZ’s wholesale 90 day rates exceeded Australia’s by an average of 0.96%, while 5 year rates were 0.57% higher; our slightly lower inflation rate means that real rate differentials are approximately 0.2% higher than this.) The product of these numbers is roughly $1.3 billion per year, or $300 per person per year. This is a net figure; borrowers would be better off, and lenders (and the tax collectors) worse off.
This is actually an extremely large figure. The traditional optimal currency area literature usually posits a cost benefit analysis in terms of the possible benefits of monetary independence against the transactions costs of a separate currency. The benefits of monetary independence would need to be very high indeed to overcome the interest rate costs of this magnitude. Of course, many people people do believe the benefits of monetary independence are large – although it is no longer clear that the Reserve Bank is fully in this camp, since their interventions in the foreign exchange market in the last two years clearly indicate that a belief that an “independent” exchange rate can be a cause of shocks in the economy as well as a means of stabilising shocks. But if the main reason for having an independent monetary policy is to stabilise the economy against shocks, and if you are paying a large interest rate premium for this privilege, then you want to be sure that monetary policy is conducted in a manner that maximises its potentially stabilising influence. This is of course my original point: the stabilisation benefits of monetary policy are unlikely to be maximised if most households arrange their affairs so that they are only influenced by fluctuations in short term interest rates with a considerable delay.
OK, but question (2) still stands. Because if lower interest rates were the end of the story, then the optimal outcome wouldn’t be to set them in line with Australia, but to set them at zero permanently. And that would just be silly.
The benefit of monetary independence is that prices can adjust to absorb shocks – that means the exchange rate as well as interest rates, with no judgement about which one does the adjusting. So the ability to borrow at fixed rates is a “problem” only to the same extent that the ability to hedge your FX exposure is a “problem”. (In fact, higher rates in NZ actually benefit exporters who hedge, as their average hedged rate will be below the average spot rate.)
As for the Reserve Bank’s “beliefs”, I’d suggest their recent FX intervention has little to do with belief and a lot to do with political pragmatism. On the other hand, their ongoing complaints that global investors are lumping the NZD in with other non-USD currencies, and more recently their threat to do something to counter this, is to me a pretty strong signal that they still see value in monetary independence.
Clearly lower interest rates are not the end of the story; but having the OECD’s highest interest rates is not necessarily the most desirable outcome for a country either, particularly if that country has a large net borrowing position. To a large extent, the point with an independent monetary policy is to choose a time path for interest rates that is correlated with the shocks hitting the local economy, not to have mean interest rates different from the rest of the world. Unless NZ has been peculiarly unlucky with its shocks in the last two decades, so we have had a series of random shocks that have meant we needed much higher local currency interest rates than other countries, it would be nice to think we don’t need a huge and costly interest rate premium.
There really are two issues. The first is whether monetary policy – changing the price, terms and conditions or the quantity of the nominal debt contracts used in the economy – is being used optimally to stabilise the economy and achieve low inflation. It is by no means obvious that the goal of stabilising the economy is achieved when the vast majority of household borrowing contracts are at fixed term rates, if for no other reason that there is a considerable delay between when short term interest rates change and when a large fraction of households change their behaviour. Having a steeply inverted yield curve for five years seems to have induced a large majority of NZ households to borrow on fixed term rates, and this may be lowering the effectiveness of monetary policy.
The second issue, which is quite different, is whether an independent monetary policy is likely to be a good policy for NZ. Monetary independence can assist the stabilisation of an economy, particularly if the most of the shocks hitting the economy are peculiar to the local economy. And, yes, monetary policy works both through interest rate and exchange rate channels. But an independent currency need not be stabilising: there are good theoretical reasons why this may not be the case, and ample empirical evidence that exchange rate volatility can be excessive and harmful. The Bank’s frequent complaints about the level of the dollar are consistent with the idea that the exchange rate can be destabilising; so is the voluminous evidence that UIP doesn’t hold because of frequent changes in risk premiums. (Neither evidence denies that the exchange rate often behaves in the expected manner in response to changes in interest rates: there is some pretty good evidence that it does depreciate when interest rates are cut, and appreciate when monetary policy is tightened. This evidence just says that the exchange rate oftentimes changes for other reasons, and these changes can be destabilising.) And even if monetary policy is on average stabilising, the economy has to balance the additional stabilising effects (say of having a NZ specific monetary policy rather than Australian monetary policy) against the additional costs associated with an independent currency. In New Zealand, these costs are probably dominated by the additional interest rate margin New Zealanders pay for the privilege of borrowing in NZ dollars, given the net debt position: and these costs have been very large for the last two decades.
But this is an old issue, one that NZ politicians and bankers are tired of (even if in my view (and the view of a majority of business people) they have made the wrong decision). The issue going forward, under the assumption that the benefits of monetary independence are much greater than the costs, is whether inflation can be controlled without needing the frequent use of a steeply inverted yield curve. One would hope the answer is yes (since most other countries can do it); and it may be the case that the solution lies in finding a better way to influence inflation expectations.
To say that “it’s not desirable for a high-debt country to have such high interest rates” gives the impression that we’re the innocent victim here. But there’s a reason why we have so much debt: we CHOSE to borrow it. The more logical conclusion is that interest rates haven’t been high enough to deter borowing or encourage saving, even if they are the highest in the OECD (and I’d junk that comparison anyway – a reluctance to break away from the pack has been the great failure of OECD central banks in the last decade, and it’s one area where the RBNZ deserves to be applauded).
The persistently inverted NZ yield curve is bizarre, but for me the puzzle is not why the short end has been so high, but why the long end has been so low. As I said in my first reply: why have foreign investors been willing to take on our interest rate risk at a consistent loss? Now I don’t actually have an answer for this. Lack of RBNZ credibility is one possibility (i.e. if investors believe, despite all past experience, that the OCR will be dropped sharply in the future), but I don’t feel that really covers it.