Global Markets Research
Data Flash (New Zealand) - Strategy
Outlook for the NZ yield curve
Summary The historical experience suggests the NZ yield curve is presently in the midst of a move from a very positive
slope to a very negative slope. However, the historical experience may not be relevant. The RBNZ has moved to an
official cash rate and some of the upward pressures on inflation in the mid-1990s are not present. Having said this, the
weakness of the currency provides a threat to the inflation outlook and a significant offset to other factors. On
balance we think the OCR will spend an extended period at or above 7%, though it is unlikely to reach this level until
next year and the very near-term pressure is for the curve to steepen slightly. Notwithstanding this, if we are correct
then the yield curve looks like being flat to inverted for a considerable period. The main risk to this view is a
significant sell-off in the US bond market.
Introduction
We expected the RBNZ to deliver a hawkish message in its August Monetary Policy Statement. As such, we were in the
process of writing a piece on how inverted the NZ curve could become. Thankfully, this was not scheduled for release
until after the RBNZ statement! In the event the RBNZ surprised on the dovish side, with the result that the curve has
become significantly less inverted. The 3Y/10Y spread moved from -15 bp the day prior to the statement to -3 bp at the
start of this week. The rest of this note looks at the possible shifts in yield curve shape over the next 6-12 months,
beginning with a review of the 1990s experience.
The NZ yield curve in the 1990s
The above chart shows that the shape of the NZ yield curve has been through a number of clear cycles in the 1990s. It
is also notable that the 90 day bank bill/10Y bond spread has tended to sit above +100 bp or below -100 bp, with little
time in between. Simply looking at the historical experience would suggest it is unlikely that the present flat 3M/10Y
spread will persist for long, with the cyclical pattern of curve movements suggesting we are heading for a period of
significant inversion. However, the past may be a poor guide to the future. The RBNZ has shifted to the use of an
official cash rate (OCR), the Bank seems more relaxed about inflation "spikes" and some of the sources of inflation in
the mid-1990s (most notably the housing market) may be absent this time around (or certainly less of a concern).
Will the shift to an OCR regime change the behaviour of the curve?
In an analysis published late last year, we looked at whether the operational change in the RBNZ's monetary policy
regime could change the behaviour of the yield curve over the course of the cycle. In particular, we noted that the
shift to the Official Cash Rate may not only reduce the day-to-day volatility of short-term interest rates, but could
also lead to a reduction in the amplitude of the short-term interest rate cycle over the course of the economic cycle.
This analysis is updated below.
The chart above shows daily 90 day bank bill yields for the first 8 months of 1998 and over the equivalent days in 2000.
Not surprisingly, the chart shows a significant reduction in the day-to-day volatility of short-term interest rates over
the two periods. This would be true no matter what period we chose to compare with the post-regime change experience.
The official cash rate has firmly anchored short-term interest rates over the near-term. The key issue, however, is not
short term volatility but the extent to which interest rates need to move over the course of the economic cycle. It is
not obvious that simply changing the mechanism by which monetary policy is operated will have an impact on the link
between interest rates and inflation. Perhaps the "fact" that the transmission of monetary policy signals is clearer
than in the past will mean a smaller cycle in interest rates than previously is possible. The change in volatility may
influence households choice of fixed versus floating, which in turn could change the transmission mechanism. These
effects are likely to be relative small , however, suggesting there is little reason to believe that on its own the
change in the operating regime will have a marked impact on the amplitude of the interest rate cycle. As always, the key
determinant of the interest rate cycle will be the strength and persistence of inflationary pressures. In this regard
there are reasons to be believe the present cycle will be different from the last one. However, there are also
offsetting influences. Some of the differences point to a lower peak in interest rates, while some suggest the pressures
on interest rates to rise could actually be greater. To some extent the decision to leave the cash rate unchanged on 16
August reflected the RBNZ's uncertainty about which set of factors would prove dominant.
Housing vs the exchange rate: the pressures on interest rates are mixed
Ultimately short-term interest rates, or more correctly monetary conditions, will rise to the level required to keep the
inflation rate towards the middle of the RBNZ's target band over the medium-term. Over the previous interest rate cycle
the peak required turned about to be around 10% in mid-1996 (counting the similar peak in mid-1998 as a policy "error"
rather than as something that was "required"). There is reason to believe the required peak in interest rates over the
present cycle will be lower than the last - perhaps by a considerable margin. This stems primarily from what is expected
to be a relatively weak housing market.
As we have noted in previous research, the level of household indebtedness is substantially greater than in the
mid-1990s. This should increase the "bite" of monetary policy, even if the greater proportion of fixed rate mortgages
means there is some delay before the bite starts to work. Associated with this, the demand pressures on the housing
market from inward migration should be considerably less than in the mid-1990s. Indeed, at present there is net outward
migration. Not all the pressures on inflation are likely to be downward in relation to the previous cycle. In
particular, it looks like the currency will have a much less dampening effect on traded good prices (and perhaps none at
all) than in the mid-1990s. At the time of writing the NZ TWI was sitting close to its record low of just under 51.0. In
contrast to previous episodes of currency weakness, the present weakness in the NZ dollar is occurring against the
backdrop of a strong global economy. As a consequence commodity prices are rising and one would typically expect the NZD
to be rising as a result. For instance, the "usual" relationship between the CRB and the TWI would suggest the currency
is almost 20% undervalued. Even if one objects to the use of the CRB because of the inclusion of oil in the index, the
gap between the NZD and commodities produced in NZ is at historical highs. As evident in the fact the NZD value of the
ANZ Commodity Price Index is at its highest ever level.
Survey evidence reveals considerable input price pressure on domestic firms as a result. It is also the case that those
parts of the economy exposed to the external sector are considerably more buoyant than those parts focused on the
domestic economy. While the current softness in the domestic economy may be restricting the ability of firms to pass
higher costs on, there may be a limit to how long this can last. Though there may not be a “blow-out” in prices, there
could be a significant amount of margin catch-up in store (unless there are significant productivity gains possible that
have not yet been tapped). Certainly, unless the currency appreciates quite dramatically the traded goods sector looks
like being a source of inflation over the next year or two. This is quite different from the experience in the mid-1990s
when traded good prices were flat and provided a significant offset to non-traded inflation that was in excess of 4% for
an extended period. Whether traded good inflation is of concern to the RBNZ may well hinge on the performance of
non-traded inflation. The MPS put considerable emphasis on the importance of non-traded prices in indicating the degree
of ongoing inflation pressures in the economy. With the construction sector weak, non-traded prices could ease over the
next quarter or two and thus delay the RBNZ's response to rising traded good prices. Unless, however, the domestic
economy remains subdued for an extended period some response from the RBNZ is eventually expected.
The outlook for the yield curve
As discussed, pressures on inflation are mixed. This makes it difficult to be definitive about the outlook for the level
of the cash rate let alone the shape of the curve. On balance we think the OCR is likely to rise from current levels, to
somewhere above 7% during the course of next year. Softness in the activity numbers may keep the RBNZ from tightening
again this year, but we expect renewed growth and rising inflation to force the Bank to eventually shift the cash rate
up. While short term interest rates seem unlikely to approach the sorts of levels seen during the previous cycle, the
relative weakness of the NZD and the associated inflation pressure will see rates remain at mildly restrictive levels
(i.e. 7% or more) for an extended period. While renewed tightening concerns are likely to be negative for the relative
performance of the NZ long end, we expect the outright level of long bond yields to be supported by the US. In the
absence of a surprising surge in US inflation or the prospect of a rapid elimination of the Budget surplus, we do not
see significant upside in US bond yields. For this reason we would be surprised in NZ 10Y bonds move above 7% for a
prolonged period. This means a flat to inverted yield curve looks likely to be in place for an extended period. Having
said this, in the very near-term, i.e. the next month or two, the pressures are for the curve to steepen further as the
market reflects on the implications of the RBNZ being on hold. If our view about the economy and inflation proves
correct, however, any tendency to steepen should be temporary.
David Plank, Fixed Income Strategist
This, along with an extensive range of other publications, is available on our web site http://research.gm.db.com
In order to read our research you will require the Adobe Acrobat Reader which can be obtained from their website
http://www.adobe.com for free.
For answers to your EMU questions, check Deutsche Bank's EMU web site http://www.db.com/emu or email our helpline business.emu@db.com.