‘Year Of The Refix’ Redux, Post RBNZ

• Mortgage positioning is looking more balanced, about halfway through a busy 12 months for mortgage refixes
• About $135b or 41% of all fixed-rate mortgages are due to reprice over the coming six months
• Evidence is growing that the typical fixing strategy is no longer just about picking the lowest rate
• RBNZ cuts cash rate 25bps, we forecast two more
• Gentle trend toward longer mortgage terms may continue, but we doubt there will be a rush

One of the things that jumped out at us at the start of the year was just how ‘short’ the mortgage book had become. Such was the preference for short fixed terms that over 80% of mortgage borrowings were set to experience a rate reset this year. That was the highest in 13 years, earning 2025 the ‘Year of the Refix’ tagline.

Now that we’re about half-way through (data to June), and with the latest Reserve Bank meeting under our belts, we thought it was a good time for an update.

Peak short in the rear view

We expressed some unease about extreme short positioning in our January note. As it turns out, the sentiment shifted relatively quickly thereafter. Indeed, the November 2024 data referenced at the time turned out to be peak short for the mortgage book.

The share of new borrowings on floating and six-month fixed terms has reduced a little, and there’s been snatches of increased interest in two-year fixed terms. Consequently, the share of new mortgage borrowing on terms of 12 months or less has fallen from a high of 94% in November, to an average of 71% over the three months to June.

Through the recent heavy period of refixing activity we’ve also seen:

a) Many take the opportunity a mortgage rollover affords to switch providers. A record amount of new residential lending in June (the latest data) – 30% – was attributable to mortgage switching.

b) A quickening pace of decline in the average interest rate paid on mortgage debt. Heightened refixing activity has sped up the process of borrowers ‘rolling down’ onto lower mortgage rates. The average paid rate was 5.66% in June, down from a 6.39% peak. It’s still got a ways to go. We estimate a decline to around 5% by the end of the year.

The gradual shifts in term preference evident in new lending flows have been enough to nudge out the term profile of the stock of mortgage borrowing.

For example, the share of outstanding fixed-rate borrowings with remaining terms of six months or less has declined from the highly concentrated 44% as of January, to 36% as at June (chart below). There’s now a slightly higher share with remaining fixed terms of 6-12 months, while the share in the 1-2 years bucket has lifted a solid six percentage points from January.

The broad conclusion is that mortgage positioning is still shorter than average, but it’s more balanced than it was.

The next question is whether the gentle trend toward longer fixed terms will continue. After all there’s still plenty of refixing due to occur over the coming six months. About $135b worth, or 41% of all outstanding fixed-rate mortgage debt.

As always, there are many angles to consider. But from a macro perspective, we think there are two major drivers worth thinking about.

1. Where will rates go?
 

First, there’s what happens to mortgage rates – particularly the relativities between rates for various terms. The vibe of yesterday’s RBNZ announcement was supportive of our view additional modest declines in mortgage rates are likely, concentrated more at shorter terms.

The chart shows the current mortgage curve along with our expectations (which are based on our wholesale interest rate forecasts). As mortgage rates have come down this year, shorter term rates have fallen by relatively more. That is, the mortgage curve has steepened. We think this trend has a little further to go.

Forecasts are, as always, only a guide. But the fact fixing for two years or longer may end up a little more costly (upfront at least), relative to fixing for six months or one-year, may encourage some to stick with those shorter fixed terms.

It's the old ‘picking the lowest rate’ strategy that, at times, has been a dominant force in term selection. But perhaps no longer!

2. Where are we in the cycle?
 

That brings us to the second big factor influencing mortgage fixing behaviour: where we are in the interest rate cycle. It appears that this more forward-looking approach has become the dominant consideration, at least in aggregate. If so, it’s probably something to be celebrated.

Consider the following chart. It plots the difference between three-year and one-year mortgage rates (yellow line) against fixing behaviour (blue line). Up until 2020, fixing activity generally ‘chased’ the available rate relativities.

So, when relative pricing moved in favour of shorter-terms (one-year rates falling relative to three-year), fixing behaviour followed suit, with the share of mortgage borrowings at terms of less than a year rising. The same was true in the opposite direction. The correlation was strong: +0.7 over the period from 2005 to 2020.

But, as we first pointed out a year ago, times are changing. The correlation in the five years since March 2020 has flipped from strongly positive to negative (-0.3). Now, borrowers – in aggregate – appear much more inclined to absorb what might be unfavourable upfront cost relativities to position themselves to potentially benefit long-term as we move through an interest rate cycle.

For example, borrowers shortened their borrowing terms well ahead of the RBNZ’s latest easing cycle, despite the extra upfront costs of doing so at the time (with short-term rates above long-term). That proved prescient as rates were subsequently cut aggressively. But as the easing cycle has advanced, borrowers have now started to push out for longer terms even though longer-term rates are above shorter-term equivalents, for the most part.

As illustrated in the chart, this break from the pick-the-lowest-rate ways of the past looks to have coincided, cleanly, with the start of the COVID era in early 2020. We’re not sure why (send in your theories!) other than to state the obvious and say all sorts of correlations and relationships changed around this time.

Bringing in the RBNZ

The Reserve Bank yesterday cut the Official Cash Rate (OCR) by 25bps to 3.00%, the seventh cut this cycle. The cash rate is now 2½ percentage points below the peak. Nothing to surprise there. Where we were surprised – in a good way – was with the Bank’s new-found conviction that additional cuts beyond this will be required.

For some time, we’d been factoring in one final 25bps cut to a low of 2.75%. We’d flagged the risk of more, and the RBNZ’s clear signalling means we’ve now acted on that. Our forecast trough in the OCR has been pulled down by another 25bps to 2.50%, by November.

There still could be an element of the Reserve Bank feeling its way to the bottom of the cycle. Whatever the exact low point turns out to be, we are edging onto the home stretch for the easing cycle. Borrowers’ small step away from shorter fixed terms appears consistent with this.

As to whether that continues, the two factors discussed above, on the face of it, will be working in offsetting directions. However, mortgage fixing behaviour now appears less sensitive to current relative pricing, and more focused on the risks and opportunities from the shape of the next interest rate cycle.

If that holds, we may see the gentle trend towards longer terms continue as borrowers gradually factor in the end of the Reserve Bank’s easing cycle. We doubt there will be a rush toward longer-term fixes though. Not while the RBNZ maintains a bias to lower interest rates further, the economy remains weak, and the risk of rate hikes remains a long way off. Our read of mortgage rate ‘break-evens’ (comparing our interest rate forecasts to implied market pricing) is also consistent with the idea that borrowers have some time up their sleeve.

© Scoop Media