Three points from the RBA’s FSR

The RBA’s October FSR was less concerned about current financial stability issues in Australia than the April FSR. This allowed the RBA to pivot to thinking about longer-term financial stability risks — which means the (mostly political) campaign against housing investors. Investors aren’t empirically riskier, so the facts don’t support their differential treatment, but the politics are much better.

The RBA seems incrementally concerned about China and EM. In China, Evergrande is the current concern, but reform on multiple fronts is the more enduring source of ‘unintended consequences’. In EM, the delayed recovery means that risks remain elevated.

More macro-prudential tightening and downside risks to Australia’s trading partners (China and EM) are a reason to think AUDUSD should be lower. It should also extend the period of low rates (though I’d note that global factors have recently dominated Australian factors when it comes to pricing the Aussie short end).

The RBA is no longer worried about bad debts

The April FSR was concerned with the possibility that there might be a rising tide of bad and doubtful debts as stimulus was wound back. Instead, we had a strong labour market and rising house prices. So, recent experience has made the RBA much more sanguine about the financial stability consequences of the Delta-shock. They simply aren’t really worried about bad debts as a consequence of the current lockdowns.

You might characterize the period up until now as being about reducing the probability of a bad debt cycle. The chart below shows that the recent house price boom has significantly lowered the distribution of LVRs (lower LVRs mean less risk). It follows that a (much) larger shock to house prices would be required to push households into negative equity and create a meaningful bad debt cycle. By allowing house prices to race up, insurance against future economic shocks has been taken out.

Of course, you can have too much of this good thing. For every grey-beard like me that receives equity during the rally, there’s a new home buyer (like my kids one day) that’s taking on more debt to get set. Or investors, buying with leverage in anticipation of capital gains … which brings us to the next point.

Slowing housing = DTI+LVR?

With sufficient insurance against a credit-supply-choking bad-debt cycle having been taken out, it’s time to wind things back. However, the RBA’s inflation target remain distant so policy-makers must look at non-cash-rate means of constraining house prices.

In that vein, there’s a new chapter titled Mortgage Macroprudential Policies (MMP). There wasn’t a chapter with this title in April FSR. The main thrust of the new chapter is that the official family are going to keep going until they get housing under control. Getting housing under control means house prices and debt growing in line with incomes (never mind that lower rates should mean that debt to income ratios rise, as the amount you can borrow rises in a non-linear way as the mortgage rate declines). The primary target of these policies seems likely to be investors.

If it seems a bit political, that’s because it is … first home buyers are a protected species in Australia.

APRA made that explicit this week when they highlighted the differential impact on investors relative to first home buyers in their mortgage-buffer press release. When explaining why they preferred an increased buffer to a higher floor rate, they made a virtue of the fact that the larger buffer had a bigger impact on Investors. The tone of the RBA’s chapter on MPP continued in that vein.

The text of the FSR implies that the next policy step will be to combine Debt-To-Income (DTI) and Loan-To-Valuation (LTV) restrictions in the second stage of macro-prudential restrictions.

The RBA noted that:

limits on high-DTI lending, depending on their calibration, may also constrain some borrowers, particularly investors, who are well placed to service their debt. Combining DTI restrictions with LVR restrictions could help to avoid this problem by capturing riskier borrowers without constraining high DTI lending to borrowers who are much lower risk

RBA, Oct 2021 Financial Stability Review (my emphasis)

Of course, this is a bit of a fake problem.

If a million dollar house pays itself off, and you own five of them, you’ll have a very high DTI but no financial stress. Does it mean you’re more risky? Well, that depends. Empirically speaking, investors are no worse than owner occupiers — but they are ‘fair game’ politically. Investors with a many properties have high DTIs because houses cost 25x their income stream. Never mind that the portfolio that pays itself off. That makes DTI limits politically attractive.

Investors have higher DTI ratios

When we look at LVR ratios, you can see that the LVR ratios at origination for Investors are lower than for owner-occupiers. That’s not an accident. If you own five houses, you probably know what you’re doing. You avoid costs like mortgage insurance. First home buyers are deposit-constrained, and therefore tend to have higher LTV ratios. This makes LTV restrictions less politically saleable.

Investors have lower LVRs

While rising house prices are a political winner in Australia, we do also have periodic crises (moments of guilt?) about housing affordability. The concern is typically the ability of first home buyers to get a foothold in the market. The solution, both in Australia and New Zealand, is always to push investors back a little to make space for first home buyers. This is occurring in NZ just now.

Investors often add a bit of late cycle momentum to the market, so the macro effect will be to slow the present upswing and extend the period of ultra-low rates. The policy-induced weakness in Aussie housing up to 2019 is a key reason that the RBA never looked like raising rates in that period.

Finally, for those that are wondering about the how the recent decision to increase the product buffer rate by 50bps works, the RBA has a nice chart that lines up pretty much exactly with the policy that was announced. According to APRA, the 50bps increase in the buffer, to 300bps, will reduce maximum loan size by about 5% (so 100 to 95 in the chart). Not everyone borrowed their maximum before, so only for some subset of borrowers would the constraint matter: these are the borrowers on the horizontal part of the pink line.

Not all borrowers were constrained under the old rules, but a few more are constrained under the new rules, so the 5% decline in the maximum loan will lead to a decrease in the total amount of credit written that’s a bit smaller than 5%.

To those that say that the new policy won’t make much difference (as mortgage specials plus buffer rates are below minimum assessment rates, or so called ‘floor rates’), my response is that it’s the signal that matters. The Council of Financial Regulators is sending a signal that they are going to make conditions for credit growth less favorable. The RBA just told us that they are going to use DTI and LTV restrictions in the next stages. This is going to work.

The pendulum is swinging back toward tightening for the first time since mid 2019. That’s important.

More worried about China and EM

The problems in China got more play in the October FSR. The RBA downgraded their assessment of China at the September RBA meeting (something that appeared in the minutes, but not the Sep or Oct post-meeting statements). These concerns were unpacked a little, for the first time, in the October FSR.

While there’s a couple of hundred words about Evergrande and Chinese property, the last sentence struck me as the truest:

In general, the numerous and simultaneous policy changes occurring at this time raises the probability of unintended consequences.

RBA, Oct 2021 Financial Stability Review (my emphasis)

This multiple-tightening problem was a feature of 2018, when China tightened credit and pollution controls concurrently and (arguably) reversed the global upswing that was in train. Of course, trade tensions between China and the US were also a factor at this time. The concurrent decline of global PMIs and the EUR were a feature of this period (it was also a decent time for rates carry trades, outside of USD rates).

Of course, there’s never been more debt in China, but that’s true most years. I am not too concerned about a financial melt-down caused by China (for example by an Evergrande-event or similar). I do, however, think that the politics of stimulus v. reform have changed (for now) and that next few years are more likely to be characterized by slower credit growth and lower steel intensity.

That means a lower Australian terms of trade, and probably a lower AUD. Given the importance of Chinese Residential Investment for steel demand, we have probably already seen the peak for Iron Ore prices and Chinese Steel steel consumption. Which means we’ve probably also seen the peak for the Australian Dollar.

The rest of EM is unlikely to make up the difference. Lessor access to vaccines means that EM isn’t expected to make a full recovery from COVID for some time. That growth differential between EM and DM should mean a weaker AUD.

Conclusion: MPP-tightening & slower China/EM

The main takeaways from the FSR are that the Council of Financial Regulators (CoFR) wants slower housing, and that the RBA is more worried about the outlook for trading partner (China / EM) growth. These things should combine to delay rate hikes and create some downside risk for the AUD. Pressure from the looming review of the RBA works in the same direction.

The October FSR should be understood as a tool to help the CoFR achieve their housing aims. While it’s true that a 50bps increase to a 300bps buffer over the ‘product rate’ remains below minimum floor rates in many cases (and so doesn’t bite), the key point is that regulators have stirred and expect to get a result. The FSR lists the tools the CoFR has to achieve their housing aims; they will use them until they get the result they want.

While the next step is likely to be announced after APRA’s review, the FSR makes it pretty clear that it’s going to be some combination of DTI and LVR restrictions. If you think it odd that the RBA is announcing this before APRA, I’d note that they semi-announced announced the increased serviceability buffers a day before APRA, in the October RBA post meeting statement.

More macro-prudential tightening means less cash rate tightening. It slows housing and the broader economy, and therefore extends the distance to the RBA’s targets. This works to prolong the period of ultra-low rates.

Increased concern about China and EM will also keep the RBA from pre-emptively raising rates.

Finally, if the RBA is right about the downside risks to China, & EM more broadly, it means downside risks for the AUDUSD. The AUDUSD is currently ~0,73: I think it’s more likely that the AUDUSD trades 0.70 before it trades 0.75

Leave a Comment