Kudos to The Australian columnist Terry McCrann for bravely calling early increases ahead of everyone else, and to CBA’s Gareth Aird for generally being ahead of the curve.
The RBA’s governor, Philip Lowe, then gave an impressive press conference, where he delivered a mea culpa on the central bank’s woeful forecasting track record, describing it as “embarrassing” and something that had to be fixed. This extremely rare bout of humility was way overdue but nonetheless carried an enormous contradiction.
It is not just that the RBA’s forecasts have been pathetically poor. It is that the RBA has then lurched into huge, multi-year policy pre-commitments on the basis of these specious perspectives.
Recall the commitment to not raise rates until at least 2024. This was also accompanied by the promise to keep buying the 2024 government bond yield to ensure its yield remained equal to the 0.1 per cent cash rate. And, more recently, the pre-commitment to wait for the May and June wages data before increasing rates, ruling out a May move. All these de facto promises have been overturned, torching the RBA’s credibility.
Despite admitting that his economists could not forecast their next footstep, Lowe then repeatedly told the world that the RBA expected to lift its cash rate to a 2.5 per cent “neutral rate”. Over and over again. On what planet can the RBA have confidence in this expectation? Why even utter these words? What happened to just being data-dependent?
All of this tells us that the RBA has learnt absolutely nothing about the perils of pre-commitments. It demonstrates that it is not yet willing to concede defeat about its deep intellectual deficiencies. It cannot help but reflex into pretending to be an all-seeing, all-knowing diviner of our destinies.
The bad news is that the RBA has an especially distinguished track record of getting the Aussie housing market’s reactions to its cash rate changes horribly wrong. This is all the more surprising because the housing cycle is exceptionally easy to forecast.
House price will force RBA to pause
Whether the RBA wants to admit it or not, its interest rate decisions will be heavily influenced by the direction of house prices, which after years of uber-cheap money have never been more inflated. This will be amplified by two dynamics.
First, there are hundreds of billions of dollars worth of circa 2 per cent fixed-rate loans that will roll into variable rate products carrying much higher interest rates in the next two years.
Second, households are much more sensitive to interest rate changes: our household debt-to-income ratio is sitting around 186 per cent, in line with all-time highs.
This column projects that the RBA will likely be forced – if it acts prudently – to pause its monetary policy tightening process after the first 100-150 basis points of hikes as a result of the start of a record 15-25 per cent decline in Aussie home values (as we outlined in October last year).
It will also be held back by banks unilaterally lifting mortgage rates out of cycle as their funding costs normalise. Put differently, banks will do some of the heavy lifting on interest rates for the RBA.
Potential losses of $1.5trn
Since the total value of residential real estate is worth $9.9 trillion, the RBA will likely impose losses on households worth some $1.5 trillion (assuming just a 15 per cent draw-down). Superannuation will also shrink in value as listed equities, infrastructure, property and private equity are smashed.
Banks have already been forced to lift their three-year home loan rates from 1.98 per cent 12 months ago to 4.5 per cent. Non-bank lenders are likewise wearing an overdue repricing in their cost of funding via a sharp increase in the credit spreads on their residential mortgage-backed securities (RMBS).
What makes the RBA’s history of housing misses so odd is that its own staff have actually developed an outstanding model of the market, which we have refined and used ourselves. While Martin Place has been historically insouciant to the insights rendered by this analysis, which explain the 2012-17 and 2020-22 booms, it did unfurl the “Saunders-Tulip” model for the first time in its latest Financial Stability Review.
Our updated version of the model points to a 33 per cent correction in house prices after a permanent 100 basis point increase in mortgage rates. Our official forecast involves a more modest, 15-25 per cent correction, which would still be the largest loss in housing market history.
The risk is that the RBA once again gets wrapped up in its flawed projections of the future, and hikes too hard and fast. Lowe himself seemed to endorse financial market expectations for 150-175 basis points of hikes this year alone, which has now become a consensus view. Goldman Sachs has taken this to an absurd extreme, alleging the RBA will hike 260 basis points by the end of 2022. Can you imagine what our world will be like when the discounted variable mortgage rate jumps from 2.25 per cent right now to 5 per cent (allowing for bank top-ups)? It would be Armageddon.
Back in 2013, we repeatedly warned the RBA – publicly and privately –that by slashing its cash rate it was going to precipitate the mother of all housing bubbles, powered by double-digit house price appreciation. Between 2011 and 2017, the RBA dropped the cash rate from 4.75 per cent to 1.5 per cent. Over this time, national prices soared a stunning 41 per cent based on CoreLogic’s all-regions index, which includes both the capital city and regional markets.
We further advised the RBA in 2013 that it would have to embrace the application of macro-prudential constraints on credit creation, which it was – at the time – opposed to. By late 2014, APRA had been compelled to start introducing these measures, which were progressively ramped up in the years that followed.
Between 2015 and 2017, APRA’s limits on credit creation forced lenders to materially jack up rates on investment loans. In 2017, we argued this would trigger a 10 per cent decline in national prices. Between 2017 and 2019, the 8 capital city index slumped 10.2 per cent (the all-regions index fell 8.3 per cent).
In mid-2019, we forecast that future RBA rate cuts would drive house prices up 10 per cent over the next year, which is what we got: the 8 capital city index rose 10.3 per cent while the all-regions index climbed 8.9 per cent.
The real howler for the RBA – and bank economists – was its forecasts for 10 per cent-plus house price declines during the pandemic coupled with its inability to anticipate the stonking recovery. In March 2020, we projected a very short, and shallow, correction of 0-5 per cent followed by capital gains of up to 20 per cent, starting in September 2020.
CoreLogic’s all-regions price index fell 2.1 per cent between March and September (the 8 capital city index declined 2.8 per cent), and thereafter started climbing quickly.
The RBA claimed there would not be a housing boom in response to its pandemic policies because population growth was non-existent. We countered that the profound change in purchasing power care of a huge reduction in the cost of debt combined with robust income growth would power a sharp, 20 per cent increase in prices.
By October 2021, the all-regions dwelling value index had indeed increased by 20 per cent. In that month, we updated our forecasts to include at least another 5 per cent of national price growth until the RBA started increasing its cash rate in the second half of 2022. More controversially, we also projected that national home values would thereafter correct 15-25 per cent after the first 100 basis points of rate increases.
Since this forecast was released, most bank economists have embraced the proposition, looking for a 10-15 per cent correction in house prices over 2022-24.
It’s worth putting this payback in context: home values across all markets have appreciated by 37 per cent since mid-2019 when the RBA first started cutting its cash rate from 150 basis points down to the 10 basis point level that prevailed before this week’s increase.
There is no precedent for the RBA hiking through a period in which house prices are falling materially. We think the RBA will try to persist through the initial drawdown. Yet it is hard to imagine the RBA tightening financial conditions once it has wiped more than 10 percentage points off Aussie households’ most valuable asset. If the RBA’s increase trigger the minimum correction we expect, households will have lost $1.5 trillion.
There are potential mitigants. Wage growth is expected to recover robustly. If the RBA is gradual with its rate increases – spreading them over a number of years – a non-trivial share of the correction in prices could come via household income growth.
And the last thing the RBA wants to do is to tip us into a recession that would deny workers their jobs and the healthy wage growth that the RBA has been trying for years to stimulate. It will, therefore, be very closely following the world-leading daily hedonic house price indices produced by CoreLogic, which revalue the entire 10.5 million dwelling housing stock every day using 100 per cent of the incoming sales (as reported by agents and checked using land titles data).
As house prices start rolling over, we expect the RBA to materially modulate the pace of its monetary policy tightening process in response to clear evidence that its transmission mechanism is working.
If Martin Place has learnt anything from the pandemic, it should be the importance of being data-dependent. But that’s a big “if”.