Forecasting House Price Change by Mortgage Loans — cases of Australia and New Zealand
Australia
Jericho (2021) put forward his inspiring hypothesis that the continuous rise in Australian house prices in 2021 was positively and strongly associated with the increase in housing finance (mortgage loans). He made a forecast that Australian house prices would increase by 30% by the end of 2021 by referring to the more than 80% increase in housing finance before that period.
Theoretically, when interest rate is ultra low, more borrowing is encouraged, demand on houses is increased, and higher house prices can be expected.
Jericho (2022) then confirmed his forecast after about six months as shown in the charts below. I put his two charts for easy comparison. The upper chart in Figure 1 with the forecast was made on August 4, 2021. The lower chart confirmed the forecast was then made on March 16, 2022. Basically, the forecast of a 30% increase is confirmed by a 24% increase.
Since he used 6-month leading data of housing finance, it becomes a leading indicator rather than simply an association. This makes it a good and valid forecasting model. That is: the change in housing finance can predict the change in house prices.
He further plotted a scatterplot between the 6-month leading data of housing finance change against house price change to identify their linear correlation. The upper chart in Figure 2 provides the best fit line and predicts a 30% increase in house prices by the end of 2021 corresponding to the then upsurge of 83% of the 6-month leading housing loans surged by 83% (top right circle in the upper chart). The lower chart provided by Jericho (2022) confirmed his prediction (the orange circle in the lower chart). According to CoreLogic (2022), the annual change in Australia’s house price index, released at the end of January 2022, rose by 29.8% in Sydney and 33.09% in Brisbane in the end of 2021.
A test of the Housing Supply Hypothesis versus the Money Supply Hypothesis
The accurate prediction of Jericho’s forecasting model reinforces the money supply hypothesis that house prices are more strongly determined by money supply (housing finance) rather than house supply. Prosper Australia (2022) also opined that “the COVID19 pandemic has provided a case study that seemingly dismisses this [housing supply] argument”, by citing Fitzgerald’s contention:
“The pandemic induced hit to net migration of -95,300 was equivalent to an extra 36,653 homes becoming available to the market. Instead of this added supply reducing pressure on land and housing prices, we saw a record surge…As this data reflects, raw market power is the driver of such excessive increases…Instead of the focus on supply, the role of record low interest rates and the demand it is directing must be recognised. That is the one common factor across economies.”
New Zealand
New Zealand’s housing market also shows a similar pattern. Figure 3 shows the monthly change of outstanding amount of residential mortgage loans in New Zealand (RBNZ, 2021a). Before the pandemic, the amount has remained stable at about NZ$1.5 billion per month. However, the amount has soared to over NZ$3 billion per month in the end of 2020, which was more than double of the previous average!
The Central Bank of New Zealand admitted that the surge in mortgage loans after the outbreak of the pandemic is related to the monetary policy changes, including the reduction of the official cash rate (OCR) from 1% to 0.25% in March 2020 (RBNZ, 2020a) resulting in a mortgage rate even lower than the inflation rate, i.e. a negative real interest rate, which is favorable for buyers to invest. In addition, the central bank also relaxed the mortgage loan restrictions and removed the upper limit on the proportion of mortgage loans that banks can undertake high-percentage mortgage loans (RBNZ, 2020b).
Among all mortgage loan borrowers, investors took actions fast. Immediately after the relaxation of the mortgage restrictions and interest rate cut, the annual growth rate of investors’ mortgage loans increased to 89% in December 2020, when the first home buyers’ increase was only 39% and the other owner-occupiers’ increase was 37%. It indicated a huge increase in house demand which imposed an impact on house prices. (Figure 4)
In fact, the Reserve Bank of NZ also admitted in its report released in February 2021 that post-pandemic measures led to a sharp rise in mortgage loans. It also relates to the sharp rise in property prices. House prices increased by about 27% in 2021 (Figure 5), a record high growth rate of house prices in New Zealand. But then when the central bank raised its interest rate, the growth rate of house prices started turning south rapidly.
Re-test of Jericho’s (2021) Forecasting Model on New Zealand House Market
Then, is Jericho’s (2021) forecasting model also valid in New Zealand’s house market? Figure 6 compares the time series of the year-on-year change of house price index of New Zealand (blue curve , left axis — HPIyoy) and the year-on-year change of mortgage loans of New Zealand in 6-month leads (orange curve , right axis — Mortgage yoy) from 2014 to 2022, with forecasting figures of house price changes from June to November 2022.
The mortgage loans data is collected from the Reserve Bank of New Zealand’s (RBNZ, 2022) New Residential Mortgage Lending by Borrower type — C31. Since it does not provide mortgage for re-financing data series, we simply use the Total lending A1 — All borrower types (A2+A3+A4+A5) LVRN.MMB1.AA in NZDm, which is not as representative as the Jericho’s (2021) series, as the refinancing series may not be related to house demand. The house price index data is collected from REINZ (2022).
Further, if the two series are plotted in scatterplot as in Figure 7, they depict a positively association. The R-squared of the best fit line is about 33%. Even though the explanatory power is not really high, the prediction of house price fall by 18% in June 2022 by the model has just been checked with the -1.7% fall. The magnitudes of the fall may not be close enough to justify an accurate forecast, but it accurately predicts the first month of negative growth in house prices after a continuous years of high rate of increase. The coming months’ forecasts will be checked later.
Summing up, the similar patterns of positive association between house price change and housing finance change in 6-month leads in both Australia and New Zealand, especially after the outbreak of COVID, supports the money supply hypothesis. Jericho’s (2022) model is easy to use as there are only two variables, but it cannot control the influences of other factors. My recent paper (Yiu, 2021) applied the real interest rate analysis in a five country study with GDPs and Unemployment Rates being controlled in the regression models. The results also confirm the money supply hypothesis. There are more and more evidence to support the money supply hypothesis that house price changes are more strongly shaped by money supply rather than housing supply.
References
CoreLogic (2022) CoreLogic Home Property Value Index — Monthly Indices, 31 January. https://www.corelogic.com.au/research/monthly-indices
Jericho, G. (2021) A drop in new home loans in June is no sign Australia’s house prices will return to reality, The Guardian, August 4. https://www.theguardian.com/business/grogonomics/2021/aug/05/a-drop-in-new-home-loans-in-june-is-no-sign-australias-house-prices-will-return-to-reality
Jericho, G. (2022) Australian home prices rose a record 24% last year. That’s insane, The Guardian, March 16. https://www.theguardian.com/business/grogonomics/2022/mar/17/australian-home-prices-rose-a-record-24-last-year-thats-insane
Prosper Australia (2022) Friendly Jordies on Housing Woes, by Charlotte George | Apr 13, 2022. https://www.prosper.org.au/2022/04/friendly-jordies-on-housing-woes/
REINZ (2022) Monthly Report of House Price Index. https://www.reinz.co.nz/residential-property-data-gallery
RBNZ (2020a) OCR reduced to 0.25 percent for next 12 months, Reserve Bank of New Zealand, March 16. https://www.rbnz.govt.nz/news/2020/03/ocr-reduced-to-025-percent -for-next-12-months
RBNZ (2020b) Reserve Bank Removes LVR Restrictions for 12 Months, Reserve Bank of New Zealand, April 30. https://www.rbnz.govt.nz/news/2020/04/reserve-bank-removes-lvr-restrictions- for-12-months
RBNZ (2020c) Reserve Bank Proposes Reinstating LVR Restrictions, Reserve Bank of New Zealand, December 8. https://www.rbnz.govt.nz/news/2020/12/reserve-bank-proposes-reinstating-lvr-restrictions
RBNZ (2021a) Regulatory Impact Assessment: Reinstating Loan-to-Value Ratio Restrictions, Reserve Bank of New Zealand, February. https://www.rbnz.govt.nz/-/media/ReserveBank/Files/regulation-and-supervision /banks/macro-prudential/LVR/Reinstating-LVRs-Risk-Impact-Assessment-02–2021.pdf
RBNZ (2021b) Loan-to-Value Ratio Restrictions, Reserve Bank of New Zealand, August. https://www.rbnz.govt.nz/regulation-and-supervision/banks/macro-prudential-policy/loan-to- valuation-ratio-restrictions
RBNZ (2022) New Residential Mortgage Lending by Borrower type — C31. Reserve Bank of New Zealand, 28 Jun 2022. https://www.rbnz.govt.nz/statistics/series/lending-and-monetary/new-residential-mortgage-lending-by-borrower-type
Yiu, CY (2021) Why House Prices Increase in the COVID-19 Recession: A Five-Country Empirical Study on the Real Interest Rate Hypothesis, Urban Science, 5(4), 77. https://doi.org/10.3390/urbansci5040077
