The Fed's lessons on property for Australian banks

In our view the major Australian banks offer very little valuation upside. In fact we are becoming increasingly concerned that Australian banks are collectively embracing the view that inflation will stay lower for longer, that mortgage serviceability will be at least maintained over the medium term and that property investments will remain prospective. Australia’s four largest banks account for nearly 30% of the ASX 200 Index. The K2 Australian Fund holds just one major bank, ANZ, and its weight in our portfolio is a mere 1%. 

We are worried that the banks are growing complacent about future problem loans. We are of the view that every action has a consequence and the recent out-of-cycle rate rises for property investors should be viewed negatively. If the banks pass the costs of the new bank levy on to borrowers then we believe that many recent property investors will transition towards a position of negative equity. Unfortunately, the banks are not prepared for this outcome.

We have been monitoring CBA and Westpac’s loans that are more than 90 days in arrears; as of March 2017 it was $5.97 billion for the two banks combined. What we find most interesting is that these arrears have risen by 38% since the start of 2015. Interestingly, over the same time frame the standard variable home loan interest rate fell by 12%. We wonder how these loans will perform if residential property markets become more challenging and interest rates are less favourable.

Lessons from the Federal Reserve

During the past year or two, there has been a tsunami of commentary regarding the perilous state of the Australian housing market. Most have pointed to elevated levels of household debt to disposable income, property prices relative to wages and general affordability for first-time home buyers. We think the most sensible way to assess property values was spelt out by the US Federal Reserve (FED) back in June 2005. Their analysis concluded that rental payments in the housing market are analogous to dividends in the stock market.  Hence the price of a house should reflect an appropriately discounted stream of expected rents. Using this framework the FED’s 2005 analysis concluded that:

• The June 2005 price-rent ratio of 27x was the highest in 35 years
• When price-rent ratios are high subsequent prices changes are lower than usual
• House prices in June 2005 were about 20% too high relative to rent.

The FED also considered loan to valuation ratios (LVR). Their work suggested that 82% of borrowers had an LVR below 80% and 18% had LVRs over 80%. They estimated that 4% of borrowers were highly leveraged and could lose all of their home equity if house prices were to fall 10%. In addition to high price-rent ratios in 2005 the FED was also concerned about the sudden popularity of interest-only mortgages. US interest-only mortgages had increased sixfold between 2003 and 2005 but were a relative small part of the mortgage market with a share of about 4%.

The FED subsequently lifted interest rates by 175 basis points between June 2005 and March 2006. The FED consistently referred to higher energy prices, pressures on inflation and improving labour market conditions. US house prices continued to rise and by March 2006 the price-rent ratio hit 28x. Despite low LVRs and relatively low levels of interest-only borrowing, US property prices still declined 32% over the next 5 years. The decline in home prices caused mortgage delinquencies to rise from 4.3% in 2005 to 10.0% by 2010.

The present state of Australian housing

Using the FED’s framework, Australian property is currently priced at over 30x rent. We estimate that 87% of the borrowers from the four major Australian banks have an LVR below 80%. Mortgage delinquencies in Australia are presently less than 0.5%. As a result, Australian banks have reduced provisions for future bad debts to just 0.4% of loans outstanding. However, 37% of major Australian banks’ housing mortgages are interest-only and these are predominantly classified as “investment” loans. At over 30x rent we would classify these loans as “speculative”. So what could go wrong?

Price discovery

Back in October 2016, the Reserve Bank of Australia stated that “the foreshadowed risk of oversupply in some apartment markets is nearing as a large volume of new apartments has started to come online.”  

It would appear that the supply of Melbourne and Brisbane apartments will increase by nearly 18% over the coming 24 months. Given this backdrop, we are concerned that rental growth will be non-existent for the recent property speculators. In fact, we believe that occupancy rates will be challenged over the coming years and rental yields may fall significantly below funding costs. Ultimately, bankers will be having some very difficult discussions with Melbourne and Brisbane apartment landlords over the medium term.

We feel that the expertise of Australian bankers will be tested over the coming year or two. In Australia, there are currently 1.7 million interest-only mortgages. In addition, nearly 15% of investment loans already exhibit LVRs over 80%. If apartment prices in Melbourne and Brisbane fell by more than $75,000 over the coming year then 100,000 speculators could move towards positions of negative equity. We wonder whether Australian bankers are sufficiently resourced and experienced enough to mitigate the vulnerabilities surrounding a property related negative feedback loop.