The great normalisation of interest rates has only just begun

Back in June 2019, the Reserve Bank (RBA) Board meeting minutes stated that households are net borrowers in aggregate”, which contributed to their view that on balance “a lower level of interest rates was likely to support growth in employment and incomes and promote stronger overall economic conditions.”  I subsequently wrote to the RBA and sought clarification, given that households are not net borrowers in aggregate. Their response advised that the comment was in reference to interest-bearing net borrowings, rather than total net borrowings. Perhaps this was the intention of the authors, however the lack of this qualification in the minutes makes the statement incorrect – households are in fact large net savers in aggregate. 

The above point is important. There is currently a general perception that high levels of debt in the private sector preclude the RBA from materially lifting interest rates. This view would appear to be reflected in the current yield curve. Ten-year government bonds are yielding 1.9%, which remains below the RBA’s target range for inflation and the bond market’s expected long term inflation rate of 2.3%. Hence bond market pricing assumes monetary policy will remain exceedingly loose, with the general level of interest rates remaining below inflation for most of the next decade (i.e., bond markets are assuming negative real rates will continue). 

Underpinning bond market pricing and the RBA’s statement above is the conventional view that a higher level of interest-bearing liabilities than assets results in household spending being negatively correlated to movements in interest rates, i.e., lower interest rates will boost spending and higher rates will restrict spending. The efficacy of this relationship, however, has been brought into question over the past decade when the spiralling down of rates to record low levels failed to show any material boost to spending, or indeed inflationary pressures. In the decade ending December 2010, the cash interest rate averaged 5.2%, which coincided with an average rate of growth in real household consumption spending of 3.8%. In the following decade ending December 2020, the cash rate average had dropped to 2.2%, with household spending growth languishing at an average growth of 1.5%. A lower propensity of households to spend, despite the materially lower interest rates in place, is indicated by a rise in the average household savings ratio over the 2 decades from 1.9% to 7.5%.

Hence the relationship between interest rates and spending is complex and may not follow expected conventions, particularly, it would seem, when interest rates are very low. Provided below are 3 examples of factors that may have worked against lower interest rates stimulating spending in the way policymakers would have expected over recent years:

  • Clearly retirees living off interest income have their spending constrained by lower interest rates. However, in addition, those approaching retirement are required to accumulate larger savings in a low interest rate environment in preparation for lower interest earnings in retirement. So, whilst pre-retirement age groups may have very little by way of interest-sensitive assets (the majority is likely to be held in the form of property and equities), the lowering of interest rates may lead to lower spending and higher savings by this cohort.

 

  • At the other end of the age scale, first home buyers are typically required to save larger deposits once low interest rates are in place, given the tendency for housing prices to rise as interest rates fall. Again, this group’s spending may be negatively impacted by lower interest rates, despite them having a relatively low pool of interest-bearing assets (or liabilities) prior to making the purchase of a house.

 

  • Those households with existing loans are the group most expected to increase spending when interest rates fall. However, in reality there will be a mixed response to lower interest rates by borrowers. Many will maintain existing repayment levels and run-down loan principal. Others will hold larger balances in mortgage offset accounts to further reduce interest expense. Increasingly over recent periods, some borrowers have opted for fixed-rate loans, removing the conventional link between policy interest rate levels and spending levels. 

 

Given the spectacular lack of responsiveness of inflation to interest rates on the way down over the past decade, there must be a serious question over the extent to which the small increments in interest rates assumed by bonds markets in the years ahead will have the necessary moderating effect on inflation to bring it into line with target. Perhaps, therefore, larger interest rate increases will be required to have any material impact on spending and inflation.

 

 

How high can interest rates go?

A common argument used against the proposition that interest rates need to return to “normal” levels is that the level of debt has become so elevated that normalised interest rates could not be sustained by the real economy. Having purchased my own first property in 1990 and subsequently dealing with mortgage rates in the 15-16% range, the proposition that the economy has minimal capacity to absorb a cash rate not much higher than 0.1% does deserve some critical analysis!

Whilst government debt has clearly ballooned in recent years, borrowing by the private sector has been far more constrained. In fact, the relatively slow rate of expansion in private sector credit is one reason why loose monetary policy over the past decade has not been associated with higher spending and inflation. In September 2021, the level of private-sector borrowings outstanding was $3.2 trillion, or 1.5 times the annualised Gross Domestic Product (GDP) of the Australian economy. Ten years earlier, the ratio of borrowings to GDP was 1.4 times. Hence, there has been remarkably little growth in the relative size of credit outstanding given the magnitude of the decline in the price of credit (interest rates). 

Part of the explanation for the relatively muted rate of growth in credit is the high rate of principal repayment on loans, which has been facilitated in part by record-low interest rates. However, even if we ignore the improvement to household balance sheets made by existing borrowers and focus on the position of recent first home buyers (the group most vulnerable to rate rises), the evidence still suggests there is capacity to absorb interest rate increases. In the 10-years to 2021, the average size of loans approved for first home buyers has increased from $318,000 to $455,000. However, despite the high principal balance, lower interest rates have meant that the interest servicing cost on these new loans has fallen from $24,800 to $20,600 per annum over the same period (assuming standard variable home loan rates – currently 4.5%). At the same time, wages have increased, resulting in the interest servicing cost as a percentage of the average wage declining from 35% to 22%. It is apparent, therefore, that much of the pain of higher housing prices for first home buyers has been felt via a much larger deposit requirement, rather than the size of the loan repayment.

Hence, given that interest rates are a fraction of what they have been throughout past economic cycles, and that there has been only a very measured increase in private sector debt, there would appear to be considerable capacity for the private sector to absorb higher interest rates. If this is indeed the case, then the current policy of holding cash interest rates at an artificially low level is going to become increasingly difficult to justify. Economies have typically operated well when the prevailing level of interest rates is above that of the rate of inflation. Positive real interest rates provide the incentive for capital accumulation and subsequent investment in productive assets. A prolonged period of negative real interest rates will increasingly direct investments to real assets, such as property, whereby the stream of rental income, being inflation-linked, is destined to compound at a higher rate than the cost of capital (the interest rate). If rents rise faster than borrowing costs into the long term, almost any capital value for the real asset can be justified mathematically (Sydney real estate being a case in point). With these metrics in place, the incentive to invest productive assets, such as machinery, which depreciate over time, is reduced on a relative basis.

The long-term policy objective of central banks should be to return the general level of interest rates to being above the level of inflation. As would appear to be increasingly likely, long-term inflation is expected to return to the target 2% to 3% range. In the past, cash rates have averaged approximately 1% higher than inflation. In addition, typically, the shape of the yield curve is positive, with 10-year government bonds being an average 0.6% higher than the cash rate. As such, a cash rate of 3.5% and a 10-year bond yield of 4.1% should not be considered extraordinary. This is, however, a long way from current bond market pricing and, indeed, central bank guidance. 

With the labour market currently strong and job vacancies at record highs, the rationale for delaying the change to normality in interest rates is becoming less clear. Inefficiency in resource allocation, inequity in wealth distribution and lower rates of homeownership are some of the prices being paid for the current policy. These considerations will start to weigh more heavily on central banks in the absence of the case for “emergency settings”. Positive real interest rates are needed for long term economic efficiency, growth and equilibrium and should not be feared by central banks, nor indeed share market investors.

 

Brad Matthews 

Brad Matthews Investment Strategies