The Reserve Bank of Australia (RBA) had markets moving this month, particularly the Australian dollar. It inadvertently did this when it put a number on the neutral rate of interest in Australia (the point at which policy is judged from being expansionary to contractionary, or vice-versa) in its June Board meeting minutes.
The number itself was of no surprise to economists, but given the demand from markets for hawkish fodder, as other central banks move to remove policy stimulus, this was interpreted somewhat as a statement of intent. It turned out to be a misinterpretation as the RBA qualified the discussion as being a theoretical one, with no implications for short-term policy settings.
Nonetheless, it did highlight a narrative that has recently emanated from several global central banks, as they contemplate removing policy accommodation. Indeed, the nominated neutral interest rate is now lower than it was previously and will be so for the foreseeable future.
The RBA itself noted in its minutes that its estimate of the neutral rate is around 3.5 per cent. The Reserve Bank of New Zealand followed up relatively quickly in a statement supporting this level in its context.
The Bank of Canada has been discussing the neutral rate for some time and re-iterated in its July 2017 Monetary Policy Report that: “The neutral nominal policy rate in Canada is estimated to be between 2.5 and 3.5 per cent.”
The US Federal Reserve (the Fed) also subscribes to this view in its quarterly longer-run forecasts that, given it assumes the economy will be running at potential over the longer run, can be interpreted this way.
All central banks have arrived at this approximate estimate from levels that were significantly higher. For example, in the January 2012 minutes of the Federal Open Market Committee (FOMC), it noted that “longer-run nominal levels [of the neutral rate] were in a range from 3.75 to 4.5 per cent”.
And the RBA itself estimated the neutral rate above 5 per cent in the pre-financial crisis, pre-resources bust period. Nonetheless, all of these estimates from central banks are just that: theoretical estimates.
No central bank knows what the neutral interest rate is – not until they reach it. Until then, they will be feeling the way with monetary policy as they remove stimulus. This potentially creates a risk to the economy should they overshoot. That is why we should expect that all central banks will remove current policy accommodation extraordinarily gradually – we are not in a ‘normal’ tightening cycle.
It is apparent an inflection point for this cycle has been reached in global monetary policy, with implications for bond yields and, significantly, risk-free rates. But lower neutral rates mean we clearly should not expect mean reversion. This is important for investors in terms of asset valuations, particularly those of longer duration.
Central banks note that it is lower potential rates of real GDP growth that are a key driver of a lower neutral rate outlook. This is as lower productivity growth, reduced labour supply (via deteriorating demographics) and reduced growth in capital inputs all weigh in this space. Indeed, neutral rates of interest and potential economic growth have likely declined in lock-step. Advanced economy potential growth, as estimated by central banks, has been in decline for the past two decades.
Our estimate of Australia’s potential economic growth rate is around 2.5 per cent per annum, slightly under the 2.75 per cent estimate of the Commonwealth Treasury and seemingly accepted by the RBA – noticeably higher than many other advanced economies.
Arguably, the reasons for this are that: (1) labour inputs have so far held up; and (2) we have had a huge run up in capital investment that is, for now, benefitting growth and flattering productivity. This is attributable to higher rates of population growth and the resources investment boom, respectively. A higher rate of potential economic growth and a higher inflation target than many advanced nations would suggest Australia’s neutral rate of interest should be higher than these economies. However, this may not be so, for at least two reasons.
The first is purely mechanical. Due to cost of funds increases for lenders (due to offshore increases in costs and those arising from regulatory and prudential changes), the margin between the RBA’s target cash rate and the effective rate borrowers receive is wider. In the mortgage space, this margin has widened from 180 basis points before the global financial crisis to around 370bp currently. For large businesses, this spread is from an average of around 230bp to now 425bp, and 132bp to 220bp on average for smaller businesses. The RBA has always said it would calibrate the level of the target cash rate to what the effective rate prevailing in the market is. This wider margin implies a lower cash rate will be required.
The second factor is largely a consequence of the actions of central banks themselves. And their maintaining aggressively accommodative monetary policies for an extended period. That is, high private debt levels and, in Australia’s case, exceedingly high household debt. Monetary policy has been relied upon too much and for too long to drag forward activity, consumption and, in particular, investment in the residential sector into the current period. Consequently, Australia’s household debt to GDP ratio is amongst the highest in the world. While this may be serviceable at current interest rates, it inevitably means that interest rates are a much more powerful tool than they have been in the past.
It is also leaves monetary policymakers with a much more asymmetrical policy tool in terms of its effectiveness. Easing policy further (if needed) would likely have a limited impact on the behaviours of households (outside of property) and businesses, and serve only to weaken the exchange rate modestly.
In contrast, raising rates will have a material impact on heavily indebted households, particularly on newer borrowers. Therefore, the RBA will be limited in how aggressively it can move to tighten policy (in the event this is necessary) due to the greater magnitude of the impact it will have – and also in terms of the level of the neutral rate due to the sheer amount of debt in the system.
For example, household liabilities have risen from $1.3 trillion just before the financial crisis to $2.3 trillion currently (expanding in excess of GDP, population, etc.).
Consequently, the level of interest rates at which these liabilities becomes more difficult to service is lower as the economic impact on household is now much larger. This factor itself should be considered in the context of a lower theoretical growth rate due to the impact of higher rates on household consumption.
In terms of the neutral rate, we may in the future be forced to consider it being lower should the current difficulties that central banks are having in generating inflation persist. That is to say, they will not have to rise as far or as rapidly to manage what could be only modest rates of inflation.
This may be a consequence of ongoing weak wages growth, the rise of technology and automation, and increased competition over an extended period. And although they may not acknowledge it, central banks may become a little more circumspect about their inflation targets and focus more on growth, full employment and financial stability.