Bubble trouble. Whether we label them bubbles, the Australian economy has experienced a series of developments that potentially could have the economy lurching from boom to bust and back. In recent years these have included:
the record run up in commodity prices and subsequent correction;
the associated boom in mining investment and current reversal;
record low bond yields;
the boom in housing construction, specifically apartments, that was spurred by the low interest rates;
and also at least partly related to low interest rates, but also to record foreign investment and other factors, rapid and sustained house price inflation. Investor concerns. Indeed, some investors, especially those overseas, worry that the bursting of record commodity prices, house prices, apartment approvals, household debt and mining investment would derail the economy. Yet so far it has proved remarkably resilient. Measurement. This note explores these issues by calculating what we call a bubble meter and its implications for the economy. To do so we:
detrended the five components of the terms of trade, mining investment, house prices, housing approvals and interest rates.
created Z scores (the number of standard deviations from the mean) for each series to gauge how deflationary or stimulatory it is for the economy.
equally weighted the 5 components to construct an overall bubble meter. The construction of the bubble meter is eclectic and therefore isn’t a definitive representation of bubbles. A number of other developments could have been included in the bubble meter, one obvious one being the record level of household debt. However, debt has so far shown no tendency towards mean reversal. It may also be proxied to some extent by house prices. To detrend each component we used a moving average technique but other methods could have been used. Detrending and calculating Z scores is not the only way of defining and measuring bubbles. And the choice of equally weighing the components was purely pragmatic given that there was no theoretical basis to weight components that included commodity prices, asset prices, interest rates and real variables.
What the bubble meter shows. Figure 2 shows our aggregate of the 5 components of the bubble meters over the past 20 years. The key points include:
The bubble meter has fluctuated between one half and one Z score over the 20 years.
Most of this time has been spent in stimulatory territory. This has reflected the rising trend in commodity prices and house prices and downward trend in bond yields.
The GFC saw the bubble indicators dip temporarily but they did not fall into deflationary territory
Over the past 20 years there have been only two episodes where the bubble meter has been deflationary. One in 2000 and it currently is now.
The deflationary episode in 2000 reflected a tightening cycle by the RBA, which together with the introduction of a goods and services tax, saw a cyclical downturn in housing activity, and falling mining investment.
The current deflationary episode is of a similar magnitude to that in 2000. But it is different to then. Global factors are more important this time.
In addition to the correction in mining investment, there is a deflationary income shock as the terms of trade have reversed from record levels and, partly offsetting that, housing construction is booming as the RBA has cut rates to record lows (Figure 3).
The length of time the bubble meter could spend in bust territory will depend largely on how much further mining investment and the terms of trade has to fall and whether the elevated level of apartment approvals and house prices reverse sharply. Our contention is that it will continue to signal a bust for longer than in the 2000 episode.
Economic outlook implications.
Neither the “bursting” of the bubble meter in 2000 or in 2015-16 has led to a hard landing in the economy (Figure 4). We offer some reasons why.
Z scores of between -½ and 1 do not signal large (standard) deviations from the mean. Generally, Z scores of 2 and above would be considered extreme. Therefore on our calculations of the bubble meter the shocks to the economy are not sufficiently large to materially destabilize the economy.
There also are good reasons to believe that regardless of the size of the shocks there are limitations to the spillovers from bubble prone variables into the broader economy. The AUD for example traditionally has played a key role as shock absorber to external developments. The RBA actively manages cycles. And governments usually have allowed fiscal automatic stabilisers to provide a countercyclical role. It also is the case that in the current boom and bust in commodity prices there has not been the hubris or exuberance that could have been expected. Household saving actually rose for example. Underlying that absence of hubris could be the cautionary global backdrop and the regulatory framework provided by APRA, ACCC and others.
Having said that, the current situation carries some risks. Housing indicators in the bubble meter are at record highs but interest rates remain at record lows. Typically monetary policy is well into tightening mode at this stage in the housing cycle. A destabilizing housing burst (both in activity and prices) is a clear risk, particularly the longer the upswing runs.
External shock equals “pop”.