Climbing up Philip’s Inflation Curve

In the AFR I write that the most daunting challenge Australia’s central bank boss, governor Philip Lowe, faces is surmounting the formidable slopes of what is awkwardly known as the “Phillips Curve”. 

This basically depicts an inverse relationship between the economy's labour market slack, or unemployment, and consumer price pressures across the economy. As the jobless rate rises, inflation tends to decline. Conversely, as that spare capacity gets exhausted through, say, stimulatory fiscal and monetary policy, the bargaining power of employees improves, flowing through to stronger wages and ultimately consumer prices.

The last time Australia experienced a real bout of consumer price inflation in 2008, the unemployment rate had declined to about 4 per cent, which pushed wages growth up to 4 per cent. The Reserve Bank of Australia’s core benchmark for consumer price inflation, called the trimmed mean measure, jumped to almost 5 per cent, materially above its 2 per cent to 3 per cent target band. This was, in fact, the biggest overshoot since the target was adopted in 1993.

In an important 2019 paper on the RBA’s inflation forecasting models, my former Sydney University colleague, Natasha Cassidy, and three co-authors, demonstrated that the RBA’s Phillips Curve model does a decent job of explaining historical trend movements in inflation, including the spike around 2008.

(Cassidy was admired at university for besting another impressive RBA peer, Gianni La Cava, on the tricky topic of “hysteresis”, which was no mean feat given La Cava secured the penultimate place in his year behind the university medallist, who will remain nameless.)

As you would expect, the RBA’s Phillips Curve model is a little more complex than simply using the jobless rate to predict inflation innovations. Martin Place’s proxy for labour market slack is actually what it calls the "unemployment gap", or the distance between the current jobless rate (5.6 per cent) and what is unhelpfully described by economists as the "non-accelerating inflation rate of unemployment" (NAIRU). The NAIRU might be better characterised as the “wage increasing jobless rate” (WIJR). It is that very low level of unemployment at which point it is hard for companies to find workers, and this competition for staff results in firms bidding-up pay levels.

The problem is that the WIJR is constantly evolving. According to Cassidy et al, the WIJR has been falling since the early 2000s from an unemployment rate north of 6 per cent to about 4.5 per cent in 2018. Importantly, governor Lowe and deputy governor Guy Debelle think that the WIJR continued to edge lower and could now be in the 3s, as has been seen in other advanced economies like the US.

In addition to the magnitude of labour market slack, the RBA’s Phillips Curve inflation-forecasting model incorporates several other variables. These include changes in the value of imported goods, which can impact domestic inflation (cheap Chinese goods historically reduced local price pressures) and variables that represent inflation expectations.

Cassidy et al’s research suggests, for instance, that the community’s subdued expectations of inflation dragged down actual inflation outcomes between 2015 and 2020. (If you are not expecting big pay rises, you are unlikely to demand them.)

Decomposing the key drivers of inflation into different variables is important because it highlights that an overheating economy and/or inflation expectations can push-up consumer prices, amongst other things (eg, supply-chain bottlenecks and the reversal of temporary price cuts during the pandemic).

With the unemployment rates in Australia and the US---5.6 per cent and 6.0 per cent, respectively---still a long way above the RBA and the Federal Reserve’s best guess of the WIJR, which are both at, or below, 4 per cent, it is unlikely that wages growth will fuel inflation in the short-term.

But in a new research paper, our chief macro strategist, Kieran Davies, observes that there is less certainty about the influence of expected inflation, especially in the US where the government is delivering significant additional fiscal stimulus.

Central banks like the Fed and the RBA have been generally good at keeping inflation close to their circa 2 per cent targets. This hard-won inflation-fighting credibility has in turn kept inflation expectations well-anchored at around the same level.

“If a crack appeared in the Fed’s credibility and expected inflation started to respond to higher headline inflation – which should temporarily spike on the recovery in oil prices, pandemic costs related to supply chain disruptions and the reversal of price cuts made during lockdowns – this could feed into higher ongoing inflation,” Davies says.

One recent concern amongst investors has been the jump in the more timely, and widely-watched, 10-year bond market “breakeven inflation rate”, which is another measure of inflation expectations that has climbed from a pandemic low of 0.9 per cent last year to 2.4 per cent in March. This is its highest level since 2013.

In contrast to surveys of inflation, the breakeven inflation rate in the bond market measures the difference in yield earned on two types of 10-year government bonds. The first is just a normal 10-year government bond that pays you an interest rate that is the market’s best guess of where the central bank’s cash rate will be over the next decade plus a bit of extra interest to compensate you for the risk that this estimate is wrong (ie, it turns out to be different to the actual cash rate).

The second is an inflation-linked bond, where the principal is indexed to the change in the official consumer price index (CPI) over time. So as the CPI increases, you get more income.

The gap between these two yields provides an indication of the market’s best estimate of CPI inflation over time.

Davies points out, however, that life is not this simple. The 10-year government bond yield also includes a premium for inflation risk, which is the compensation investors expect for bearing the risk of higher prices over the next 10 years. And then inflation-linked bonds are impacted by their illiquidity, or low trading volumes, which mean that investors demand higher yields to compensate them for this risk.

Davies says that once you adjust for these two corrupting influences, “expected inflation has increased much more modestly from last year’s virus-affected low of 1.6 per cent to 2.0 per cent in March, which matches the Fed’s official target”. That is, the spike in breakeven inflation expectations is not as worrying as it seems.

While this is good news, there have been abrupt shifts in inflation expectations in the past. As former IMF chief economist Olivier Blanchard has observed, “the history of the Phillips Curve is one of shifts”.

To assess the potential impact of such a shock, Davies estimated the effect of a 1 percentage point increase in inflation expectations on US financial markets using a vector autoregression (VAR) model. This is a relatively large rise compared to the 0.6 per cent annual standard deviation of expectations in the US.

Davies’s key conclusion is that a material jump in inflation expectations would result in sharply higher interest rates and a large fall in stock prices. The process starts with tighter monetary policy to bring inflation under control, with Davies’ estimating that the Fed’s cash rate would climb by almost 2.5 percentage points (from near-zero today). Persistence in sticky inflation expectations and tighter monetary policy settings drives a big, 1.5 percentage point rise in the 10 year government bond yield. Higher discount rates hurt stock prices, which decline by about 15 per cent over the next 3 to 4 years in real, or inflation-adjusted, terms.

Lots of people have lazy opinions about inflation and interest rates. A more rigorous approach to thinking about these questions often furnishes superior insights.

Access Coolabah's intellectual edge

With the biggest team in investment-grade Australian fixed-income, Coolabah Capital Investments publishes unique insights and research on markets and macroeconomics from around the world overlaid leveraging its 13 analysts and 5 portfolio managers. Click the ‘CONTACT’ button below to get in touch