Stagflation Explored
- 7 September 2022
- 18 min read
The term “stagflation” has increasingly been bandied about in the past few months. Sustained upward pressure on inflation (in some cases to levels not seen in decades), an increasingly strong monetary policy response and rising concern about the impact of tighter liquidity conditions on economic growth, explains the heightened interest. The significance of this should also be considered against the background of the disruption to economic activity and price levels by the worst global pandemic in a century and the Russia/Ukraine war, the first major cross-border conflict since the end of the Second World War.
This thought piece endeavours to unpack the concept of stagflation, its causes and its potential impact. We also share our view on the probability of this unfolding globally and, more specifically, in South Africa. This will be followed by our take on the most appropriate investment strategy in response to such an event.
WHAT IS STAGFLATION?
The term “stagflation” was first coined in the mid-1970’s in response to the simultaneous occurrence of economic stagnation, rising unemployment, and a rapid rate of inflation.
Stagflation can be defined in more than one way:
A period of rising inflation together with rising unemployment (the so-called twin evils of macroeconomics).
Persistent high inflation with high unemployment and stagnant aggregate demand.
A period when slow economic growth and joblessness coincide with rising inflation.
Referring to an economy that has inflation, a slow or stagnant economic growth rate and a relatively high unemployment rate.
The word “recession” does not explicitly feature in any of the above definitions. Stagflation simply refers to a period when economic activity slows down enough to cause job shedding, while the rate of inflation accelerates simultaneously.
WHEN DID THE WORLD FIRST EXPERIENCE STAGFLATION?
The world was introduced to this concept for the first time in 1973. This was followed by a second period of usage that lasted from 1979 to 1981.
At the time, stagflation contrasted with general economic experience to the extent that it was even (wrongly) believed by some that this phenomenon “defies the laws of economics”. Economic reference included the Great Depression of the 1930’s, when economic activity, employment and prices collapsed: the post-WWII experience of an inflationary boom, which had been followed by low inflation during the recession-prone 1950’s and 1960’s. However, the stagflation periods of 1973 - 1975 and 1979 - 1981 proved that unemployment and inflation could coexist in unhappy wedlock.
WHAT WERE THE CATALYSTS FOR THIS STAGFLATION?
Economic history suggests that these periods of stagflation were triggered by significant price shocks to critical production inputs and/or consumption goods. During 1973, the price of crude oil quadrupled as a direct consequence of the war between Israel and its Arabian neighbours. This coincided with poor grain harvests around the world both in 1973 and 1974, which caused the prices of other raw materials to skyrocket. Furthermore, wage and price controls were lifted in the US during late 1973/early 1974 as the country eliminated the price control programme that had been introduced two years earlier by the Nixon administration. This contributed to widespread price increases on the products and services that had been capped by price controls. Not only did the above lead to US inflation reaching 12% during 1974, but also contributed to a surge in inflation around the world. Meanwhile, real US Gross National Product (GNP) growth began to slide in 1974, eventually turning into the most severe recession since the 1930’s, with the unemployment rate rising from 5% to 9%.
The period 1979 – 1981 experienced a repeat of the above results (although the causes differed), with another spike in crude oil prices, this time caused by the Iran revolution.
A THEORETICAL EXPLANATION OF STAGFLATION
Theoretically, stagflation is best illustrated by the Phillips curve. The Phillips curve (introduced by A.W. Phillips in 1958) is mainly used to explain the relationship between inflation and the unemployment rate. In a well-functioning economy - specifically in the absence of price controls and disruptive organised labour - a drop in the unemployment rate will, all else constant, give rise to higher wages and eventually inflation. The Phillips curve suggests that stagflation can occur when inflation expectations in the well-functioning economy adjust to rising and higher inflation, which, in turn, forces the unemployment rate higher until a new equilibrium level between inflation and the employment rate is reached.
Figure 1: The Phillips Curve and Stagflation
Source: Macroeconomics (Dornbusch & Fischer) 3rd print, 1985, Futuregrowth
The opposite also holds; a rise in unemployment leads to lower wage increases and eventually lower inflation. However, it is important to note that this relationship tends to be asymmetrical. In other words, wages rise in response to a drop in the unemployment rate (or excess demand for labour) and/or higher inflation, but do not fall at the same rate in the case of rising unemployment and/or disinflation. This is illustrated by the kinked curve (C) in Figure 1. The relationship is even more asymmetrical in an economy like South Africa’s, where the structure of the labour market does not allow for much flexibility, if any at all.
While the Phillips curve is based on the relationship between inflation and the unemployment rate, the intimate link between economic activity and the labour market allows for the curve to be extended to economic growth.
HOW DO WE RESOLVE STAGFLATION?
While standard economic theory provides an explanation for stagflation, finding a policy solution that is reliable and economically painless, and can be applied broadly, is more problematic. Theoretically, the most obvious solution is to increase economic output by way of productivity gains. Since higher productivity implies higher output and, by implication, lower unit costs, this (all else constant) also leads to higher economic growth but at relative price stability (perhaps even lower inflation). This productivity-induced acceleration in economic activity would allow monetary authorities to raise policy rates to tame inflation. Of course, this solution assumes a well-functioning economy with a highly skilled labour force that can easily be retrained or replaced by (for instance) a higher level of automation.
Economic history offers an example of a real-life solution. The most prominent is the US policy response to the 1979 – 1981 period of stagflation. At the time, the US Federal Reserve (under Chairman Volcker who intentionally wanted to shake off the 1970’s stagflation denialism) raised interest rates to levels high enough to break a vicious inflation cycle. Of course, this came at a price, as economic activity slowed, and jobs were shed. In the end, the bitter monetary medication had to be administered to stop the snowball effect of subsequent rounds of price increases. This created a sound base for sustained economic growth, rising employment, and low inflation in the subsequent decades. One caveat is that central banks require political backing.
Reducing unemployment is a different matter altogether. The level of ease in resolving stagflation varies between well-functioning economies with a flexible and highly skilled labour force and economies like South Africa that sadly battle with structural issues, such as an inflexible labour market and poor skill levels.
IS THE GLOBAL ECONOMY SLIPPING INTO STAGFLATION?
Judging by market and media commentary, the fear of stagflation has not been this elevated in decades. While the current rate of inflation, both at producer and consumer levels, is deemed excessive in several advanced and emerging countries, economic activity and labour markets are not yet at levels that are synonymous with stagflation. That said, if consensus forecasts are anything to go by, it appears that the general flow is heading towards a higher inflation/lower growth environment. This point is illustrated in Figure 2 below. Increasingly hawkish monetary policy response to elevated inflation is feeding expectations of lower economic growth, in some instances, even fear of recessionary conditions.
Figure 2: The global economy is heading closer to stagflation territory
Source: Bloomberg, Futuregrowth
What caused the current stagflation scare?
Broadly speaking, the current scare started with the sustained global supply side price shock that was sparked by COVID-related hard lockdowns, as supply chains were disrupted. More recently, this supply chain disruption (which is still in the process of normalising) received another setback from events in eastern Europe. The Russia-Ukraine war led to sharp increases in crude oil, gas, and grain prices. This caused significant upward pressure on energy and food prices and boosted already elevated inflation to levels not seen in decades in many of the advanced economies. With the word “transitory” now discarded by central banks, the acceleration in the rate of inflation forced a belated concerted monetary policy response. This, in turn, has raised broad-based concerns that the fragile post-COVID economic recovery would also suffer.
WHERE DOES THIS LEAVE SOUTH AFRICA?
Overall, South Africa is moving in the wrong direction, even though the rate of increase in consumer inflation is lagging that of several emerging and even advanced economies. However, it could be argued that the country has been much closer to periods of stagflation in the past, as illustrated in Figure 3 below. A concerning fact is that the country’s economy has been growing at a tepid pace since the mid-2000s, while it has also been faced with high structural unemployment. By implication, it would not take much to fall into a stagflation trap. Being a small, open economy, this increases the risk, should the globe move closer to such an economic outcome.
Figure 3: South Africa has experienced episodes of stagflation in the past
Source: Bloomberg, Futuregrowth
WHAT ARE THE POTENTIAL IMPLICATIONS OF STAGFLATION FOR PUBLIC SECTOR FINANCES?
All else equal, lower real GDP growth would negatively impact the fiscus in the form of lower tax receipts. Our calculations (which consider the relationship between real GDP growth and annual tax revenue growth) indicate a strong positive relationship between the two variables. The effect of lower GDP usually impacts tax revenue with a lag. Our analysis suggests that it usually takes up to nine months for the full effect of the lower real GDP growth to be assimilated and reflected in tax revenue receipts, more so for income-type taxes such as personal income and corporate income tax. Consumption taxes, such as value-added tax tend to respond more quickly to changes in real GDP growth.
Figure 4: A strong positive relationship exists between real GDP and tax revenue growth
Source: Bloomberg, Futuregrowth
Figure 5 below shows the correlation between real GDP growth and growth in personal income, corporate income and value-added tax (VAT) when lagged over various periods. The relationship between changes in real GDP growth and personal income tax is strongest when we lag personal income tax by six months. Similarly, corporate income tax has a peak positive correlation to real GDP when lagged by nine months. Value added tax tends to exhibit a weaker correlation as we increase the lagged period, indicating a diminished effect as time passes.
Figure 5: Lower real GDP growth tends to impact revenue growth 6-9 months down the line
Source: National Treasury, Futuregrowth
On the expenditure front, wage growth is more likely to come under pressure in response to higher inflation and, by consequence, high inflation expectations. South Africa’s wage bill accounts for approximately 30% of consolidated government spending and has in the past grown at rates in excess of inflation. Typically, higher inflation expectations tend to trigger higher wage demands, which puts more pressure on government expenditure. South Africa’s most recent stagflationary period (which was triggered by the 2008 global financial crisis) is instructive. Although inflation peaked at 11.3% on a fiscal year basis, compensation growth increased materially more during the lead up, and continued to grow by double digits despite the aggressive moderation in headline inflation subsequent to the peak. Between fiscal year 2004/05 and fiscal year 2012/13, compensation expenditure almost tripled from R128.3 billion to R370.8 billion. Had the compensation budget been adjusted to account purely for inflation, government would have spent a materially lower amount of R221.8 billion by fiscal year 2012/13 on public-sector compensation.
Figure 6: Fiscal compensation expenditure growth vs Headline CPI (y/y)
Source: National Treasury, Futuregrowth
Occurrences of a stagflationary environment domestically have historically had the effect of impairing revenue and increasing expenditure. All else equal, this would put additional pressure on an already stretched fiscal position, more so if the potential stagflationary cycle becomes persistent and entrenched.
HOW WOULD THE THREE INTEREST RATE BEARING ASSETS CLASSES PERFORM UNDER STAGFLATION?
To assess how the three domestic interest-bearing asset classes might perform in an environment characterised by below potential macroeconomic growth and elevated inflation, it is instructive to review the fundamental return drivers across the asset classes as well as the empirical findings on the performance of the interest-bearing asset classes. Empirically, assessing returns from an inflation adjusted perspective to ensure the comparability of our results across varying macroeconomic backdrops, we find an intuitively pleasing result when only considering the returns across the asset classes under “normal” macroeconomic conditions. In this environment, incorporating monthly index return data from March 2000, cash is the worst performing interest-bearing asset class with a modest annual real return of 1.6% relative to the 3.5% real return for inflation-linked bonds and the marginally better 3.6% return for nominal bonds. The linearity in the relationship between risk and return is met under normal market conditions domestically, with low-risk cash underperforming more risky bonds.
In an environment of sub-par macroeconomic growth and elevated inflation pressure, we unsurprisingly find that cash (0.95%) yields the best inflation-adjusted annual return across our investment universe. While there is only a modest performance differential between the risk assets, we find that nominal bonds (0.9%) marginally outperform inflation-linked bonds (0.87%).
Figure 7: Inflation-adjusted interest rate asset class performance under stagflation and non-stagflation periods
Source: Futuregrowth
Our findings suggest that on an inflation-adjusted basis, cash remains a suitable refuge from capital risk in an environment of heightened inflation pressure and sub-par macroeconomic growth. The evidence from inflation-adjusted nominal and inflation-linked bond performance, on the other hand, is consistent with our historic observation of the elevated inflation breakeven rates in the domestic bond market, and the marginal capital protection this can provide to nominal bonds in periods of market stress.
HOW WOULD WE POSITION FOR STAGFLATION?
Based on the current term structure of interest rates, and considering our analysis above, we would have an allocation preference for cash relative to nominal and inflation-linked bonds in our Interest Rate Asset Allocation funds - if we had a high conviction that a sustained period of domestic stagflation was imminent.
In terms of our allocation preference for bonds, while we would expect inflation-linked bonds to outperform nominal bonds in inflationary periods, all else equal, our analysis suggests that the additional effect of sub-par macroeconomic growth (and the bearing this has on the fiscus) impairs the return on sovereign bonds broadly. This somewhat erodes the appeal of inflation-linked bonds as a defensive trade (relative to nominal bonds) in a period marked by stagflation risk. Given the current term structure of interest rates across both the nominal and inflation-linked bond markets, (that are characterised by a steeply sloped nominal bond curve and elevated inflation breakeven rates) we would maintain a marginal allocation preference for medium-dated nominal bonds relative to inflation-linked bonds in a stagflation environment.