Real Assets: Understanding inflation shocks
Kurt Winkelmann and Raghu Suryanarayanan model asset returns under inflation shocks as a first step to designing genuinely inflation-sensitive portfolios – itself the first step towards broader macro-sensitive portfolios
Making informed asset allocation decisions involves assessing the impact of current and future macroeconomic conditions. Since inflation erodes nominal asset values, it is sensible for investors to treat inflation as an important component of portfolio risk – while today’s levels of current and expected inflation appear to be benign – even at 30-year lows – asset allocation is meant to anticipate future trends.
Although there are multiple scenarios to consider when modelling inflation risk, this article will focus on how we observe assets behaving differently under inflation stress, how we employ a quantitative model to disentangle inflation from other macro conditions, and how we use our model to build inflation-sensitive portfolios. In addition, our approach can be applied to other macro conditions with the same success.
We analyse potential increases in inflation in a structured way, starting with a definition of macroeconomic risk as a persistent shock to trends in, for example, real GDP growth or inflation. We then construct asset-pricing models that incorporate our definition of macro risk. Finally, we use the asset pricing models to assess the impact of macroeconomic shocks on asset returns.
How inflation hurts nominal bond returns
For government bonds, yield changes (which drive bond returns) come from either shocks to real output or shocks to inflation. Theory, data and practical experience combine to tell us that when real GDP growth contracts, real bond yields decline and the return to both real and nominal bonds should be positive.
Since nominal bonds are priced in nominal terms, bursts of inflation correspond to declines in nominal prices. Prices adjust downward to keep expected real returns constant; the real return of the nominal bond should be the same pre- and post-shock. (In fact, our analysis shows that the real return to nominal bonds is also negatively impacted by positive inflation shocks, reflecting a premium for inflation risk). Hence, we would anticipate nominal bond returns to be negatively affected by positive inflation shocks.
Using a simulation generated by our model, figures 1 and 2 show the impact on real and nominal zero-coupon bond prices of negative real GDP growth and positive inflation shocks over both one-year and five-year horizons. Three features stand out from the figures:
• The impact of inflation shocks on nominal bond returns is much larger than the impact of real GDP shocks;
• Real bond returns are positively affected by inflation bursts;
• Horizon effects are less important for nominal bonds; inflation hurts immediately and persists for five years.
Why would bursts of inflation be positive for real bond returns? If inflation and real growth are uncorrelated, then inflation would have no impact on real yields or real bond returns. Historically, however, bursts of inflation have been associated with lower real growth, and vice versa – US data show that historically an increase of the inflation rate by 2% has translated into a 30 basis point decline in the growth rate of real GDP, and qualitatively similar results are observed in other countries. This pattern has been incorporated into our model. If this pattern persists, then our model projects that higher inflation translates into lower real growth, which would translate into positive real bond returns. While figure 2 summarises the effects of inflation on real bond returns, the actual transmission mechanism is the analysis shown in figure 1.
Inflation influences equity returns over longer horizons
Our model focuses on analysing the impact of inflation on equities by looking at the impact of persistent inflation shocks on real dividend growth rates and discount rates, consistent with the basic concept that an asset’s value is the discounted value of current and future cash flows.
By approaching the problem this way, we are able to incorporate our definition of macro risk as persistent shocks to trend (inflation, in this case); to explicitly account for the impact of investment horizon on asset prices; and to separate the impact of inflation from other important drivers of equity performance.
For equities, we measure the impact of persistent shocks to inflation through their impact on real dividend growth rates. When we link real dividend growth rates to inflation and real GDP growth, it turns out that significant differences emerge across types of equity portfolios. These differences, however, only emerge over longer time horizons; the short-horizon effects are not appreciably different from one another. For example, over a five-year horizon, value portfolios are much more sensitive to real GDP growth shocks than are growth portfolios, as are small-cap portfolios relative to large-cap portfolios.
We can analyse the impact of inflation shocks by first looking at the impact on market-cap equity returns and then looking at sector and strategy portfolios. Figure 3 shows the impact of an increase in the trend inflation rate of 3% on equity returns over a five-year horizon. The initial impact of the inflation shock is muted, consistent with very low contemporaneous correlations between equity returns and inflation. However, over a five-year period a market-cap equity portfolio is projected by our model simulation to experience a 3% decline in real value. This suggests that in the early years, the dominant transmission mechanism is the impact of inflation on discount rates, while in the later years the transmission mechanism is dividend growth rates.
In figure 4 we explore the cumulative impact of a 3% increase in the average inflation rate on sector and strategy portfolios, again over a five-year horizon. We see the impact of persistently higher inflation on equities depends on the choice of sector and strategy.
Perhaps surprisingly, returns to the materials and consumer discretionary sectors are negatively affected by positive inflation shocks. Industrials and healthcare appear to have had positive returns for a positive inflation shock. Similarly, the quality, momentum and small cap strategies have been adversely affected by positive inflation shocks, while the value weighted, minimum volatility and risk-weighted strategies have been positively affected.
The figure shows the cumulative impact of a 3% permanent increase in US inflation over five years on the real return to the MSCI USA Sector and the MSCI USA Risk Premia indices over the same five-year horizon.
Our approach to understanding the impact of inflation on equity returns centres around understanding the dynamic links between the components of equity valuation and macroeconomic drivers. As shown by our figures, the impact of inflation shocks on market-wide equity returns is negative. However, our analysis also suggests that about 80% of the risk in equity portfolios is persistent shocks to real GDE growth. In other words, equities really are the growth asset.
Designing macro sensitive portfolios
How should an investor formulate an inflation-sensitive investment strategy? Our analysis suggests that this problem can be approached in a very structured way. First, investors can complement their views about future inflation by calculating the likelihood of this view, derived from a baseline quantitative model. Second, they can assess the possible effects of their view and the baseline view on cash flows and returns over multi-year horizons. Finally, since returns now reflect views about inflation, an inflation-sensitive asset allocation can be developed.
The same basic approach can be used to consider other macro variables, either in terms of standalone macro shocks, or as a combined macro outlook. In our view, designing an inflation-sensitive portfolio strategy is a component of the broader problem of designing macro-sensitive portfolio strategies.
Kurt Winkelman is head of risk & analytical research and Raghu Suryanarayanan is vice-president in research, asset allocation and macroeconomic risk at MSCI