Inflation differentials and equity returns
There is evidence that equity markets fail to adjust to persistent cross-country shifts in inflation in a timely and efficient manner. While equity investors focus on tracking firm-specific price effects and cash flows, they seem to pay less attention to aggregate local inflation and appear sluggish in adjusting long-term discount factors across countries. Since equity is a long duration asset even small calibration errors in discount factors have a large impact. Empirically, real equity returns in lower-inflation markets tend to outperform those in higher-inflation markets. No such effect can be found in fixed income markets.
The post ties in with the subject of “best practices for tracking macro trends”, particularly the case for cross-asset perspective, as explained on the summary page on macro trends.
The below are excerpts from the paper. Headings and some other cursive text has been added for context and convenience of reading.
Realized inflation and asset returns
“The below figure plots the time-series average of log nominal (real) returns (annualized) against the time-series average of log inflation in the left (right) panel at the 1-month horizon for portfolios sorted by the 1-month lagged year-over-year inflation rate… We sort countries into quintiles by lagged inflation…The upper panel plots nominal returns. The lower panel plots real returns. We plot stock returns (“diamonds”), bond returns (“circles”), and returns on T-bills (“stars”). The lines denote two standard error bands…The sample starts with nine countries in 1950, and ends with thirty countries in 2012.”
“Countries with higher lagged inflation do experience higher [nominal bond and treasury returns but not higher equity returns]…Real stock returns…decline almost one for one with realized inflation.”
“A large body of empirical literature on inflation hedging documents that real stock returns decrease after countries experience higher than average inflation.”
Equity markets seem to underestimate inflation
“We have established that the pass-through of expected inflation to nominal stock returns is slow and incomplete. The local component of expected inflation is a powerful predictor of real returns on stocks in the cross-section of countries: When a country’s expected inflation rate is higher than the global average, subsequent real returns and excess returns are lower…This is not true in the time-series dimension: When a country’s rate of inflation is higher than average for that country, this has a small, negative but statistically insignificant effect on real returns.”
“The price-dividend ratio is high when expected inflation is higher than usual, leading to lower subsequent expected real returns under the actual measure… The log price-dividend ratio [in the market] can be inferred from the discounted sums of subjective expectations…of nominal future cash flows…[minus]…nominal discount rates… Local stock market investors seem slow to adjust nominal discount rates in response to news about the future path of local inflation.”
“Given the large ratio of firm-level cash flow to inflation variance, it is natural that stock investors with a limited capacity to process information would update firm-level nominal cash flow forecasts more frequently than economy-wide inflation forecasts in discount rates… investors respond faster to firm-level price shocks on the cash flow side than to aggregate discount rate shocks.”
“Consistent with the sticky discount rate hypothesis, we find that the difference in real stocks returns between the lowest and highest inflation quintile countries is larger in countries which have only recently experienced high/low inflation.”
“As predicted by theories of rational inattention, agents reallocate resources towards inflation forecasting when inflation is high and volatile. In our sample, we find no evidence of stickiness in nominal stock returns in those countries with the most volatile inflation history.”
“In a…model of sticky information, we show that lagged inflation predicts lower real stock returns in the future. As inflation increases, investors are slow to update their nominal discount rates, leading to high current valuations and low subsequent real stock returns. Even small departures from rational inflation expectations deliver substantial real stock return predictability that is quantitatively similar to that in the data: Small mistakes have large effects because stocks are high-duration assets. Provided that inflation is sufficiently persistent, the model matches the return predictability in the data if 10% of investors fail to adjust the discount rate in any given year.”
Fixed income markets seem to be more attentive to inflation
“There is a large difference in the dynamics of the bond and stock returns in response to a change in expected inflation: Bond prices respond immediately to inflation news, but stocks respond slowly.”
“The nominal returns on local short-term and long-term government bonds respond much faster to changes in inflation. So do exchange rates. As a result, an increase in the local rate of expected inflation shrinks the local equity premium over bills and bonds, but not the equity premium on a basket of foreign stocks, because the local currency depreciates.”
“Given that inflation is not one of the main drivers of stock return variation, but it is for bonds and currencies, this finding may reflect rational inattention: Investors specialized in stock valuation may decide not to continuously monitor inflation, allocating limited bandwidth elsewhere, while bond and currency market investors do.”
“Thus, our results present a challenge to leading asset pricing models which impute rational inflation expectations to its agents.”