Horizon sensitivity of the inflation hedge of stocks

https://doi.org/10.1016/S0927-5398(00)00013-XGet rights and content

Abstract

In this paper, we study the potential of stocks as a hedge against inflation for different investment horizons. We show that stocks can be a hedge against inflation even if stock returns are negatively correlated with unexpected inflation shocks, and only moderately positively related to expected inflation. Depending on the investment horizon, the optimal hedge ratio can be either positive or negative. The crucial parameter for the results is the persistence of inflation.

Introduction

One of the biggest fears for investors is increasing inflation, because it reduces the real return on investments. As with any other risk in the financial market, the investor might want to try to reduce the risk exposure by adjusting the composition of the portfolio.

In this paper, we focus on the risk associated with inflation and the extent to which stocks can be used as a hedge against inflation. The theoretical basis for this strand of the literature is the Fisher hypothesis, describing the link between real and nominal returns. Applied to stocks, the Fisher hypothesis implies that there should be a one-to-one relation between expected nominal stock returns and expected inflation. However, in contrast to the Fisher hypothesis, many empirical studies observed a negative relation between inflation and stock returns.1 It seems, however, that there is an important horizon effect. More recent studies find a positive relation at a horizon of 5 years or longer.2 In this paper, we focus on the hedge potential and examine how it is influenced by the investment horizon.

An explanation for the short-term negative hedge potential of stocks is offered by Fama (1981), who argues that inflation simply acts as a proxy for real-activity variables in relations between inflation and stock returns. Higher expected economic activity would lead to an increase in stock returns, but due to the short-term non-neutrality of money, increasing inflation leads to lower economic activity and thus to lower stock returns.

In contrast to this negative effect, one would expect the Fisher hypothesis to hold in the long run. For example, Campbell and Shiller (1988) explain that inflation has two effects in a present value relation linking the stock price to the expected discounted future dividends. First, higher inflation increases the discount rates, which lowers returns. The second effect of increasing inflation is the rise of future dividends and therefore the rise of expected stock returns. Due to nominal price rigidities in the short run, the price elasticity of future cashflows is not necessarily equal to one. This means that the net effect is ambiguous in the short run, but will be positive in the long run.

Time series characteristics are important when we explore the horizon sensitivity. First, the volatility of stock returns is usually much higher than the volatility of inflation, which makes it difficult to test for correlation between stock returns and inflation. Even more important is the widely recognized strong persistence of inflation, probably related to inertia of monetary policies carried out by the central banks.

The horizon sensitivity is very important for investors who have to deal with inflation risk. An investor might not be interested in short-term performance at all. Institutional investors like pension funds have a very long horizon. Under a defined benefit plan, they often have inflation-linked liabilities with a duration of 15 years or more. It is therefore interesting to study the effects of changing correlation coefficients when the investment horizon increases. The changing covariance structure of stocks and inflation will affect the hedge potential of stocks. Horizon effects could lead to stocks being a poor hedge against inflation in the short run, but implying a positive long-run hedge ratio.

In this paper, we examine the portfolio problem of a long-term investor who faces inflation risk. We develop a model that accommodates the two following stylized facts: a short-term negative hedge ratio and long-term positive hedge ratio. The remainder of the paper is organized as follows. In Section 2, we set up the theoretical model. In Section 3, we discuss parameter values for the model. In Section 4, we present the results. In Section 5, we discuss the relations with other literature and in Section 6, we conclude.

Section snippets

The long-term inflation hedge potential of stocks

In this section, we use a stylized model of stock returns to explain how the hedge ratio will change with the investment horizon. We assume that the one-period stock return is generated by:Rt+1=c+βEtt+1]+φηt+1t+1The stock return Rt+1, which has a constant part c, depends on the expected inflation Et[πt+1], unexpected inflation ηt+1, and a specific risk term ϵt+1 with mean zero and variance σϵ2. The strength of the Fisher relation is represented by β, which measures the relation between

Parameter estimates

To illustrate the hedging potential of stocks in the model, we use benchmark parameters found in the literature. The parameters α, β and φ have been estimated for many different countries, sample periods and observation frequencies.

The starting point for empirical results for the inflation hedge regression model (Eq. (1)) are the studies of Bodie (1976) and Fama and Schwert (1977). According to their empirical evidence φ is almost certainly negative. The Fisher parameter β could also be

Results

Since the one period hedge ratio (Eq. (5)) does not depend on α and β, we can substitute the numbers for φ, ση and σϵ in the first line of Table 1 to get a hedge ratio of −2%, which means that the investor should short 2% of the portfolio in stocks to have the optimal protection against inflation with a 1-year investment horizon.

Table 2 presents the infinite horizon hedge ratio for different values of the inflation persistence and the coefficient for the Fisher hypothesis. If there is little

Discussion

The previous section showed that stocks have a hedge potential over longer investment horizons. It also showed how the hedge potential depends on key parameters like inflation persistence, the Fisher effect, and the inflation beta of stocks. In this section, we discuss a number of further issues and related work on the hedge potential of stocks.

The weight of stocks in the optimal portfolio is based on the assumption that the remaining part of the portfolio is invested in an asset that pays a

Conclusion

This paper showed that the negative inflation hedge potential of stocks can become positive if the investment horizon changes. In the model, a short-term negative sign is consistent with the positive hedge ratio implied by the Fisher hypothesis in the long run. The crucial parameter is inflation persistence. The higher the inflation persistence, the better the performance of stocks as a hedge against inflation. Even if the Fisher coefficient is only slightly positive, the inflation persistence

Acknowledgements

We would like to thank Casper De Vries, Clemens Kool, Christian Wolff, Franz Palm (the editor) and participants at the 1998 ESEM in Berlin and the JEF-LIFE conference on risk management for valuable comments. All errors remain the responsibility of the authors.

References (30)

  • J.Y. Campbell et al.

    Who Should Buy Long-Term Bonds?

    (2000)
  • N. Canner et al.

    An asset allocation puzzle

    American Economic Review

    (1997)
  • W.J. Crowder et al.

    The long-run relationship between nominal interest rates and inflation: the fisher equation revisited

    Journal of Money Credit and Banking

    (1992)
  • M.D.D. Evans et al.

    Do expected shifts in inflation affect estimates of the long-run fisher relation

    Journal of Finance

    (1995)
  • E.F. Fama

    Short term interest rates as predictors of inflation

    American Economic Review

    (1975)
  • Cited by (62)

    • Pension de-risking choice and firm risk: Traditional versus innovative strategies

      2022, International Review of Financial Analysis
      Citation Excerpt :

      Pension schemes often have long-time horizons, with new members likely to be drawing a pension many years later, and therefore need to make long-term investment decisions to meet their liabilities. In particular, the horizon sensitivity is very important for investors who have to deal with inflation risk (Schotman & Schweitzer, 2000). Hence, firms with longer pension plan investment horizons (i.e. maturity) are more likely to implement de-risking as they are exposed to a greater pension plan risk (Amir, Guan, & Oswald, 2010).

    • New evidence for the inflation hedging potential of US stock returns

      2020, Finance Research Letters
      Citation Excerpt :

      The first strand highlights the earlier studies which essentially cover the US and the findings do not seem to support the hypothesis (see Lintner, 1973; Oudet, 1973; Jaffe and Mandelker, 1976; Geske and Roll, 1983; Cochran and DeFina, 1993). The second strand attempts to resolve the perverse inflation hedge of stocks from methodological perspective by investigating the long-run hedging properties of stocks using long span of data (see Cagan, 1972; Boudoukh and Richardson, 1993; Solnik and Solnik, 1997; Lothian and Simaan, 1998; Schotman and Schweitzer, 2000; Lothian and McCarthy, 2001), and cointegration analysis (see Ely and Robinson, 1997; Anari and Kolari, 2001, 2010; Al-Nassar and Bhatti, 2019). We offer a different perspective to the generalized Fisher hypothesis.

    • Regulation and pension fund risk-taking

      2018, Journal of International Money and Finance
    View all citing articles on Scopus
    View full text