CHAPTER 3: A CHRONOLOGICAL REVIEW OF THE LITERATURE
3.2.1 The Empirical Evidence that Contributed to the Proxy Hypothesis
(Bodie, 1976; Jaffe and Mandelker, 1976; Nelson, 1976; Fama and Schwert, 1977) Fama (1981) states that a significant amount of evidence exists of a negative relationship between common stock returns and inflation between 1953 and the publication of his paper, in 1981. Fama (1981) referred to studies by Bodie (1976), Jaffe and Mandelker (1976), Nelson (1976) and Fama and Schwert (1977) as examples of studies that have documented a negative relationship between stock returns and both expected and unexpected inflation. Fama (1981) notes that these results are confusing when considering the previously accepted theory of the Fisher Hypothesis that advocated equities as an effective inflationary hedge.
In the study by Bodie (1976) the author used the framework of the Markowitz-Tobin-Mean- Variance model of portfolio choice in order to determine the effectiveness of common stocks as an inflationary hedge. Bodie (1976) measured this as the attainable reduction in the variance of the real return on a representative portfolio constructed of single period nominal bonds for which there was no risk of default. Bodie (1976) focused primarily on a minimum variance portfolio, from the perspective that that an inflationary hedge is basically taking the aforementioned bond as the initial point and combining it with other securities to reduce the variance of its real return as much as possible.
Bodie (1976) acknowledged that the definition of an inflationary hedge used in this paper is not the only possible definition and that the term has been employed in other distinct ways in the literature. Bodie (1976) referred to one such definition, by Reilly, Johnson and Smith (1971), which defined a security as an effective hedge against inflation if it offers “protection”
against inflation, in that it reduces the possibility that the real rate of return on the security will fall below a “floor” value such as zero. Bodie (1976) notes that studies that used this definition of an inflationary hedge often found that the nominal rate of return on common stock was frequently less than that of inflation and so it does act as an effective inflationary hedge, due primarily to the significant variance experienced by equity returns. The author refereed to a second definition of an inflationary hedge as a security that has a real return which is independent of the rate of inflation, as used by Branch (1974), Fama and Macbeth (1974) and Oudet (1973). This theory relates closely to the Fisher Hypothesis in that it proposes that because stocks represent a claim to real capital assets, which maintain a constant real value during inflationary periods, they maintain a real value which is independent of inflation – therefore the real value of stocks should match any changes in the rate of inflation. This is expressed as a positive relationship between the rate of inflation and the rate of return on equity.
Using a formula derived in the same paper which measures the effectiveness of equity as an inflationary hedge, Bodie (1976) proceeded to determine the magnitude of this capacity by determining the parameters on which the effectiveness of the capacity of common stocks as an inflationary hedge was dependent. The first of these parameters was the ratio of the variance of the non-inflation component of the real return on common stocks to the variance of unanticipated inflation – for which the author notes that the greater the ratio, the less effective equity was as an inflationary hedge. The second parameter was the discrepancy between nominal returns on nominal bonds and the coefficient of unanticipated inflation in an equation considering real equity returns. Bodie (1976) notes that the more significant this difference was in absolute value terms, the more effectively equity was able to act as an inflationary hedge.
Bodie (1976) used annual, quarterly and monthly data between 1953 and 1972 using the Consumer Price Index (CPI) as the inflationary variable and nominal returns derived from holding NYSE common stocks as the equity variable. Risk-free nominal returns were based on Treasury Bills with a matched maturity (Bodie, 1976). The results indicated that, contrary to the popular beliefs of the period, that the real return on equity is negatively related to both anticipated and unanticipated inflation. Bodie (1976: 469) states this to be a “somewhat
disturbing conclusion that to use common stocks as a hedge against inflation one must sell them short.”
Nelson (1976) states that in 1968 there was little doubt in both the academic and non-academic communities that the rate of return on common stocks moved directly with the rate of inflation in accordance with the Fisher Hypothesis. However, Nelson (1976) notes that over the years belief in the application of Fisher’s hypothesis to stock returns deteriorated as a result of the appearance of contradictory evidence. Nelson (1976) also investigated the relationship between common stocks and the rate of inflation and discovered evidence of a negative relationship between stock returns and both anticipated and unanticipated changes in the rate of inflation during the period. Nelson (1976) used a series of regression tests on monthly data between 1953 and 1972 in the United States, with the stock returns variable being based on a portfolio of common stocks and the inflationary variable being based on the consumer price index. To summarise the results obtained by the regression tests, Nelson (1976: 474) is quoted as writing:
“The most striking features of these results is the uniformly negative and generally statistically strong correlation between rates of return and inflation.”
Jaffe and Mandelker (1976) state that the relationship between the real interest rate and the inflation rate has been empirically tested with differing results further noting that relatively few studies directly examine the relationship between inflation and returns on risky assets. To address this, Jaffe and Mandelker (1976) used the Lawrence Fisher Index, which represented an equally-weighted portfolio of all of the securities listed on the NYSE as the stock price variable and the consumer price index (CPI) as their measure of the price level for the period 1953 to 1971. Using a regression analysis, in which stocks were defined as either a complete or partial hedge against inflation when the relationship between the variables was positive, i.e.
the coefficient of determination was positive, or were shown to be unable to act as a hedge when the relationship was negative, they found a significant negative relationship between inflation and common stock returns. They concluded that the stock market does not act as even a partial hedge against inflation, and that the diminishing effects on real wealth caused by inflation would be compounded by the low returns on the stock market.
These studies, among others, served as the primary motivation for the development of a theory that could explain the significant negative relationship between stock returns and inflation, which eliminated the stock market as an effective hedging option. Fama (1981) attempted to fill this gap in the literature with the development of the Proxy Hypothesis using a similar
regression analysis as previously discussed. Working with the hypothesis that the negative relationship between inflation and stock returns can be attributed to proxy effects, specifically that the negative relationship is a direct consequence of the negative relationship between inflation and real activity Fama (1981) conducted a series of empirical assessments. These tests provided evidence of a positive relationship between real stock returns and measures of real activity such as capital expenditures, the average real rate of return of capital and output.
Additionally, they determined consistent evidence of a significant negative relationship between the rate of inflation and these measures of real activity. Fama (1981) states that a detailed examination of the only regression which completely explains the effects on stock returns of expected inflation, that of the regression which includes the base growth rate, revealed that either the regression between real stock returns and either the base growth rate or the expected rate of inflation were spurious. Fama (1981) concluded that both common stocks and bond returns are determined in the real sector and so the proxy effect exists, but that spurious negative relationships between inflation and the expected real rate of returns are then caused by the fact that variations in real money demand in response to variations in real activity have been dealt with by offsetting inflationary variations instead of through growth in the nominal money supply.
The following studies have also attempted to address the issue of a lack of a clear and consistent methodology that measures the relationship between inflation and real equity returns. Due to the extent of the literature it would be impractical to review every study conducted. The studies which are presented below have been selected in order to demonstrate the conflicting views over the period prior to the development of cointegration theory.