The residual market factor,the APT,and mean-variance efficiency

The rapidly increasing volume of both published and unpublished work on the arbitrage pricing theory (APT) of Ross (1976) has given rise to a number of misunderstandings at the interface of theoretical and econometric work. In this article we extend the theoretical structure of our previous work (McElroy and Burmeister, 1985, 1988; Burmeister and McElroy, 1987, 1988) to provide a broad yet rigorous framework both for econometric estimation and for better economic interpretation of new empirical results. We begin with the case where allK factors are observed, and then present the second case ofK−1≡J observed APT factors and one unobserved factor, theresidual market factor introduced in McElroy and Burmeister (1985). The economic interpretations for equivalent specifications of this model are discussed, and we enumerate several immediate payoffs to these specifications. The main new results are concerned with the sometimes intricate relationships among APT models withK factors and APT models withK factors that are constrained to satisfy mean-variance efficiency restrictions. These results are not only of theoretical interest, but more importantly they provide the basis for econometric estimation and testing of nested hypotheses. These econometric issues are discussed in detail.     

The Current State of the Arbitrage Pricing Theory

This paper provides a simple proof of a recent theorem presented by Reisman (1992) , concerning the use of proxies for the factors in the return-generating process of the arbitrage pricing theory (APT). In the single-factor case, the theorem asserts that any variable correlated with the factor can serve as the benchmark in an approximate APT expected return relation. The significance of this result is considered and a new direction for empirical work on “arbitrage pricing” is outlined.     

An Asset-Pricing Theory Unifying the CAPM and APT

This study shows that the competitive-equilibrium version of the APT may be extended to develop an exact model if idiosyncratic risks obey the Ross separating distribution. The results indicate that one only need add the market portfolio as an extra factor to the factor model in order to obtain an exact asset-pricing relation. Thus, this study presents an extension and integration of the CAPM and APT. The “empirical” APT is also generalized to allow for some factors to be omitted from the econometric model employed to test the theory. The developed model is extremely robust and may be reduced to the CAPM or expanded to approximate Ross's APT depending upon the number of omitted factors. Further, the importance of the market portfolio is shown to be a monotonic increasing function of the number of omitted factors. Finally, the study demonstrates that, in a finite economy, the pricing-error bound of the Ross APT in a correlated-residuals factor structure is an increasing function of the absolute value of market-residual beta, rather than the weight of the asset in the market portfolio as is the case of uncorrelated factor residuals. However, under the normality assumption, the pricing error becomes an extra component related to the market-portfolio factor, and the exact asset-pricing relation is once again obtained.     

Do Assumptions About Factor Structure Matter in Empirical Tests of the APT?

The effect of assumptions about factor structure on empirical tests of multifactor models such as the Arbitrage Pricing Policy Theory has received little attention in the literature. Using data on securities traded on the London Stock Exchange, we examine whether returns are best described by an approximate factor structure and whether assumptions about correlations across idiosyncratic returns have a significant impact on estimated prices of risk and their significance. Our findings suggest that returns are best described by an approximate factor structure and, if this is taken into account when empirically testing the APT, six factors carry significant prices of risk. However, if a strict factor structure is imposed, no factors carry significant prices of risk. These findings suggest that assumptions about factor structure matter in empirically testing the APT.     

Simulating and Forecasting Utility Stock Returns: Arbitrage Pricing Theory vs. Capital Asset Pricing Model

Just when the capital asset pricing model (CAPM) has become accepted by public utility regulators as a method for estimating a utility's screening rate, academic criticism of the model's theoretical and empirical shortcomings has led to empirical testing of the alternative arbitrage pricing theory (APT). This paper expands on recent APT-CAPM performance comparisons by simulating returns of public utility stocks using versions of both models, as was done by Bower, Bower, and Logue in a 1984 paper. In addition, the models are used for ex-post forecasting of returns in a subsequent time period. The Litzenberger-Ramaswamy method is used to correct for errors-in-variables in the CAPM cross-sectional equation. This allows for estimating the security market line using firm betas. The same methodology is used in the APT stages. Three different criteria—the Theil inequality, the sources of mean square error, and Chen's estimated weights of expected return-are used to compare CAPM and APT simulation and forecasting of the equity screening rates. Tested on a sample of 128 public utility companies, results show that neither model is clearly dominant. There is a tendency for reversal of performance. The model that is superior for simulating returns tends to be inferior for forecasting them, and vice-versa.     

Multivariate regression tests of the arbitrage pricing theory: The instrumental-variables approach

This article expands the theoretical basis upon which empirical testing of the arbitrage pricing theory (APT) rests. Specifically, it specifies linear restrictions for worlds in which the APT holds. These restrictions may, in principle, be tested. Since the regressors in the model are only “noisy” proxies for a specific linear transformation of the factors or mimicking portfolios, testing regressions suffer from an errors-in-variables problem. The standard econometric treatment for this problem is the instrumental-variables approach. A size-based example is employed to compare the test results derived from the instrumental-variables approach to those obtained via the ordinary least squares (OLS) method. The results from both methods cannot reject a two-factor APT for the size-sorted portfolio sample. The authors appreciate the helpful comments of Edwin Burmeister, Raymond Chiang, Steve Pruitt, participant at the 1989 Western Finance Association annual meetings, Indiana University, and University of Miami, and especially Shmuel Kandel.     

Equity prices and fundamentals: a DDM–APT mixed approach

This paper focuses on the linkage between equity prices and fundamentals for 27 individual shares belonging to the French stock price index (CAC40). To assess fundamental value, the traditional dividend discount model (DDM) is coupled with the arbitrage pricing theory (APT), which assumes that investors hold efficient portfolios. This yields a simple equity valuation relationship for which the APT determines the long-term risk premium included in the DDM. Accordingly, equity risk premia are determined by common factors reflecting the non diversifiable risk. These factors are not a priori identified by the theory, and therefore must be exhibited through an empirical analysis. Four domestic and three international common factors are found, all being among those identified by empirical analyses of the APT in the literature. While studies related to stock price indices showed that DDM fundamental values are very smooth compared to stock indices, our DDM–APT model reproduces both trends and major fluctuations of share prices. Further, as for studies based on stock indices, a mean-reverting process of equity prices towards fundamentals is highlighted, but the linear error correction model that was considered contains shortcomings suggesting a more complex adjustment process.     

The Arbitrage Pricing Theory: Is it Testable?

This paper challenges the view that the Arbitrage Pricing Theory (APT) is inherently more susceptible to empirical verification than the Capital Asset Pricing Model (CAPM). The usual formulation of the testable implications of the APT is shown to be inadequate, as it precludes the very expected return differentials which the theory attempts to explain. A recent competitive-equilibrium extension of the APT may be testable in principle. In order to implement such a test, however, observation of the return on the true market portfolio appears to be necessary.     

Some Empirical Tests of the Theory of Arbitrage Pricing

We estimate the parameters of Ross's Arbitrage Pricing Theory (APT). Using daily return data during the 1963–78 period, we compare the evidence on the APT and the Capital Asset Pricing Model (CAPM) as implemented by market indices and find that the APT performs well. The theory is further supported in that estimated expected returns depend on estimated factor loadings, and variables such as own variance and firm size do not contribute additional explanatory power to that of the factor loadings.     

Stock Return Anomalies and the Tests of the APT

This paper shows that the empirical tests of the Arbitrage Pricing Theory (APT) model are very sensitive to the anomalies observed in January in the stock returns data. There is a strong seasonal pattern in the estimates of the risk premia from the APT model. The most important implication of the findings in this paper is that the APT model can explain the risk-return relation mostly for January. Once the January returns are excluded from the data, there is no significant relation between the expected stock returns and the risk measures predicted by the APT model.     

An explicit bound on individual assets' deviations from APT pricing in a finite economy

Ross's Arbitrage Pricing Theory (APT) is a tractible and reasonable alternative to the mean-variance model. Nonetheless, understanding of the theory has been obscured by the complexity of the sequence economy models used for motivation. By contrast, we give a simple and direct derivation of the APT in a finite economy. Using an explicit bound on the deviations from APT prices across assets, a coarse calculation shows that theoretical deviations from APT pricing are negligible in our economy.     

Arbitrage Asset Pricing under Exchange Risk

This paper extends the APT to an international setting. Specifying a linear factor return-generating model in local currency terms, we show that the usual risk-diversification rule in the APT does not yield a riskless portfolio unless currency fluctuations obey the same factor model as asset returns. We then consider an arbitrage portfolio whose exchange risk is hedged by foreign riskless bonds. Under the resulting no-arbitrage conditions, the expected returns are not on the same hyperplane, unlike the closed-economy APT, unless they are adjusted by the cost of exchange risk hedging.     

Estimation of the underlying structure of systematic risk with the use of principal component analysis and factor analysis

We present an improved methodology to estimate the underlying structure of systematic risk in the Mexican Stock Exchange with the use of Principal Component Analysis and Factor Analysis. We consider the estimation of risk factors in an Arbitrage Pricing Theory (APT) framework under a statistical approach, where the systematic risk factors are extracted directly from the observed returns on equities, and there are two differentiated stages, namely, the risk extraction and the risk attribution processes. Our empirical study focuses only on the former; it includes the testing of our models in two versions: returns and returns in excess of the riskless interest rate for weekly and daily databases, and a two-stage methodology for the econometric contrast. First, we extract the underlying systematic risk factors by way of both, the standard linear version of the Principal Component Analysis and the Maximum Likelihood Factor Analysis estimation. Then, we estimate simultaneously, for all the system of equations, the sensitivities to the systematic risk factors (betas) by weighted least squares. Finally, we test the pricing model with the use of an average cross-section methodology via ordinary least squares, corrected by heteroskedasticity and autocorrelation consistent covariances estimation. Our results show that although APT is very sensitive to the extraction technique utilized and to the number of components or factors retained, the evidence found partially supports the APT according to the methodology presented and the sample studied.     

A Critical Reexamination of the Empirical Evidence on the Arbitrage Pricing Theory

This paper demonstrates that the Roll and Ross (RR) and other previously published tests of the APT are subject to several basic limitations. There is a general nonequivalence of factor analyzing small groups of securities and factor analyzing a group of securities sufficiently large for the APT model to hold. It is found that as one increases the number of securities, the number of “factors” determined increases. This increase in the number of “factors” with larger groups of securities cannot readily be explained by a distinction between “priced” and “nonpriced” risk factors as it is impermissible to carry out tests on whether a given “risk factor is priced” using factor analytic procedures.     

Reference Variables,Factor Structure,and the Approximate Multibeta Representation

The Arbitrage Pricing Theory (APT) implies that if asset returns have a factor structure, then an approximate multibeta representation holds with respect to the factors as reference variables. This paper assumes that asset returns satisfy a factor structure and derives a condition under which the approximate multibeta representation holds with respect to a set of reference variables which may not be the factors. This condition is that the regression matrix of the reference variables on the factors is nonsingular. Implications for the testability of the APT are also discussed.     

Joint Estimation of Factor Sensitivities and Risk Premia for the Arbitrage Pricing Theory

The APT is represented as a multivariate regression model with across-equations restrictions. Both observed and unobserved (latent) macroeconomic factors are included, thus generalizing and unifying two previous strands of literature. Large portfolios representing unobserved factors are treated as endogenous, and nonlinear 3SLS estimates are shown to differ sharply from estimates that ignore this endogeneity. Using monthly stock returns and six factors, we cannot reject January effects. The following results are invariant with respect to the inclusion of January effects: we reject the CAPM in favor of the APT; however, we cannot reject the APT restrictions on the linear factor model.     

THE ARBITRAGE PRICING THEORY AND COST-OF-CAPITAL ESTIMATION: THE CASE OF ELECTRIC UTILITIES

Capital asset pricing model (CAPM) and alternative arbitrage pricing theory (APT) methodologies are used to estimate the cost of capital for a sample of electric utilities. The statistical factors APT method is found to produce significantly different estimates depending on the number of factors specified and the set of firms factor analyzed. The use of macroeconomic factors is explored, and it is shown that this methodology has advantages over the statistical factors APT and the market model.     

No Arbitrage and Arbitrage Pricing: A New Approach

We argue that arbitrage-pricing theories (APT) imply the existence of a low-dimensional nonnegative nonlinear pricing kernel. In contrast to standard constructs of the APT, we do not assume a linear factor structure on the payoffs. This allows us to price both primitive and derivative securities. Semi-nonparametric techniques are used to estimate the pricing kernel and test the theory. Empirical results using size-based portfolio returns and yields on bonds reject the nested capital asset-pricing model and linear APT and support the nonlinear APT. Diagnostics show that the nonlinear model is more capable of explaining variations in small firm returns.     

The Pricing of Interest-Rate Risk: Evidence from the Stock Market

This paper addresses the issue of whether firms are required to pay an ex ante premium to investors for bearing the risk of interest-rate changes. A two-factor APT model with the market and changes in the yield on long-term government bonds as factors is employed. The paper shows that, empirically, most of the interest-sensitive stocks are in the utility industries, and that there is reasonable evidence that the interest factor is priced in the sense of the APT. Several sources for the interest sensitivity are considered, and regulatory lags are focused on as a likely candidate.     

A Semiautoregression Approach to the Arbitrage Pricing Theory

This paper developes a semiautoregression (SAR) approach to estimate factors of the arbitrage pricing theory (APT) that has the advantage of providing a simple asymptotic variance-covariance matrix for the factor estimates, which makes it easy to adjust for measurement errors. Using the extracted factors, I confirm the finding that the APT describes asset returns slightly better than the CAPM, although there is still some mispricing in the APT model. I find that not only are the factors “priced” by the market, but the factor premiums move over time in relation to business cycle variables.