A Unified Model of Investor Utility and Asset Pricing

Traditional asset pricing theories and models are facing more and more challenges in empirical study. Machine learning provides a new tool for asset pricing research. Due to the low signal-to-noise ratio and concept drift of financial data, the theoretical constraints of economics are very important for the applicability of machine learning in asset pricing. Firstly, this paper introduces seven multi-factor asset pricing models based on ad hoc sparsity constraints, summarizes the characteristics and shortcomings of traditional asset pricing models. Then, we display the challenges of machine learning facing in empirical application of asset pricing, formulate the targeted economic constraints. Finally, we further discuss the possible future trends of machine learning algorithms in asset pricing.

Journal of Accounting Advances, (2020) 12(1): DOI: 10.22099/JAA.2021.39285.2077 Journal of Accounting Advances (JAA)Journal homepage: www.jaa.shirazu.ac.ir/?lang=en Developing Fama and French Multi-Factor Pricing Model Using a Fundamental Factor Based on Accounting CharacteristicsSanaz Aalamifar1, Abdollah Khani2*, Hadi Amir3 1. Ph.D. Candidate, Department of Accounting, Faculty of Administrative Science and Economics, University of Isfahan, Iran. sanazaalamifar@ase.ui.ac.irCorresponding author, Associate Prof., Department of Accounting, Faculty of Administrative Science and Economics, University of Isfahan, Isfahan, Iran. a.khani@ase.ui.ac.irAssistant Prof., Department of Economic, Faculty of Administrative Science and Economics, University of Isfahan, Isfahan, Iran. h.amiri@ase.ui.ac.ir ARTICLE INFABSTRACT Received: 2020-12-21Accepted: 2021-04-03 The purpose of the present research is to introduce a fundamental factor, based on related accounting characteristics (including earnings to price, book to price, sales growth rate, accruals, investment and growth in net operating assets), as a factor in the structure of Fama and French asset pricing model. The mentioned factor has been deducted from consumption theory and accounting principles and assumptions. In order to test the hypotheses, data of 345 companies listed in the Tehran Stock Exchange (TSE) and Iran Farabourse market, during the period 2006 to 2020, were used. To evaluate the performance of the multi-factor asset pricing model, test assets were ranked in two categories (once considering expected return characteristic, and once without considering the company’s expected return characteristic). In the following, using time series regression approach, the performance of augmented multifactor asset pricing models and corresponding conventional ones are compared. The results of this research showed that development of the research models with the fundamental factor based on mentioned accounting characteristics, can lead to improving the performance of these multi-factor models in explaining the variation in (expected) stock returns, and the test assets that considered the company’s expected return performed better compared to those that did not. The findings of this study indicate that the information in the financial statements has information content and can play an undeniable role in determining the expected return. * Corresponding author: Abdollah Khani Associate Prof., Department of Accounting, Faculty of Administrative Science and Economics, University of Isfahan, Isfahan, Iran. E-mail: a.khani@ase.ui.ac.ir 1-IntroductionIdentifying the correct asset pricing model has long been an important topic in the thematic literature of financial economics. Such a model not only explains stock returns, but also increases the ability to predict abnormal returns. The first models for estimating returns date back to the 1960s, when Markowitz’s (1952) new theory of securities attracted the attention of researchers. The first model for estimating returns was the capital asset pricing model (CAPM) which was presented by William Sharp (1964). In his research, William Sharp showed that return on asset was a function of line of market risk premium. But from 1975 to 1990, deviations and anomalies related to the CAPM model gradually became apparent. Following the recognition of these anomalies in accounting, in this study, based on the research of Penman and Zhou (2018), a fundamental factor based on accounting characteristics is introduced. For this purpose, consumption theory and accounting principles and assumptions will be used for initial identification; and empirical tests will be used for final identification of accounting characteristics that affect earnings growth and expected returns. Then these identified characteristics are summarized in a factor called the fundamental factor. Therefore, the purpose of this study is to evaluate the possibility of improving the performance of the asset pricing factor models in explaining the stock returns by adding a fundamental factor based on accounting characteristics. The hypothesis, methods, result and discussion and conclusion have been explained below. 2-HypothesisThe aim of this research is to introduce a fundamental factor based on accounting characteristics as a factor in the structure of the Fama and French asset pricing models. For this purpose, data of 345 companies, listed in the Tehran Stock Exchange, during the period 2006 to 2020 have been used. In order to achieve the objectives of this research, the following hypothesis are developed:H1: Adding a fundamental factor, based on accounting characteristics, to Fama and French three-factor model, improves its performance in explaining the stock returns.H2: Adding a fundamental factor, based on accounting characteristics, to Carhart four-factor model, improves its performance in explaining the stock returns.H3: Adding a fundamental factor, based on accounting characteristic, to Fama and French five-factor model, improves its performance in explaining the stock returns.H4: Adding a fundamental factor, based on accounting characteristics, to Fama and French six-factor model, improves its performance in explaining stock returns. 3- MethodsThis is an applied research in terms of purpose and an inferential and descriptive research, in terms of method. For data analysis and hypothesis testing, the data have been collected from 345 companies listed in the Tehran Stock Exchange for a period of 15 years (2006 to 2020). For initial calculations, the Excel and Ox Metrics software tools, and for the final analysis, Eviews and State software tools were used. 4- ResultsThe results of this research showed that a research model with fundamental factor, based on accounting characteristics, has a better performance in explaining the stock returns compared to the corresponding multifactor pricing models; and the tests that considered the company’s expected return, performed better compared to those that did not. 5- Discussion and Conclusion Findings of this research indicate that adding a fundamental factor based on accounting characteristics to Fama and French three-factor, Carhart four-factor, Fama and French five-factor, and Fame and French six-factor models, improves their performance in explaining the stock returns. In other words, developing a model with a fundamental factor, based on accounting characteristics, can lead to an improvement in the asset’s pricing model. This is a result of all the efforts that have been made since the time of Sharpe (1964). In addition, the research findings show that despite the research and efforts that have been made in the field of assets’ pricing, it is still possible to further develop this model from other angles in the financial field.

Previous articleNext article FreeCommentGeorge W. EvansGeorge W. EvansUniversity of Oregon and University of St. Andrews Search for more articles by this author University of Oregon and University of St. AndrewsPDFPDF PLUSFull Text Add to favoritesDownload CitationTrack CitationsPermissionsReprints Share onFacebookTwitterLinked InRedditEmailQR Code SectionsMoreIntroductionExpectations clearly play a central role in modern macroeconomics. Households and firms are assumed to be dynamic optimizers, making decisions about work, consumption, savings, production, and investment, based in part on current economic conditions, but also to a great extent on the future state of the economy. Thus, in particular, household saving and portfolio decisions depend on expected future interest rates, inflation, and taxes and on the likely future trajectory of equity dividends and prices. Because of the key role of expectations in economics and finance, theories of expectations have been central to modern economic theory. Since the rational expectations (RE) revolution of the 1970s—associated with John Muth, Robert Lucas, and Thomas Sargent—the benchmark theory has been that expectations are formed rationally, in the sense that they are consistent with the true model and yield forecast errors that are orthogonal to agents’ information sets.In their paper in this volume, Andreas Fuster, Benjamin Hebert, and David Laibson present an asset-pricing model in which RE is replaced by natural expectations (NE). Under NE, agents misspecify the time-series model in a “natural” way: they chose a parsimonious model of dividends that omits longer lags. This captures short-run dynamics but misses long-run mean reversion. An earlier paper, Fuster, Laibson, and Mendel (2010), made similar arguments about other macroeconomic time series.1 Taken together, the two papers suggest the even bolder possibility of NE as a general stylized description of expectation formation.In the current paper, Fuster et al. insert an NE dividend forecasting equation into a Lucas-type consumption-based asset-pricing model with constant absolute risk aversion (CARA) preferences and habit persistence. Using this setup, Fuster et al. can replicate a number of stylized facts and puzzles about asset price data and consumption. These include the findings of excess volatility of stock prices, that excess returns are negatively predicted by lagged excess returns, price to earnings ratios, and consumption growth, and the existence of a large equity premium.Outline of Their ArgumentFuster et al. consider a variation of the Lucas-type “tree” model of asset prices. In keeping with Lucas (1978) there is an endowment economy with a single risky asset, trees, which provide an exogenous stochastic dividend of the perishable, homogeneous consumption good. The principal variation is that Fuster et al. consider an open economy version in which agents can borrow or lend internationally at a fixed interest rate R. In addition, exponential (CARA) preferences with habits are used. Finally, Δdt, the first difference in dividends, is assumed to follow to follow a stationary AR(p) process for some p > 0.The exponential preferences give a type of mean-variance setup, and Fuster et al. show that the asset price pt satisfies Fuster et al. also show how to obtain closed-form expressions for pt and for consumption, ct, given beliefs about the AR(p) process for Δdt. Here R is the inverse of the discount factor, α is the CARA measure of risk aversion, and γ is the habit-persistence parameter.The key assumptions of the Fuster et al. model concern the stochastic process actually followed by dividends and the perceived process followed by dividends. The true dividend process is assumed to be a high-order stationary AR(p) for the first difference in dividends dt; that is, , where ϵt is white noise and . Furthermore, Φ(L) is assumed to be such that there is a hump-shaped impulse response function for dividend levels . Put differently, dt is assumed to have a unit root with dynamics that lead to long-run mean reversion.Evidence for this is given in Fuster et al.’s figure 2, which presents based on empirical estimates of Φ(L) for AR(p) processes with alternative values of p. (In the empirical work, Fuster et al. use earnings rather than dividends.) The long-run level of persistence is given by and for p ≥ 15 we have . Thus there is mean reversion in the sense that an innovation ϵt has a reduced permanent impact. Fuster et al. assume that large values of p (e.g., p = 40) correspond to the truth.In contrast, for p ≤ 10 estimates of long-run persistence are . That is, based on low-order AR(p) estimates, one would come to the conclusion that one should extrapolate innovations in dividends—that the long-run effects are larger than the impact effect. Fuster et al. assume that low-order AR(p) estimates correspond to the perceived dividend process; that is, to the view held by economic agents.The essence of Fuster et al.’s approach is thus that the beliefs of agents differ from the truth and do so in a particular way. Agents use simpler low-order time-series models that lead them to accentuate the importance of short-run trends and to neglect longer-run corrections in which dividends revert toward an underlying trend. This central feature leads to the empirical implications noted before.What is the rationale for the discrepancy they assume between truth and perception? Fuster et al. give two types of argument—statistical / econometric and psychological. The statistical argument is that econometricians have often argued that in forecasting there is an advantage in using parsimonious models in preference to more complex models. Furthermore, standard statistical procedures for model selection based on Akaike Information Criterion (AIC) and, especially, Bayesian Information Criterion (BIC) often select low-order models. The psychological argument is that agents, for a variety of reasons, prefer to use simple, parsimonious models in preference to complex models when trying to understand the world and make decisions.Fuster et al. are usually careful not to be too dogmatic on this point. In essence they say: there is some evidence, as seen in their table 1 and figure 2, that a higher-order AR(p) process of Δdt might well be correct, while agents might plausibly believe in a low-order process. Their paper then explores the full implications of this assumption.Links to the Macro Learning LiteratureIf the truth is that p is large, but agents believe that p is small, then clearly agents do not have RE. There is a now extensive macro literature, which started around 1980, in which RE is replaced, for example, by adaptive or econometric learning (see Sargent 1993 and Evans and Honkapohja 2001, 2009).A major argument for the adaptive learning approach is what might be called the cognitive consistency principle.2 According to this principle, agents should be assumed to have the same level of rationality as the economic modeler or policymaker (in contrast to both old-style adaptive expectations and to RE). On the adaptive learning approach agents are assumed to make forecasts in the same way that econometricians do—formulating models, estimating their parameters, and updating estimated coefficients over time as new data become available. When parameters are updated using a form of least-squares, this is known as least-squares (LS) learning.The early macro learning literature focused on whether or not LS learning would converge over time to RE in self-referential models, in which the variables being forecasted are affected by the forecasts. Conditions were worked out that determined whether or not REE (RE equilibria) were indeed stable under LS learning. Stability under learning could then be used as a selection criteria in models with multiple REE, since in some cases only a subset of REE were stable under learning.More recently, another major strand has been to show how learning can generate transitory or persistent “learning dynamics”; that is, dynamics different from RE. Much of the recent macro learning literature has emphasized learning dynamics induced by one or more of the following factors: (a) misspecified forecasting models (misspecified “perceived laws of motion” or PLMs); (b) discounted LS learning (downweighting past data due to concern about unknown structural change); and (c) dynamic predictor selection (selecting between alternative PLMs based on past performance) or Bayesian model averaging. Applications of the approach that emphasize learning dynamics include: the rise and fall of inflation in the United States, hyperinflation, business cycles, output and inflation inertia, optimal monetary and fiscal policy, and asset price anomalies.One useful concept from the recent macro learning literature has been that of a restricted perceptions equilibrium (RPE), in which agents make the best forecast they can, given their misspecified PLM.3 A special case of interest has been models that are underparameterized, either in terms of variables or lag lengths. The argument here has been precisely that econometricians recognize the value of parsimoniously specified models, and thus the cognitive consistency principle dictates that we should examine the implications of underparameterization. One can, for example, work out stability conditions for an RPE when agents use LS learning to update coefficients of a particular underparameterized model.Thus the Fuster et al. approach fits well with the recent macro learning literature. The principal contribution of this paper, in this context, is that it posits a particular, plausible type of misspecification by agents, which can arguably explain several puzzling features of asset prices, and which may also be of more general applicability.DiscussionI certainly find plausible Fuster et al.’s key assumption that economic agents underparameterize their forecasting models. This assumption is consistent with the cognitive consistency principle, given that many applied econometricians place value on parsimony and recognize the likelihood of misspecification. This hypothesis also fits well with the observation that many economists believe there is long-run mean reversion that is nonetheless difficult to detect.4 That is, the misspecification that Fuster et al. assume is particularly plausible.Other aspects of the Fuster et al. model are also attractive: the closed-form solutions under CARA preferences, for the class of perceived dividend processes examined, is likely to be more generally useful, and the simultaneous fit of a range of stylized facts is impressive.It is therefore hard not to like this paper: the model is both simple and powerful. However, of course the model has some weaknesses, several of which are noticeable from the macro learning viewpoint. This in turn suggests a number of natural extensions.CriticismsIn my critical discussion I will focus on three main issues. The first concerns the information set available to agents when making forecasts. By assumption, Δdt is an exogenous univariate process, which leads Fuster et al. to examine alternative univariate forecasting models. However, within macroeconomics the norm is to consider multivariate forecasting models, and this issue is pertinent to the question of long-run persistence and to the plausibility of the form of underparameterization assumed. For example, in the early discussion of long-run GDP persistence, Campbell and Mankiw (1987) focused on univariate techniques, and found persistence levels greater than one. However, both the unemployment rate and the consumption-output ratio Granger cause output growth, and lower levels of persistence, with mean reversion, are found in multivariate models (e.g., see Evans 1989 and Evans and Reichlin 1994). In the current context Timmermann (1994), for example, has argued that stock prices Granger cause dividends. Thus a simple bivariate forecasting model might lead to different persistence results. The issue is whether simple—that is, low-order vector autoregressions—might show long-run mean reversion more clearly, in which case this feature of the data would be less plausibly missed by economic agents. Of course, many agents might still in practice use “natural” low-order univariate models, but a heterogeneous expectations model might then be more realistic.My second concern is the fixed parameter assumption of Fuster et al. Suppose first that we agree that agents plausibly underparameterize Δdt as an AR(1). From the learning viewpoint this leads to the corresponding RPE as the appropriate equilibrium to which the system would, if stable, converge. However, the cognitive consistency principle suggests that agents would not know the parameters of this process a priori, but, like real-world econometricians, would estimate the parameters and update their estimates over time. Furthermore, if agents are concerned about potential structural change, they might discount older data, leading to persistent learning dynamics around the RPE. This particular issue could easily be addressed by simulations in which fixed parameter natural expectations were replaced by discounted LS learning with the same AR(1) PLM.Related to both of the previous two points, if long-run estimates of persistence are crucial for good decision making in their portfolio choices, one might expect agents to focus on this issue in their choice of forecasting models. They might estimate mean reversion directly and allow for uncertainty concerning its value in their decisions. Alternatively, they might adopt decision-making rules that are robust to errors in this dimension, along the lines of Hansen and Sargent (2007).The third issue, which is probably most central from the learning perspective, is that the Fuster et al. model is not self-referential. Agents simply forecast dividends, which is treated as an exogenous process, and do not have a forecasting model for stock prices. Some learning models emphasize short-horizon decision making, in which the demand for stocks depends on short-horizon expected returns, and possibly also on the estimated conditional variance of returns. See, for example, Brock and Hommes (1998); Lansing (2010); Adam, Marcet, and Nicolini (2010); and Branch and Evans (2011). Indeed, one way to formulate the most basic risk-neutral model of stock prices is to assume that prices are determined by the sum of expected dividend and expected stock price in the coming period. These models are self-referential in the sense that asset prices today depend on the expected price tomorrow, so that the evolution of the variable being forecasted depends on the expectations themselves.Self-referential models, because of this feedback, give a much greater role to expectations, and this makes more likely asset-price bubbles: self-fulfilling or nearly self-fulfilling asset price movements with complex dynamics in which prices can become detached from fundamentals for extended periods. My own view is that this dynamic plays a central role in asset prices.Example: A Simple Model of BubblesAn example of the scope for dramatic learning dynamics in self-referential asset price models is given in my work with William Branch, presented in Branch and Evans (2011). We use a simple mean-variance linear asset pricing model. The setup can come from an overlapping generations model in which agents have two period planning horizons, CARA preferences, and a choice between a risky stock and a risk-free asset. The central equation is where zst is the iid random supply of the risky asset, denotes the subjective expectations of agents, and is their estimate of the conditional variance of returns. We assume the dividend process is known and that agents therefore need estimates of the price process to make their decisions.With iid dividend and supply shocks, the REE for pt is a constant + white noise. Under learning, agents forecast pt as an AR(1) using discounted LS and they estimate using a simple recursive algorithm. Because agents discount past data, prices under learning will occasionally break free from their fundamentals and exhibit bubbles and crashes. This results from the self-referential feature of the model.An illustrative simulation is shown in figure 1 (see Branch and Evans 2011 for analysis and for other simulations). The figure shows the realized price pt under learning and also the time series of estimates of two key learning parameters, the AR(1) coefficient ct and the estimate of the conditional variance . The figure shows the price process initially very close to the REE, which is a constant plus white noise in our setup. However, under learning, asset prices occasionally break free into a bubble regime in which stock prices are believed to follow a pure random walk (c = 1). In this regime pt is particularly sensitive to changes in the estimate of risk .Fig. 1. Asset price dynamics in the Branch and Evans (2011) modelView Large ImageDownload PowerPointIn summary, self-referential learning models have great scope for generating some of the more extreme partially self-fulfilling movements of stock prices often described as bubbles and crashes. Intuitively, the reason for this is that the stock price pt depends on expected price with a coefficient β < 1 that is close to one.Other Types of Learning DynamicsIn various settings learning dynamics have also been shown in self-referential models to lead to: (a) inertia of inflation and output, as in Orphanides and Williams (2007) and Milani (2007); (b) overshooting and nonmonotone impulse response functions (e.g., Eusepi and Preston 2011 and Evans, Honkapohja, and Mitra 2009); and (c) regime-switching and parameter drift (e.g., Sargent 1999, Cogley and Sargent 2005, and Branch and Evans 2007). For numerous examples and references, see Evans and Honkapohja (2011).With this in mind, consider again the question of whether it is plausible that agents believe Δdt is a specific constant coefficient AR(p) process with known parameters. Parameter drift and regime switching appear to be standard features of the data, as emphasized by Sims and Zha (2006), Cogley and Sargent (2005), and Sargent, Williams, and Zha (2006). The cognitive consistency principle suggests that agents should therefore allow for the possibility of structural change in their parameter estimation and through model selection, model averaging, or robust decision making.In the Fuster et al. setup, under RE the risk premium is very small. In the late 1990s some people argued (“Dow 30,000”) that the rise of the stock market was due to a recognition that the market risk premium was too high. An implication of Fuster et al. is that this view is fundamentally correct. Is it not, however, more plausible to believe that the risk premium reflects the uncertainty that economists and agents share?ConclusionsAlthough I have indicated a number of reservations, overall I find the Fuster et al. story very attractive. The setup is conceptually simple, and it is based on the plausible premise that agents underparameterize their forecasting model. This is in line with standard econometric advice to estimate parsimonious models, as well as evidence from psychology that people are inclined to make decisions based on simple heuristics. The Fuster et al. model is disciplined and delivers a number of important empirical implications that appear to be in line with the data.I would prefer to extend the model to include additional insights from the adaptive learning literature, but even as it stands the Fuster et al. model provides an impressive but simple benchmark model of asset-price behavior, which is sure to receive considerable attention.EndnotesFor acknowledgments, sources of research support, and disclosure of the authors’ material financial relationships, if any, please see http://www.nber.org/chapters/c12406.ack.1.Actually, in Fuster et al. (2010) the term “natural expectations” is used to denote an average between RE and what is called NE in the current paper.2.See Evans and Honkapohja (2009, 2011).3.Closely related concepts are those of “self-confirming equilibria” and “consistent expectations equilibria.”4.Estimating long-run persistence is equivalent to estimating the spectrum of dt at zero, and is understood to be subject to great uncertainty.ReferencesAdam, Klaus, Albert Marcet, and Juan Pablo Nicolini. 2010. “Learning and Stock Market Volatility.” Working Paper.First citation in articleGoogle ScholarBranch, William A., and George W. Evans. 2007. “Model Uncertainty and Endogenous Volatility.” Review of Economic Dynamics 10:207–37.First citation in articleCrossrefGoogle Scholar———. 2011. “Learning about Risk and Returns: A Simple Model of Bubbles and Crashes.” American Economic Journal: Macroeconomics 3:159–91.First citation in articleCrossrefGoogle ScholarBrock, William A., and Cars H. Hommes. 1998. “Heterogenous Beliefs and Routes to Chaos in a Simple Asset Pricing Model.” Journal of Economic Dynamics and Control 22:1235–74.First citation in articleCrossrefGoogle ScholarCampbell, John Y., and N. Gregory Mankiw. 1987. Quarterly Journal of Economics 102: 857–80.First citation in articleGoogle ScholarCogley, Timothy, and Thomas J. Sargent. 2005. “The Conquest of US Inflation: Learning and Robustness to Model Uncertainty.” Review of Economic Dynamics 8:528–63.First citation in articleCrossrefGoogle ScholarEusepi, Stefano, and Bruce Preston. 2011. “Expectations, Learning and Business Cycle Fluctuations.” American Economic Review 101:2844–72.First citation in articleCrossrefGoogle ScholarEvans, George W. 1989. “Output and Unemployment Dynamics in the United States: 1950–1985.” Journal of Applied Econometrics 4:213–37.First citation in articleCrossrefGoogle ScholarEvans, George W., and Seppo Honkapohja. 2001. Learning and Expectations in Macroeconomics. Princeton, NJ: Princeton University Press.First citation in articleGoogle Scholar———. 2009. “Learning and Macroeconomics.” Annual Review of Economics 1:421–51.First citation in articleCrossrefGoogle Scholar———. 2011. “Learning As a Rational Foundation for Macroeconomics and Finance.” In Rethinking Expectations: The Way Forward for Macroeconomics, Roman Frydman and Edmund S. Phelps (eds.), Princeton, NJ: Princeton University Press, forthcoming.First citation in articleGoogle ScholarEvans, George W., Seppo Honkapohja, and Kaushik Mitra. 2009. “Anticipated Fiscal Policy and Learning.” Journal of Monetary Economics 56:930–53.First citation in articleCrossrefGoogle ScholarEvans, George W., and Lucrezia Reichlin. 1994. and of the Business Journal of Monetary Economics Fuster, A., Laibson, and 2010. Expectations and Fluctuations.” Journal of Economic citation in articleCrossrefGoogle and Thomas J. Sargent. 2007. Princeton, NJ: Princeton University Press.First citation in articleGoogle 2010. and Bubbles Economic Journal citation in articleCrossrefGoogle Robert in an citation in articleCrossrefGoogle 2007. “Expectations, Learning and Journal of Monetary Economics citation in articleCrossrefGoogle and John 2007. Monetary Policy with Journal of Monetary Economics Sargent, Thomas J. in Macroeconomics. University Press.First citation in articleCrossrefGoogle Scholar———. The Conquest of American Princeton, NJ: Princeton University Press.First citation in articleGoogle Thomas Williams, and and The and of American American Economic Review citation in articleCrossrefGoogle and There in US Monetary American Economic Review citation in articleCrossrefGoogle 1994. Agents to Rational Some on and Stability of Learning in the Stock Economic Journal citation in articleCrossrefGoogle Previous articleNext article by Macroeconomics Annual by the of Economic on this by The of Economic articles this

In recent years, scholars have explored hundreds of asset pricing factors built upon the foundation of the Fama five-factor model, sparking widespread discussion in the academic community. Simultaneously, the advancement of machine learning techniques has brought innovation to asset pricing factors. This article provides an overview of typical asset pricing factors and the application of machine learning in pricing models. It begins by discussing the construction of new asset pricing factors and then delves into the innovations brought about by machine learning in asset pricing models. The diversity of factors increases the fit of asset pricing models but also presents corresponding challenges. The application of machine learning techniques addresses issues such as overfitting in pricing models, further enhancing model effectiveness. The primary value of this article lies in summarizing various perspectives on asset pricing factors in recent years, exploring the significant applications of machine learning in asset pricing models, and providing a forward-looking view on the development of asset pricing models.

This chapter starts with explaining the term “Asset Pricing”. It covers discussion on the different school of thoughts of asset pricing studies. The capital asset pricing model (CAPM) is discussed in the line of its goal, assumptions, validity, and significance. Thereafter detailed discussion was made on the different asset pricing models that evolved over a period of time namely from ICAPM to the various multifactor asset pricing models. Detailed discussion is also made on the (Fama and French, Journal of Financial Economics 33:3–56, Fama and French, 1993) portfolio construction methodology. Econometrics of the linear factor pricing models are covered in detail. Testing one model versus the other model is vital in asset pricing studies. Critical discussions are made on it taking suitable examples. This unit concluded with the actual implementation of the different asset pricing models using EViews. Furthermore, the panel regression is also covered in detail with suitable illustrations.

My three essays contain three studies using multi-factor asset pricing models, where all the data are based on the US market. The first study extends intertemporal CAPMs with a few macro pricing factors: inflation or the cycle of industrial production (IP). I regard this specification of such models as a multi-factor pricing model, where this multi-factor linear pricing model can alternatively be derived from a consumption-based model from a theoretical perspective. I find significant evidence that the augmented multi-factor models outperform the original ICAPM. The results show that inflation is a key additional factor in the pricing models for the 25 size/book-to-market portfolios, while the cycle of IP is another vital additional factor in pricing models for the 25 size/momentum portfolios. Moreover, I find that most pricing information contained in the momentum factor is the inclusive information of the IP cycle, where the cycle of IP is generated by using the Hodrick–Prescott filter. The second study extends another two ICAPMs and Hou, Karolyi and Kho's three-factor model with inflation. The evidence shows that inflation significantly aids the original models in pricing 25 size/book-to-market portfolios in cross-sectional tests. Hence, I provide further robust evidence that inflation is the vital factor in the factor pricing models for the 25 size/book-to-market portfolios and a few other portfolios. Inflation provides additional explanatory power beyond Fama-French’s five factors in pricing the cross-sectional variation of 25 size/book-to-market portfolios. The third study investigates the performance of multi-factor asset pricing models in explaining the cross-section variation of the large number of expanding portfolios and a set of different portfolios, where the multi-factor models refer to the Fama-French three-factor model augmented by other pricing factors. I investigate the performance of several well-regarded multi-factor models by using Hansen’s general method of momentum (GMM), which is another alternative and very robust complement/guarantee to the only regression-based procedure in the previous literature. The results continuously support the superiority of the augmented multi-factor models. In general, augmented multi-factor models outperform the original models in a sound portion of different portfolios, where the original model refers to Fama-French’s three-factor model. In conclusion, my essays shed light on a fresh type of linear asset pricing model with sound theoretical background, and my research justifies the superiority of the multi-factor pricing model over Fama-French’s three-factor model in explaining the cross-sectional variation of equity returns with robust evidence.

Behavioral finance is a new issue raised by some financial intellectuals over the past two decades and has been quickly addressed by professors, experts, and students throughout the world. Investigating the factors affecting investment decisions is carried out in the field of behavioral finance; in other words, the focus of behavioral finance is on the specific charac-teristics of human behavior and applying them in asset pricing. Empirically, pricing models rarely include psychological factors, but the noticeable point is that nowadays, researchers have found behavioral factors influencing empirical asset pricing models that can manipulate returns on asset mispricing. Behavioral asset pricing is the result of applying behavioral finance theories within traditional asset pricing theories. Thus, despite the existence of many asset pricing models, due to their weaknesses and lack of comprehensiveness, as well as the necessity of reviewing behavioral factors, this study aims to model asset pricing through behavioral models. Using the data from 141 listed firms in Tehran Stock Exchange over the years 2008 to 2017 and multivariate regression, this study is an attempts to model asset pric-ing through employing behavioral models and Fama-French approach. Using Fama-French approach, the results showed that accounting information risk, investors’ trading behavior, and investors' sentiment have a direct and significant impact on asset pricing.

Purpose of the study: The study shows how the financial factor of Leverage affects the empirical model of asset pricing together with other financial factors, i.e. Size, Book to Market, Operating Profit, and Investment. The contribution of Leverage in asset pricing will be tested, and its effect will be shown in the excess return of the asset.&#x0D; Methodology: The methodology used in this paper is based on the Fama and French model of asset pricing with additional factors added in the model. Data processing follows the Fama-Mc Beth procedure. Data comes from the Indonesian Stock Market, which consists of more than 500 stocks for ten years period of observation.&#x0D; Main Findings: The financial factor of Leverage affects the empirical model of asset pricing together with, i.e. Size, Book to Market, Operating Profit, and Investment. All the financial factors in the model are stationary around their mean, or they are non-stationary due to unit-roots. All the independents' variables have P-Value less than 10%.&#x0D; Implications: This study will be useful for financial investors in building an effective portfolio stock investment. By applying this model to their portfolio investment, the investors could effectively manage their portfolio return. On the management side, managing their financing structure, e.g. Leverage is the objective of the firm to maximize returns of the firms.&#x0D; Novelty/Originality of this study: The empirical research with the involvement of the financial factor of Leverage has not been performed in Indonesia. The Leverage as the single factor of asset pricing has been considered as a significant financial factor for asset pricing, however, how the Leverage contributes to asset pricing compares to other financial factors has not examined yet.

Problem/Relevance: Measuring the risk of an asset and the economic forces driving the price of the risk is a challengingtask that preoccupied the asset pricing literature for decades. However, there exists no consensus on the integrated asset pricing framework among the financial economists in the contemporaneous asset pricing literature. Thus, we consider and study this research problem that has greater relevance in pricing the risks of an asset. In this backdrop, we develop an integrated equilibrium asset pricing model in an intertemporal (ICAPM) framework. Research Objective/Questions: Broadly we have two research objectives. First, we examine the joint dynamics of the human capital component and common factors in approximating the variation in asset return predictability. Second, we test whether the human capital component is the unaccounted and the sixth pricing factor of FF five-factor asset pricing model. Additionally, we assess the economic and statistical significance of the equilibrium six-factor asset pricing model. Methodology: The human capital component, market portfolio, size, value, profitability, and investment are the pricing factors of the equilibrium six-factor asset pricing model. We use Fama-French (FF) portfolios of 2  3, 5  5, 10  10 sorts, 2  4  4 sorts, and the Industry portfolios to examine the equilibrium six-factor asset pricing model. The Generalized method of moments (GMM) estimation is used to estimate the parameters of variant asset pricing models and Gibbons-Ross-Shanken test is employed to evaluate the performance of the variant asset pricing frameworks. Major Findings: Our approaches led to three conclusions. First, the GMM estimation result infers that the human capital component of the six-factor asset pricing model significantly priced the variation in excess return on FF portfolios of variant sorts and the Industry portfolios. Further, the sensitivity to human capital component priced separately in the presence of the market portfolios and the common factors. Second, the six-factor asset pricing model outperforms the CAPM, FF three-factor model, and FF five-factor model, which indicates that the human capital component is a significant pricing factor in asset return predictability. Third, we argue that the human capital component is the unaccounted asset pricing factor and equally the sixth-factor of the FF five-factor asset pricing model. The additional robustness test result confirms that the parameter estimation of the six-factor asset pricing model is robust to the alternative definitions of the human capital component. Implications: The empirical results and findings equally pose the more significant effects for the decision-making process of the rational investor, institutional managers, portfolio managers, and fund managers in formulating the better investment strategies, which can help in diversifying the aggregate risks.

Chapter 6 - Fast-and-frugal asset pricing

Since Sharpe (1964) derived the CAPM, it has been the benchmark of asset pricing models and has been used to calculate the cost of equity capital and other asset pricing determinations for more than four decades. Many researchers have tried to relax the original assumptions and generalize the static CAPM. In addition, Merton (1973) and Black (1976) have generalized the static CAPM in terms of intertemporal CAPM. In this chapter, we survey the important alternative theoretical models of capital asset pricing and provide a complete review of the evolution of both static and intertemporal asset pricing models. We also discuss the interrelationships among these models and suggest several possible directions for future research. In addition, we review the asset pricing tests in terms of individual companies’ data instead of portfolio data. Our results might be used as a guideline for future theoretical and empirical research in capital asset pricing.

This dissertation is composed of three chapters addressing the connections between investor beliefs and asset pricing. Specifically, I focus on one prevailing pattern of investor beliefs in the finance literature, return extrapolation. The idea is that investor expectations about future market returns are a positive function of the recent past returns. In this dissertation, I use this concept to understand a number of facts in the asset pricing literature. Return extrapolation attracts growing attention in the literature, not only because it well explains real-world investors' expectations in the survey, but also because it significantly drives investor demand towards stocks. Therefore, we should anticipate a connection between return extrapolation measurement and the stock market dynamics. However, contrary to the intuition, previous empirical studies fail to document a significant connection. In Chapter 1, Time-varying Impact of Investor Sentiment, I recover this connection. Specifically, I formally define investors who extrapolate past returns as extrapolators and incorporate their wealth level into analysis. My main finding is that return extrapolation interacts strongly with extrapolators' wealth level in predicting future market returns. Therefore, conditional on extrapolators' wealth level, return extrapolation significantly explains stock market returns. The return extrapolation concept also raises challenges to the asset pricing models under the rational expectation frameworks. Specifically, rational expectation theories lead to a positive correlation between expectations and future realized returns, whereas return extrapolation indicates a negative correlation. Given this discrepancy, there is a clear demand for a behavioral asset pricing model that can simultaneously explain survey evidence on investor expectations and the classical asset pricing puzzles. In Chapter 2, Asset Pricing with Return Extrapolation, coauthored with Lawrence Jin, we present a new model of asset prices based on return extrapolation. The model is a Lucas-type general equilibrium framework, in which the agent has Epstein-Zin preferences and extrapolative beliefs. Unlike earlier return extrapolation models, our model allows for a quantitative comparison with the data on asset prices. When the agent's beliefs are calibrated to match survey expectations of investors, the model generates excess volatility and predictability of stock returns, a high equity premium, a low and stable risk-free rate, and a low correlation between stock returns and consumption growth. In Chapter 3, Dark Matter of Finance in the Survey, I investigate another attribute of investor beliefs—tail risk perceptions. Although tail risks play significant roles in explaining asset pricing puzzles, researchers have very limited knowledge about them because tail events are difficult to observe. I use Shiller tail risk survey to empirically investigate tail risk perceptions. In this survey, investors are asked to report their estimated probability for a crash event in the U.S. stock market. However, when using survey data to understand investors’ perception of tail risks, there are two fundamental challenges. First, is tail risks survey reliable? Second, to avoid cherry-picking, is there a unified framework to explain different attributes of investor beliefs? My analysis provides positive answers to both questions. First, I show that Shiller tail risk survey is reliable. More importantly, I show that return extrapolation can serve as a unified belief formation framework to explain not only variations in investor expectations but also in tail risk perceptions.

Latent variable models in finance originate both from asset pricing theory and time series analysis. These two strands of literature appeal to two different concepts of latent structures, which are both useful to reduce the dimension of a statistical model specified for a multivariate time series of asset prices. In the CAPM or APT beta pricing models, the dimension reduction is cross-sectional in nature, while in time-series state-space models, dimension is reduced longitudinally by assuming conditional independence between consecutive returns, given a small number of state variables. In this paper, we use the concept of Stochastic Discount Factor (SDF) or pricing kernel as a unifying principle to integrate these two concepts of latent variables. Beta pricing relations amount to characterize the factors as a basis of a vectorial space for the SDF. The coefficients of the SDF with respect to the factors are specified as deterministic functions of some state variables which summarize their dynamics. In beta pricing models, it is often said that only the factorial risk is compensated since the remaining idiosyncratic risk is diversifiable. Implicitly, this argument can be interpreted as a conditional cross-sectional factor structure, that is, a conditional independence between contemporaneous returns of a large number of assets, given a small number of factors, like in standard Factor Analysis. We provide this unifying analysis in the context of conditional equilibrium beta pricing as well as asset pricing with stochastic volatility, stochastic interest rates and other state variables. We address the general issue of econometric specifications of dynamic asset pricing models, which cover the modern literature on conditionally heteroskedastic factor models as well as equilibrium-based asset pricing models with an intertemporal specification of preferences and market fundamentals. We interpret various instantaneous causality relationships between state variables and market fundamentals as leverage effects and discuss their central role relative to the validity of standard CAPM-like stock pricing and preference-free option pricing.

This paper examines the effect of pandemic vaccination on asset prices in a simple asset pricing model à la Lucas 1978. In this model, asset prices depend on susceptible individuals’ saving motives to insure against a reduction in labour income due to getting they get the virus. Hence distributing vaccine reduces precautionary saving motives and asset prices. This implies that reducing the income gap between susceptible and infected individuals, such as by cash handouts, eases the negative effect of vaccine supply on asset prices.

Corporate managers often justify certain fairly common corporate practices by describing them as attempts to enlarge their company's investor base. These practices include stock splits, the hiring of shareholder relations officers, meetings with security analysts, the issuance of ADRs, and the listing of their company's shares on major domestic and international stock exchanges. Consider, for example, the news report announcing the decision by Coastal Healthcare to have its shares listed on the New York Stock Exchange: Coastal Healthcare said the listing will "enhance the liquidity of our shares, allow us to diversify our shareholder base, broaden our recognition with the investment community and further enhance shareholder value." (Wall Street Journal, 11/29/94); or this report that accompanied Microsoft's 2-for-1 stock split: Jon Shirley, Microsoft's president and chief operating officer, said the split "reflects the company's desire to make our stock more accessible to a broader base of investors." (Wall Street Journal, 8/4/87); or this report that accompanied the initiation of Sandoz Ltd.'s ADR program: Sandoz Ltd. of Basle, Switzerland, established a sponsored American depositary receipt program to broaden its international shareholder base. (Dow Jones News Wire, 11/07/91). The notion that investor base has an effect on share value has intuitive appeal and is strongly supported by "streetlore." But standard finance theory, as represented by the familiar Capital Asset Pricing Model (CAPM) or its recent challenger, the Arbitrage Pric-This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.