Andreas Andrikopoulos (2013)
It is important to know the true value of a financial asset. Profits and losses depend on it, as financial theory suggests. We need to know, for instance, whether an IPO reflects real business prospects and whether the price of a stock reflects its true value. The question of the true value of a financial security has important implications for security design, market microstructure, and even the fragility of the financial system: corporate financing decisions are assessed with respect to the message that prices convey about the value of a stock, reforms of market microstructure aim at a fast and efficient “price discovery process” in which market prices ultimately reflect true values, and, characteristically, stock market crashes are usually assumed to happen when bubbles burst and financial securities rapidly fall from the heights of false values toward the modest vicinity of true ones.
The value of a financial security is archetypically defined as the present value of future proceeds from ownership of this security; this is the true value with respect to which market outcomes and investors’ sentiments may diverge. As truth and value are jointly evaluated by financial theorists and market participants alike, we need to assess the possibility of a mind-independent reality in the valuation of financial assets. If the truth of propositions like “The Wall Street is located in Manhattan” or “Bernard Madoff is in prison” can be assessed with respect to the actual facts of the location of Wall Street or the residence of Bernard Madoff, can the same principle apply to the proposition “the value of the stock is five dollars”? I will proceed to argue that it cannot. Unlike the location of, say, Wall Street, the value of a stock does not exist independently of our thoughts and specifically our valuation models. The truth of the proposition “the value of the stock is five dollars” depends, among a plethora of other things, on our adoption of a particular modeling approach (such as, e.g., Gordon’s growth model) and also on our sentiments about corporate prospects and market-wide economic activity. Therefore, if truth is accepted as a mind-independent property of propositions, then truth cannot be a property of propositions about the value of a financial security.
The epistemic status of propositions about the value of financial securities can be discussed in the context of some fundamental theories of truth. The correspondence theory of truth is often at the beginning of truth-theoretic discourses. According to the correspondence theory of truth, a proposition is true if a) it corresponds to the facts, and b) the proposition accurately conveys the intended meaning, given our linguistic conventions. The proposition “the key is on the door” is true if the fact is that the key is actually on the door and we have adopted the conventions of (some official version of) the English language; therefore, the proposition “la clé est sur la porte” is not true since it may correspond to the facts but it is linguistically devoid of meaning, in the context of the English language. Can the correspondence theory of truth apply in financial valuation? Whereas the position of the key with respect to the door is largely independent of what I think about it (setting aside my ability to take the key off the door), the value of a stock is not independent of what I think. My assumptions of market participants’ risk preferences will affect the discount rate of prospective dividends, my assumptions of the consumers’ preferences will affect the forecasts of future cash flows and, my adherence to, say, the Capital Asset Pricing Model, instead of the Arbitrage Pricing Theory will, in effect, shape the conclusion of my valuation analysis. Therefore, the concept of truth, as a mind-independent correspondence to facts, is irrelevant to the value of a financial security.
This context-dependent character of security valuation is not incompatible with all truth-theoretic traditions. The coherence theory of truth defines the truth of a proposition as its coherence with a set of other propositions. Clearly, coherence is a necessary property of successful theorizing in general, and asset valuation in particular. However, there are problems with such a truth-theoretic assessment of asset valuation propositions. The proposition “the value of the stock is five dollars” may be consistent with a given set of coherent propositions and incoherent with respect to another (e.g. the first set could include the proposition “the investors’ horizon is six years” whereas the second set could include the proposition “the investors’ horizon is six days”). Moreover, there can be a lot of coherent systems of propositions. For instance, the proposition “the value of the stock is three dollars” may be consistent with the proposition “the value of the stock equals the present value of expected future dividends”; the proposition “the value of the stock is five dollars”, however, may be consistent with the proposition “the value of the stock equals the average analyst’s valuation of that stock”; this leaves us with the problem that while these two valuation propositions are inconsistent with each other (the true value cannot be three and five dollars at the same time), they each satisfy the condition of coherence with a (different) set of propositions. Therefore, the discovery of true value depends on our choice of a coherent set of propositions; since it is a matter choice, the true value of a financial security does not exist prior to or independent of the process and principles of valuation.
The pragmatist approach to truth is perhaps most relevant to the ways in which the concept of truth is employed by financial economists. In this truth-theoretic approach, the truth of a proposition is assessed in the context of the consequences which would follow from the application of the proposition in practice. If action, based on the proposition, yields unfavorable results, then the proposition cannot be true because it led us to act in ways that are not in accordance to the true state of the world (and this accounts for the unfavorable outcome of our action). Despite the importance of truth for the properties of experience, Charles Sanders Peirce stressed the distinction between the truth of a proposition and our experience from its application. He also suggested that truth will be discovered with extensive, rational inquiry; such inquiry is the process of scientific research and, as scientific arguments are sound and impartial, the end of inquiry will yield a consensus on the truth of a proposition. In terms of finance, this means that extensive, rational research on corporate fundamentals and stock market fluctuations will yield the true value, and, given the efficiency of such research, the outcome of such inquiry will yield a consensus on the value of a financial security (if markets are efficient, no such research is needed; the market always discovers and reflects the true value). However, extensive research on corporate fundamentals and market dynamics has not produced a consensus between analysts (quite the opposite is the case); a plethora of scientific methods has yielded a plethora of outcomes on the true value of a financial security (e.g. multifactor models of valuation produce different results than single-factor ones). Instead, the use of term “true value” in the theory and practice of finance always refers to a fragile consensus among market participants. When we hear that the current market price is different from the true value, what is often implied is that market price will converge to the true value sooner or later and hence the observed deviation from the true value generates a profitable opportunity; and everyone acknowledges that such convergence (often termed “equilibrium”) is always fragile. Besides, even if we were to assess truth by focusing on the quality of experience -as the result of acting upon acceptance of a proposition as true- the personal, tacit, context-dependent character of experience would lead us to multiple and, most importantly, mutually exclusive accounts of the true value of a financial security.
We have seen that accounts of true values in financial economics are always context-specific and model dependent. Is there anything wrong with that? Well, if social science aims to discover and explain the truth about the structure of human conduct and the operation of society, then it must be the case that the structure of human conduct and the operation of the society preexist the utterance of the scientific question and the process of scientific discovery; that the object of inquiry preexists the inquiry itself is a condition for the possibility of discovery. If the object of inquiry is not external to the inquirer, we are left with no benchmark with respect to which we are to assess the performance of scientific explanation. While the process of inquiry is of an undeniably context-specific and theory-laden character, the progress of inquiry requires that the outcomes of inquiry are assessed with respect to the reality that they set out to explain; such assessment requires the externality of the reality (that is to be explained) with the scientific propositions (that produce the explanations). Financial economics is (part of a) social science. As such, it must try to explain the realities of market dynamics and investor behavior which are external to the explanatory models of financial economics. The structural association between the fear of the unexpected and choice of investment outlets was effective long before the invention of modern portfolio theory; the causal effect of tax laws on investment and financing decisions preexisted the emergence of the Modigliani-Miller theorem; the causal effect of a sovereign debtor’s creditworthiness on the interest rate of government bonds existed long before the formulation of models of credit risk. However, the “fact” that the value of the stock is five dollars does not preexist our adoption of a particular asset pricing model; the truth of an asset-valuation proposition is never assessed with respect to a reality that is external to the asset valuation process; all we can see is whether a numerical estimate of stock value is consistent with the parameter values and the modeling assumptions of our choice. The discovery of truth with respect to the value of financial security is not a viable goal of scientific inquiry in finance. Truth in finance is about the structure of investors’ behavior and the market’s operation (it is to this end that asset pricing models, and their econometric specifications, have been so useful).
Since asset valuation is always model-dependent, truth (with respect to valuation models) has to be sought at the level of social and behavioral structures that determine the popularity and academic status of competing valuation frameworks. What preexists any valuation model are the determinants of the persuasiveness of scientific arguments.
A pdf version of this can be found here.
About the Contributor
Andreas Andrikopoulos is an Assistant Professor of Finance at the Department of Business Administration of the University of the Aegean. He holds a PhD in Financial Modelling from Athens University of Economics and Business. He teaches courses on microeconomics, financial management and econometrics. His research focuses on corporate social responsibility of financial institutions, the liquidity of stock markets, and methodologies of financial economics.