Asset Pricing Theory

Asset pricing theory asks a single question in many forms: what should a risky asset be worth today?

10 frameworks, from 1930 to today · Financial Economics · Economics

still a working view ended · each leads to its entry below

Overview

Asset pricing theory asks a single question in many forms: what should a risky asset be worth today? A share of stock, a bond, a house, a derivative contract—each promises uncertain payments in the future, and each trades at a price set by people who must decide whether holding it is worth the risk. The field's answer, built over roughly seventy years, is that price equals the expected future payoff discounted for time and for risk, and that the hard part is measuring risk12. What is at stake is enormous: the cost of capital for every firm, the design of every pension, the regulation of every market, and the central bank's reading of whether asset prices reflect fundamentals or fever. The history of asset pricing is a sequence of increasingly precise attempts to say what "risk" means and how markets price it.

Present Value and the Classical Foundations

Long before academic finance existed, practitioners valued assets by discounting3. The idea that a dollar tomorrow is worth less than a dollar today—because it could be invested—is ancient, and by the nineteenth century actuaries and engineers used present-value formulas routinely. Irving Fisher's The Theory of Interest (1930) gave the idea its canonical economic form: an asset's price is the present value of its expected future cash flows, discounted at a rate reflecting time preference and risk3. This "present-value" framework was the common sense of finance for decades, but it left the discount rate unexplained. Why should one asset's cash flows be discounted more heavily than another's? The answer required a theory of risk, and that theory arrived in the 1950s.

Mean-Variance Portfolio Theory: Risk Becomes Measurable

Harry Markowitz, a graduate student at the University of Chicago, published "Portfolio Selection" in 1952. His insight was that an investor does not choose among individual securities but among portfolios, and that a portfolio's risk is not the average of its parts' risks4. If two stocks move imperfectly together, combining them reduces the variability of the whole5. Markowitz characterized each asset by its expected return and its variance (the statistical measure of how much its returns fluctuate), and by the covariance between assets (how much they move together)54. The investor's problem became a trade-off: for a given level of expected return, minimize variance; for a given variance, maximize return4. The set of such optimal portfolios traces an "efficient frontier."

The practical case was vivid. A portfolio split between two assets that move in opposite directions can have lower variance than either alone—diversification is not just a proverb but a mathematical result5. Markowitz's framework made risk a quantity one could compute, and it won him a share of the 1990 Nobel Memorial Prize in Economic Sciences4. Its limitation was that it described the investor's problem without saying what prices the market would set. That step came next.

The Capital Asset Pricing Model: One Risk That Pays

In the early 1960s, William Sharpe, John Lintner, and Jan Mossin independently extended Markowitz's logic to the whole market. If every investor holds an efficient portfolio, and if a riskless asset (a government bond with certain return) is available, then all investors hold the same risky portfolio—the market portfolio, containing every asset in proportion to its value6. From this, Sharpe (1964) derived the Capital Asset Pricing Model: an asset's expected excess return over the riskless rate is proportional to its "beta," the sensitivity of its returns to the market's returns. Beta measures non-diversifiable risk—the part of an asset's fluctuation that cannot be eliminated by holding many assets78. Risk that can be diversified away earns no premium, because investors can avoid it for free.

The CAPM's claim was sharp and testable: expected return rises linearly with beta, and nothing else about an asset should matter. Its appeal was that it reduced the pricing of every asset to one number. It became the workhorse of corporate finance (for estimating the cost of equity) and of performance evaluation (for judging whether a fund manager beat the market or merely took market risk). Early empirical tests in the 1960s and 1970s, notably by Sharpe and by Fischer Black, Michael Jensen, and Myron Scholes (1972), seemed broadly supportive. But the model rested on strong assumptions—that investors care only about mean and variance, that they share the same beliefs, and that borrowing and lending at the riskless rate is unrestricted. Cracks appeared as data accumulated.

Option Pricing and the Arbitrage Revolution

A parallel problem—pricing a contract whose payoff depends on a stock—produced the field's most striking result. Fischer Black and Myron Scholes (1973), with Robert Merton's contemporaneous and more general treatment, showed that a European call option's price could be derived from the stock price, the strike price, time to maturity, volatility, and the riskless rate—without knowing investors' risk preferences at all. The method was replication: a portfolio combining the stock and a riskless bond can be constructed to mimic the option's payoff exactly, so the option must sell for the cost of that replicating portfolio. If it did not, an arbitrageur could buy the cheap one and sell the dear one for a riskless profit9.

This "no-arbitrage" logic was a break from the equilibrium tradition of Markowitz and Sharpe. It did not ask what risk premium investors demand; it asked only that prices admit no free lunch. The Black-Scholes formula, published just as organized options trading began at the Chicago Board Options Exchange in 1973, was adopted almost immediately by traders. Merton and Scholes received the 1997 Nobel Prize (Black had died in 1995)10. The formula's assumption of constant volatility and continuous trading would later be strained by market crashes, but its replication argument became the template for pricing derivatives of every kind1011.

Arbitrage Pricing Theory and the Factor Turn

Stephen Ross's Arbitrage Pricing Theory (1976) offered a middle path12. Instead of assuming a particular market portfolio, Ross assumed that returns are generated by a small number of common factors—macroeconomic influences such as industrial production or interest rates—plus asset-specific noise. In a market with many assets, arbitrage should force expected returns to line up with exposures to those factors12. The theory was less restrictive than the CAPM but also less specific: it did not say what the factors were. That ambiguity invited empirical work, and the empirical work would eventually challenge the CAPM directly.

Anomalies and the Empirical Factor Models

Through the 1970s and 1980s, tests accumulated showing that beta alone did not explain returns. Basu (1977) found that stocks with low price-to-earnings ratios earned higher returns than the CAPM predicted. Banz (1981) found that small firms outperformed large ones13. Most consequentially, Eugene Fama and Kenneth French (1992, 1993) showed that two additional characteristics—size and book-to-market ratio (a measure of value versus growth)—explained the cross-section of average returns better than beta did, and that beta's explanatory power largely vanished once these were included. Their three-factor model became the standard empirical benchmark.

The pattern repeated. Jegadeesh and Titman (1993) documented momentum: stocks that rose over the past year tended to keep rising over the next few months. Later work added profitability and investment factors (Fama and French, 2015)13. Each new factor was, in effect, an admission that the simple CAPM was incomplete, and a challenge to explain why the premium existed—compensation for risk, or a mispricing that arbitrage had failed to correct. The debate between "risk-based" and "behavioral" explanations of these anomalies became one of the field's central fault lines.

Consumption-Based Pricing and the Stochastic Discount Factor

A deeper theoretical current ran alongside. Rather than starting from portfolio choice, the consumption-based approach starts from the investor's desire to smooth consumption over time14. An asset's value depends on how its payoff correlates with consumption: assets that pay off well when consumption is already high (and marginal utility is low) are worth less14. This idea, developed in the late 1970s by Rubinstein, Lucas, and Breeden, yields the "stochastic discount factor"—a random variable that varies across states of the world and prices every asset by the expectation of payoff times discount factor. The framework is general enough to nest the CAPM, the consumption model, and factor models as special cases, and it became the field's unifying language.

Its empirical problem was severe. The simplest consumption model, with a representative investor and standard preferences, predicted an equity premium far smaller than the roughly 6-percentage-point historical excess return of stocks over bonds—the "equity premium puzzle" identified by Mehra and Prescott (1985). Explaining that gap—through habit formation, long-run risk, rare disasters, or heterogeneous investors—occupied theorists for decades.

Behavioral Asset Pricing

The behavioral strand argued that some price patterns reflect investor psychology rather than rational risk compensation. De Long, Shleifer, Summers, and Waldmann (1990) showed that noise traders—investors trading on sentiment rather than fundamentals—could move prices away from fundamentals and earn higher expected returns, while rational arbitrageurs faced limits to correcting them. Barberis, Shleifer, and Vishny (1998) built models in which investors overweight recent evidence or underreact to news, generating momentum and value effects. The behavioral program did not replace the rational framework; it supplied candidate explanations for the anomalies the factor literature had catalogued, and it sharpened the question of which premiums are risk and which are error.

Machine Learning and the Modern Cross-Section

By the 2010s, the number of documented return predictors had grown into the hundreds, raising the question of which were real and which were statistical artifacts of data mining. Machine-learning methods—regularization, cross-validation, and tree-based ensembles—entered empirical asset pricing as tools for handling many predictors at once. Gu, Kelly, and Xiu (2020) showed that flexible machine-learning models could forecast returns better than linear factor models, and that the gains came largely from nonlinear interactions among characteristics. The approach is young and contested: its advocates see a more honest accounting of the cross-section, while critics warn that flexible models can overfit history and that interpretability matters for testing economic theory.

Where the Field Stands

Today asset pricing is a family of approaches rather than a single model. The stochastic discount factor provides the common grammar; the CAPM survives as a benchmark and a teaching tool; multi-factor models dominate empirical work; no-arbitrage methods underpin derivatives pricing; and behavioral and machine-learning research push at the edges. The open questions are the old ones in new dress: why the equity premium is so large, whether the many documented factors are compensation for risk or symptoms of mispricing, how frictions and heterogeneous beliefs shape prices, and whether the next generation of data-driven models will reveal structure or merely fit noise. The field's history is a record of making the question of risk precise enough to test—and of discovering, each time, that the answer was more complicated than the last model allowed.

References

  1. Asset pricing (Interrelationship), Wikipedia
  2. Stochastic discount factor, Wikipedia
  3. Discounted cash flow (History), Wikipedia
  4. Modern portfolio theory, Wikipedia
  5. Markowitz model, Wikipedia
  6. Capital market line, Wikipedia
  7. Capital asset pricing model, Wikipedia
  8. Beta (finance), Wikipedia
  9. Risk-neutral measure (Origin of the risk-neutral measure), Wikipedia
  10. Black–Scholes model (History, introduction), Wikipedia
  11. Volatility smile, Wikipedia
  12. Arbitrage pricing theory, Wikipedia
  13. Factor investing (History), Wikipedia
  14. Consumption-based capital asset pricing model, Wikipedia

The frameworks of Asset Pricing Theory

Each framework is a way of working the field has used: when it held, and the question it answers.

  1. 1930-1965

    Present value

    What is a future payment worth today?

    Fly to its tower in Nousfarer

  2. 1952-Present

    Modern portfolio theory

    How should an investor combine risky assets?

    Fly to its tower in Nousfarer

  3. 1964-Present

    Capital Asset Pricing Model

    What return should a risky asset earn?

    Fly to its tower in Nousfarer

  4. 1973-Present
  5. 1973-Present

    Option Pricing Theory

    What should an option contract cost?

    Fly to its tower in Nousfarer

  6. 1976-Present

    Arbitrage Pricing Theory

    Which common factors drive expected returns?

    Fly to its tower in Nousfarer

  7. 1978-Present
  8. 1990-Present

    Behavioral Asset Pricing

    Do psychology and sentiment move asset prices?

    Fly to its tower in Nousfarer

  9. 1992-Present

    Empirical Factor Models

    Which characteristics predict average stock returns?

    Fly to its tower in Nousfarer

  10. 2015-Present

How the ideas connect

Each arrow runs from an earlier idea to the later one it shaped. The list below says what passed between them.

Present valueModern portfolio theoryCapital Asset Pricing ModelNo-Arbitrage and Replication PricingOption Pricing TheoryArbitrage Pricing TheoryConsumption-Based Pricing and the Sto…Behavioral Asset PricingEmpirical Factor ModelsMachine Learning in the Cross-Section

Beyond the field

Where Asset Pricing Theory's ideas came from in other fields, and what they led to there: the strongest 12 of 75 links.

A city in Nousfarer: a red-carpeted stair climbing terraces to a crowned tower on an acropolis, more towers round it, forest and snowy mountains beyond.
A city in Nousfarer, the game played in the browser.

Asset Pricing Theory is a planet in Nousfarer

Its city has a tower for each of these 10 frameworks. Land, walk in, and climb one concept at a time.

Land on Asset Pricing Theory

Related fields

Fields its ideas are linked to

More in Financial Economics

See where Asset Pricing Theory sits on the Map of Human Knowledge