Loading field map
Finance as an academic subfield has been shaped by a persistent tension: should financial markets be understood through formal models that assume rational, optimizing agents, or through frameworks that incorporate the messy realities of institutions, information, and human psychology? The history of the subfield is a sequence of attempts to answer this question, with each new framework building on, challenging, or coexisting with its predecessors. The result is a field that today operates with a pluralistic toolkit, where models of perfect markets serve as benchmarks and frameworks that relax those assumptions provide richer explanations of actual financial behavior.
The earliest academic work in finance was largely institutional and descriptive. From roughly 1900 to 1935, the Institutional and Descriptive Finance framework focused on documenting how financial institutions—banks, stock exchanges, and regulatory bodies—actually operated. It addressed practical questions about market structure and legal rules but offered no formal theory of asset prices or corporate decisions. This framework was not so much replaced as absorbed: its institutional knowledge became background context for later, more analytical approaches.
The first major break came with Security Analysis and Fundamental Valuation, crystallized in Benjamin Graham and David Dodd's 1934 book Security Analysis. This framework shifted the focus from describing institutions to determining the intrinsic value of individual securities. It argued that careful analysis of financial statements, earnings, and assets could identify undervalued stocks, and that markets sometimes mispriced securities in ways a disciplined investor could exploit. Where the institutional framework had been content to describe, security analysis aimed to prescribe—to give investors a systematic method for separating value from price. This framework remains influential in professional investing, but its reliance on qualitative judgment and case-by-case analysis left it without the mathematical rigor that would soon transform the field.
The 1950s and 1960s brought a revolution that recast finance as a formal, mathematical science. The first pillar of this revolution was Modern Portfolio Theory (MPT) , introduced by Harry Markowitz in his 1952 article "Portfolio Selection." MPT addressed a question that earlier frameworks had largely ignored: how should an investor combine risky assets to maximize expected return for a given level of risk? Markowitz showed that diversification is not just a rule of thumb but a mathematical consequence of the correlation between asset returns. By formalizing the trade-off between risk and return, MPT provided a foundation for all later asset pricing models. It did not reject security analysis so much as change the unit of analysis from individual stocks to entire portfolios.
At almost the same time, Franco Modigliani and Merton Miller launched a parallel revolution in corporate finance. Their Modigliani-Miller Framework, first presented in their 1958 article "The Cost of Capital, Corporation Finance and the Theory of Investment," argued that under idealized conditions—no taxes, no bankruptcy costs, perfect information—a firm's value is independent of its capital structure. This was a deliberately provocative null hypothesis. It forced researchers to ask: if capital structure does not matter in theory, what real-world frictions make it matter in practice? The Modigliani-Miller Framework did not replace MPT; it coexisted with it, addressing a different question (how firms finance themselves rather than how investors build portfolios). But it shared MPT's commitment to rigorous, parsimonious modeling and became the benchmark against which all later corporate finance theories would be measured.
The next step was to derive a theory of how individual assets are priced in equilibrium. Equilibrium Asset Pricing, most famously the Capital Asset Pricing Model (CAPM) developed by William Sharpe, John Lintner, and others in the mid-1960s, built directly on MPT. If all investors hold the same optimal portfolio (the market portfolio), the CAPM argued, then the expected return on any asset depends only on its covariance with the market—its beta. This was a dramatic narrowing of the factors that could explain returns, reducing the rich qualitative analysis of security analysis to a single number. The CAPM did not reject MPT; it extended it by adding an equilibrium condition that gave MPT's efficient frontier a pricing implication.
The Efficient Markets Hypothesis (EMH) , articulated by Eugene Fama in his 1970 review article "Efficient Capital Markets," drew the logical conclusion from the CAPM and earlier work on random walks in stock prices. If markets are efficient, prices fully reflect all available information, and no trading strategy based on public information can consistently beat the market. The EMH did not replace equilibrium asset pricing; it provided a testable implication of that framework. If the CAPM is correct and markets are efficient, then stock price movements should be unpredictable, and deviations from the CAPM's predictions should be quickly arbitraged away. The EMH became the dominant paradigm in academic finance for the next two decades, and it remains a powerful benchmark even as later frameworks have challenged its strongest claims.
The 1970s saw two major extensions of the formal toolkit that had been built by MPT, the CAPM, and the EMH. Contingent Claims and Financial Engineering, launched by Fischer Black, Myron Scholes, and Robert Merton in 1973, provided a method for pricing options and other derivative securities. The key insight was that an option could be replicated by a dynamic portfolio of the underlying asset and a risk-free bond, and that the absence of arbitrage forced the option's price to equal the cost of this replicating portfolio. This no-arbitrage approach was methodologically distinct from the equilibrium logic of the CAPM. It did not require assumptions about investor preferences or the distribution of returns; it only required that markets be free of arbitrage opportunities. The Black-Scholes model transformed financial markets by making options tradeable at theoretically justified prices, and it provided the intellectual foundation for the entire field of financial engineering.
Arbitrage and Factor Pricing, most notably Stephen Ross's Arbitrage Pricing Theory (APT) from 1976, offered an alternative to the CAPM that was closer in spirit to the no-arbitrage approach. The APT argued that asset returns are driven by multiple systematic factors (not just the market), and that in the absence of arbitrage, expected returns must be linearly related to exposures to these factors. Unlike the CAPM, the APT did not require identifying the market portfolio or assuming that all investors hold the same portfolio. It was more flexible and more general, but it was also less precise: it did not specify what the factors were. The APT coexisted with the CAPM as a competing framework, and later empirical work (such as the Fama-French three-factor model) effectively merged the two traditions by using the APT's factor structure while retaining the CAPM's emphasis on systematic risk.
By the late 1970s, the core models of modern finance had achieved remarkable elegance and predictive power, but they rested on strong assumptions about rationality and perfect markets. Two frameworks emerged to relax those assumptions, each focusing on a different kind of imperfection.
Agency and Contracting Theory, developed by Michael Jensen, William Meckling, and others starting in 1976, relaxed the Modigliani-Miller assumption that managers act in shareholders' interests. In their landmark article "Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure," Jensen and Meckling argued that the firm is not a black box but a nexus of contracts, and that conflicts of interest between principals (shareholders) and agents (managers) create costs that affect firm value. This framework did not reject Modigliani-Miller; it showed that when agency costs are introduced, capital structure and corporate governance matter precisely because they can mitigate or exacerbate these conflicts. Agency theory transformed corporate finance by providing a rationale for debt, equity, and executive compensation contracts that the Modigliani-Miller framework had treated as irrelevant.
Behavioral Finance, which gained momentum in the 1980s, challenged the Efficient Markets Hypothesis more directly. Drawing on cognitive psychology, researchers like Werner De Bondt and Richard Thaler showed in their 1985 article "Does the Stock Market Overreact?" that investors systematically overreact to new information, creating predictable patterns of price reversals. Behavioral finance did not reject the EMH entirely; it argued that markets are not perfectly efficient because investors are subject to biases such as overconfidence, loss aversion, and herding. This framework coexists with the EMH in a state of living disagreement. Proponents of behavioral finance point to anomalies—momentum, value, and size effects—that the EMH cannot explain, while defenders of the EMH argue that these anomalies are either statistical artifacts or compensation for risk. The debate remains unresolved, and both frameworks continue to generate active research.
The 1990s brought a new set of concerns: how to measure and manage financial risk in practice, and how to model the systemic consequences of market imperfections.
Value-at-Risk (VaR) and Quantitative Risk emerged in the early 1990s as a practical framework for measuring the maximum loss a portfolio could suffer over a given time horizon with a given probability. Developed by J.P. Morgan's RiskMetrics group, VaR provided a single number that summarized a portfolio's exposure to market risk. It did not replace earlier frameworks; it operationalized them. VaR calculations typically rely on the same statistical tools—variance, covariance, and normal distributions—that MPT and the CAPM had made central. But VaR also revealed the limitations of those tools: during financial crises, correlations break down, and extreme events occur far more often than normal distributions predict. This tension between the formal elegance of VaR and the messy reality of tail risk would become a central theme in later work.
Financial Frictions and Systemic Risk, which took shape in the mid-1990s, addressed the limitations of the Modigliani-Miller and EMH frameworks by modeling how real-world imperfections—asymmetric information, collateral constraints, and interconnectedness—can amplify shocks and create systemic crises. The financial accelerator model of Ben Bernanke, Mark Gertler, and Simon Gilchrist (1994) showed that small shocks to borrowers' net worth can be magnified through credit markets, leading to large fluctuations in investment and output. This framework did not reject Modigliani-Miller; it showed that when financial frictions are present, the irrelevance propositions break down in ways that matter for macroeconomic stability. Financial frictions and systemic risk research has become increasingly central since the 2008 global financial crisis, and it now coexists with earlier frameworks as a distinct lens for understanding the role of finance in the broader economy.
Today, no single framework dominates academic finance. The leading frameworks—Modern Portfolio Theory, the Modigliani-Miller Framework, Equilibrium Asset Pricing, the Efficient Markets Hypothesis, Contingent Claims and Financial Engineering, Agency and Contracting Theory, Arbitrage and Factor Pricing, Behavioral Finance, Value-at-Risk and Quantitative Risk, and Financial Frictions and Systemic Risk—all remain active research programs. They agree on several core principles: that risk and return are related, that diversification reduces risk, that arbitrage opportunities are limited, and that financial markets are not frictionless. But they disagree sharply on how to model those frictions, how much weight to give behavioral biases, and whether markets are best understood as efficient or as prone to systemic instability.
The division of labor among these frameworks is largely determined by the questions they ask. MPT and the CAPM remain the standard tools for portfolio construction and cost-of-capital estimation. The EMH provides the null hypothesis for event studies and market efficiency tests. Behavioral finance explains anomalies that the EMH cannot. Agency theory guides research on corporate governance and executive compensation. Contingent claims models are used to price derivatives and manage risk. VaR and its successors (such as expected shortfall) are embedded in regulatory capital requirements. Financial frictions models inform macroprudential policy and our understanding of financial crises. This pluralism is not a sign of weakness; it reflects the richness of a subfield that has learned to use different frameworks for different purposes, while continuing to debate their underlying assumptions.