Stochastic Discount Factor (SDF) Framework
Consumption-Based Pricing Equation · Interactive Simulation Demonstration
γ — Coefficient of relative risk aversion (0.5–10), measuring aversion to consumption fluctuations
μc — Mean log consumption growth rate (annualized)
σc — Standard deviation of log consumption growth rate (annualized)
λ — Elasticity of asset payoff to consumption growth
Plain-Language Introduction
The Stochastic Discount Factor (SDF) can be understood as a "pricing engine." Imagine you have a future cash flow that you may receive, and the SDF tells you how much you should pay for it today. The core idea of the SDF is: money has different value at different times and in different states. During an economic recession and weak consumption, a unit of income is especially "valuable" (the SDF is large); during economic prosperity and abundant consumption, the same unit of income is less "valuable" (the SDF is small). This is why stocks that rise and fall with the market need to offer a risk premium—they give you returns "when you do not need money" and perform poorly "when you need money most."
What Problem Does It Solve?
The SDF framework unifies asset pricing theory. Before it appeared, each asset seemed to need its own pricing model (CAPM for stocks, term structure models for bonds, Black-Scholes for options). The SDF framework shows that all assets can be priced using the same SDF, with differences arising only from the covariance between their payoff structures and the SDF. It seamlessly connects macroeconomics (consumption, utility) with finance (asset prices, risk premiums).
Why Is It Important?
The SDF is a theoretical cornerstone of modern financial economics. It provides a unified perspective for understanding the source of risk premiums: risk premium = −Rf · Cov(M, R). This concise formula tells us that the expected excess return of an asset depends entirely on the covariance between its return and the SDF. In addition, the SDF framework directly gives rise to the "equity premium puzzle" (Mehra-Prescott, 1985)—under reasonable risk aversion parameters, the model predicts an equity premium far lower than historical data, which has driven the development of behavioral finance and macro-finance as a whole.
Application Scenarios
- Asset pricing: theoretical pricing of stocks, bonds, and derivatives
- Risk management: identify sources of systematic risk and evaluate hedging strategies
- Monetary policy analysis: connect interest rates with the real economy through the SDF
- Portfolio optimization: optimal asset allocation within the SDF framework
- Macro-financial modeling: the asset pricing module in DSGE models
- Factor investing: explain the economic logic behind Fama-French factors
Main Deficiencies and Limitations
- Equity premium puzzle: under reasonable parameters, it cannot match the historical equity premium
- Risk-free rate puzzle: the model-predicted risk-free rate is often too high
- The CRRA utility assumption is too strong: in reality, investor preferences are more complex (habit formation, long-run risk, ambiguity aversion, etc.)
- Consumption data problems: consumption data contain measurement errors and do not match the frequency of asset returns
- Homogeneous agent assumption: ignores heterogeneity, borrowing constraints, and market segmentation
- Tail risk: the lognormal assumption may underestimate the probability of extreme events, affecting disaster risk pricing