The Certainty Equivalence Principle <Back to Last Page>     <Full Glossary>

Definition of The Certainty Equivalence Principle: Imagine that a stochastic objective function is a function only of output and output-squared. Then the solution to the optimization problem of choosing output will have the special characteristic that only the conditional means of the future forcing variables appear in the first order conditions. (By conditional means is meant the set of means for each state of the world.) Then the solution has the "certainty equivalence" property. "That is, the problem can be separated into two stages: first, get minimum mean squared error forecasts of the exogenous [variables], which are the conditional expectations...; second, at time t, solve the nonstochastic optimization problem," using the mean in place of the random variable. "This separation of forecasting from optimization.... is computationally very convenient and explains why quadratic objective functions are assumed in much applied work. For general [functions] the certainty equivalence principle does not hold, so that the forecasting and opt problems do not 'separate.'" (Econterms)

Terms related to The Certainty Equivalence Principle:

• Forcing variables
• First order conditions
• Certainty eqiuvalent

About.Com Resources on The Certainty Equivalence Principle:
None

Writing a Term Paper? Here are a few starting points for research on The Certainty Equivalence Principle:

Books on The Certainty Equivalence Principle:

• Sargent, Thomas J. 1979. Macroeconomic Theory. New York: Academic Press.

Journal Articles on The Certainty Equivalence Principle:
None

<Return to Main Page>