Expected value of perfect information for a capital decision · Calculation guide

Choose between two mutually exclusive actions using signed net payoffs.

Calculation guide

Perfect information reveals the state before the choice. Weight by nonnegative relative probabilities.

EVPI = E[max(A,B)] − max(E[A],E[B])

Inputs

Relative scenario probability

Example entry 1.

Net payoff of action A

Example entry 100 (in the selected currency).

Net payoff of action B

Example entry 0 (in the selected currency).

Worked example

Mutually exclusive states

Relative scenario probabilityNet payoff of action ANet payoff of action B
11000
10100
ResultValue
Expected value of perfect information50
Expected payoff with perfect information100
Best expected payoff without information50

How each input affects this example

Each row starts from the worked example and changes one input.

InputOriginal inputNew inputAffected resultOriginal resultNew result
Relative scenario probability (row 1)12Expected value of perfect information5033.333333
Net payoff of action A (row 1)100110Expected payoff with perfect information100105
Net payoff of action B (row 1)01Expected value of perfect information5049.5
Expected value of perfect information for a capital decision →

Sources