How many bets to reach an information ratio of 1.0?

Question

You want an annualized information ratio of 1.0, and your forecasting signal has an information coefficient (IC) of 0.04 per bet. How many independent bets per year do you need?

Accepted Answer

The Fundamental Law, rearranged. The information ratio is skill times the square root of breadth: IR = IC 	imes \sqrt{BR}. Solving for the breadth (the number of independent bets per year): BR = \left(\frac{IR}{IC}ight)^2. Plug in the numbers. With target IR = 1.0 and IC = 0.04: BR = \left(\frac{1.0}{0.04}ight)^2 = 25^2 = 625 	ext{ bets per year}. Interpretation. With a per-bet skill of 0.04, you need about 625 independent bets a year to reach an information ratio of 1.0. The quadratic relationship is punishing: if your skill were half as good (IC of 0.02), you would need four times as …

How many bets to reach an information ratio of 1.0?

Your answer

More Statistics questions