What does the information coefficient (IC) stand for?
One way to judge the skill of an investment analyst or active portfolio manager is to look at the information coefficient (IC). The information coefficient shows how well the analyst’s predictions match financial results. If the IC is -1, it means that the analyst’s predictions had nothing to do with what happened, and if it’s 1, it means that the analyst’s predictions were spot on.
How to Figure Out the IC
IC = (2 × proportion right) – 1, where proportion correct is the number of right predictions the analyst makes.IC=(2×ProportionCorrect)−1, where ProportionCorrect is the number of correct predictions the analyst makes.
How to Explain the Information Coefficient
The information coefficient shows how closely predicted and actual stock returns match up. It is sometimes used to figure out how much of a difference a financial analyst makes. If the IC is +1.0, there is a perfect linear relationship between the expected and actual returns. If the IC is 0.0, there is no linear relationship. An IC value of -1.0 means the analyst can’t make a good guess.
If the information coefficient (IC) score is close to +1.0, the scientist is very good at making predictions. What happened, though? If “correct” means that the analyst’s prediction met the direction (up or down) of the actual results, then there are equal chances of getting the forecast right. Even a naughty expert would likely have an IC of around 0, meaning half of their predictions were correct, and half were wrong. A number close to 0 means that the analyst’s predictions are the same as what could happen by chance. This means that ICs close to -1 are not expected.
The information ratio (IR) is different from the IC. We can determine how good an investment manager is by comparing their extra results to the risk they are willing to take.
The Fundamental Law of Active Management says that a manager’s success (IR) depends on their skill level (IC) and how often they use their skills (IR).
A Picture of the Information Coefficient
For example, let’s say a financial analyst made two guesses, all of which were correct. The information coefficient would be:
IC=(2×1.0)−1=+1.0IC=(2×1.0)−1=+1.0
What would happen when an expert was wrong about half of the time?
IC=(2×0.5)−1=0.0IC=(2×0.5)−1=0.0
Not one of the forecasts was right, though. In that case,
IC=(2×0.0)−1=−1.0IC=(2×0.0)−1=−1.0
What the Information Coefficient Can’t Do?
The IC is only valid if you are an expert making many predictions. This is because random chance may be able to explain a lot of the results if there aren’t many guesses. The information coefficient is +1.0 if there are only two guesses and both are right. However, if the IC is still at or near +1.0 after a dozen guesses, it is much more likely that the person was right than they were lucky.
Conclusion
- One way to judge how good an investment analyst or active portfolio manager is to look at the information coefficient (IC).
- If the IC is +1.0, the predicted returns were precisely what they were. If it is 0.0, it means that there is no linear link. An IC value of -1.0 means the analyst can’t make a good guess.
- The information ratio (IR) is different from the IC. When you compare an investment manager’s extra gains to the amount of risk they take, you get the IR.
