 This topic has 4 replies, 2 voices, and was last updated Oct18 by ctownballer03.

AuthorPosts


Can anyone explain my confusion w/ IC?
From page 432 of CFA Derivs/PM Text:
“IC is the ex ante (i.e., anticipated) crosssectional correlation between the N forecasted active returns, Mu_i, and the N realized active returns, R_A_i”.I’m so confused by this.. Can someone explain the difference between Mu_i and R_A_i in the case of an exante measurement? This makes sense to me in the case of an expost measurement, where our difference would be the difference between our forecasted active returns and our actual active returns; however, I don’t understand what R_A_i is in an exante sense. How do we have realized active returns if this is measure before the returns happen?
Thanks for any clarification!

Exactly, your confusion is mutual ha ha. Exante IC is the correlation between their expected return and expected actual realized return (confusing I know). But when you said “How do we have realized active returns if this is measure before the returns happen?”, the answer is we do not have realized retuns before the returns happen, its impossible. It is the EXPECTED actual realized return. Naturally managers will assign themselves high exante IC’s because they think their gods and can predict their error margins, which will be small.. The expost IC, correlation between expected returns and actual returns is more of a judge of performance than the exante.

Such an absurd thing, but I think I follow. So we are essentially saying that our R_A_i (expected realized active returns) is our expected return and we are getting this formally through some pricing model such as CAPM, and then our Mu_i is our own subjective forecast? Then we are measuring the correlation between our own subjectively forecasted expected return and our expected return from a formal model? Is this right? It seems if this isn’t correct, then we would just be subjectively forecasting both expected returns and expected realized returns, and we would just assign ourselves an IC coefficient of 1 for perfect correlation lol..
I agree w/ your last statement, the metric seems absurd, but an expost IC would have some validity.

You got it bud. It is ridiculous lol. But I’m pretty sure you got it. But remember this, if a manager has high ex ante ic and low ex post then he/she isn’t doing very well. No one gives money to someone solely based off their prediction of subjectively forecasting the correlation of his/her expected active return with the forecasted expected actual active return. For example, I predict I’ll generate active returns of 2% but I think I’ll be wrong by .25%. Essentially forecasting my own error margin.



AuthorPosts
 You must be logged in to reply to this topic.