In the Kaplan Book, it mentions that Transfer coeeficient (TC)
can be thought of as the correlation between actual active weights and optimal active weights.
in another paragraph, it states that " more accurately, transfer coefficient is the cross-sectional correlation between forecasted active returns and the actual weights adjusted for risk
I read these two definitions again and again and just dont see they can be the same (connection). Can someone help explain why TC can be explained in these two different ways? Thanks!
The curriculum does the same sort of thing that Kaplan does: says that it’s the correlation between the forecasted active returns and the active weights, then later says, “Well, what we really meant is . . .” the definition above.
I consider it extremely unlikely that you’d have to compute a transfer coefficient.
in another paragraph, it states that " more accurately, transfer coefficient is the cross-sectional correlation between forecasted active returns and the actual weights adjusted for risk —> in case of multi-factor models i.e when we write Return = a + b1F1 + b2F2…
The model would have multiple factors (infaltion, GDP growth rate) instead of just Market risk premium. You can find derivation for the same in Reading: Multi-factor Models!!