Galli
#1
Multiple ways to derive these formulas, driving me nuts. Are these correct?
Beta:
Beta = cov (i,m) / variance (m)
Beta = correlation * (st dev i / st dev m)
Beta = correlation * st dev i * st dev m / variance m
Beta = variance i * variance m / variance m
CoVariance
Cov = variance i * variance m * correlation(i,m)
Cov = B * variance m
Cov = B1B2 * variance m (ICAPM)
Cov = correlation * (st dev i / st. dev m)
B= Var i* var m/ var m is absolutely wrong. It equates to B= Var i; which is not right
Cov = cor * SD i * SD m (not Var; SD=SQRT of Var)
They’re all related. Just get the basic ones and then play with algebra
for eg: B= Cor * SD i/ SD m . & = Cor SDi * SD m / (SD m * SD m) = Cor SD i* SDm/Var m
another example:
Cov= Cor * SDi* SDm ; multiply the right term by SDm/SDm ; leads to Cov= Cor* SDi * (SDm*SDm)/SDm;
(SDM * SDM) = Var m & (Cor*SDi/SDm) = B ; this leads to Cov= B * Var m
Just few practices and playing with the terms and you’ll imprint this in your brain
I only remember below. It should be able to derive others.
Beta = cov (i,m) / variance (m) = cov (i,m) / st dev (m)^2 Correlation (i,m) = cov (i,m)/ st dev i * st dev m
so which one of these are wrong?
Galli
#5
the last Beta one is wrong