Fundamental Factor model vs. APT

[mwvt9]

Can someone detail their understanding of the differences in these models?

 

[gmac2007]

APT provides little identification of risk factors, its different from fundamental factor i think

 

[McLeod81]

Completely different models:
APT: Intercept = Rf
Fund. Multi Factor: Intercept = Expected rate when factor sensitivities are zero (Expected return of avge stock)
APT: The multiple “Risk Premiums” are the premiums ABOVE Rf that result from a portfolio exposed only to that RISK
FMF Model: The factors = returns associated with various fundamental factors
APT: Sensitivities = regression sensitivities for each risk factor
FMF Model: Sensitivities = STANDARDIZED sensitivities for each Fundamental factor (ie (Factor - Avg Factor)/ Sigma of factor). Not from regression
APT: No-Arb assumption (FMF Model doesn’t have this)
Hope that helps.

 

[mwvt9]

McLeod81 Wrote:
——————————————————-
> Completely different models:
>
> APT: Intercept = Rf
>
> Fund. Multi Factor: Intercept = Expected rate when
> factor sensitivities are zero (Expected return of
> avge stock)
I thought this was the expected return for stocks with factor sensitivities equal to the market-wide average…..?
>
> APT: The multiple “Risk Premiums” are the premiums
> ABOVE Rf that result from a portfolio exposed only
> to that RISK
>
> FMF Model: The factors = returns associated with
> various fundamental factors
>
> APT: Sensitivities = regression sensitivities for
> each risk factor
>
> FMF Model: Sensitivities = STANDARDIZED
> sensitivities for each Fundamental factor (ie
> (Factor - Avg Factor)/ Sigma of factor). Not from
> regression
>
> APT: No-Arb assumption (FMF Model doesn’t have
> this)
>
> Hope that helps.
Thanks for the breakdown. You are going to kill this thing McLeod.

 

[McLeod81]

Thanks Mwvt. 2 more weeks to study our a$$es off so that we don’t have to sit and worry about the outcome all summer.

 

[mwvt9]

Did you see the point I made above. I would like your thoughts if you have the time.
You might not have to worry, but I will be…

 

[McLeod81]

The intercept is equal to the E(R) for a stock that has fundamental factors sensitivities equal to the market-wide averages. This will be the E(R) for a stock with the average fundamental factors (P/E, P/B, EPS, or whatever). Since the FFM uses standardized sensitivities and not regression slopes:
Stock avg. Sensitivity = (P/E - P/E avg) / Sigma avg
P/E = P/E avg, so the Standardized sensitivity = 0 (ie zero standard deviations from the norm).
Does that make sense?

 

[mwvt9]

Perfect sense. All terms will cancel except the intercept. Well explained.