Basic CAPM question

[mathemagician]

Hiya ladies and gentlemen!

Currently taking an undergraduate investment course and we just began the CAPM (the very basic version)
Sorry for the lack of LaTeK, but its basically:

Expected rate of return on security S = risk-free rate + Beta of S * ( market risk premium)

One student asked a good question but phrased it badly; in essence I think he was trying to say/ask

"We're given an expected rate of return on a security S, but over what period of time is this expected rate of return to be realized?"

To try and elaborate a little more, lets say we have an expected rate of return on security S of 13.5% - clearly this rate of return isn't instantaneous (as soon as I purchase the security it doesn't immediately generate the 13.5% return), and I think its safe to assume that it won't return the expected rate after a few hours or a few days. So, how long until we are able to receive this return?

If there is no time period, is there a model that can somewhat accurately predict when the return will be earned?

Thanks in advance!

 

[Parallax]

CAPM is a single period model, with no required definition of what that period is. It could be a month, a year, a decade, etc..

A key part is that all of the underlying inputs (risk free rate & market risk premium) are based on the same time period. The assumed/estimated beta should be based on the same time period as well, but this is a more complicated topic of discussion (relationships can be different over different time periods, but most financial models assume stable covariance).

The most common implementation is a one-year return, but it can be whatever you want (once again, assuming all components of the model are on the same time scale).

Furthermore, it's better to view the CAPM as generating a required return, as opposed to an expected return. It is the return you require for bearing that amount of systematic risk; remember, it is a "pricing model". In equilibrium they will be the same.

 

[mathemagician]

CAPM is a single period model, with no required definition of what that period is. It could be a month, a year, a decade, etc..

A key part is that all of the underlying inputs (risk free rate & market risk premium) are based on the same time period. The assumed/estimated beta should be based on the same time period as well, but this is a more complicated topic of discussion (relationships can be different over different time periods, but most financial models assume stable covariance).

The most common implementation is a one-year return, but it can be whatever you want (once again, assuming all components of the model are on the same time scale).

Furthermore, it's better to view the CAPM as generating a required return, as opposed to an expected return. It is the return you require for bearing that amount of systematic risk; remember, it is a "pricing model". In equilibrium they will be the same.

Thank you very much for the thorough answer!

 

[diegosanaz]

It is standard for CAPM models using annualized returns and volatilities.

 

[quant_wannabe]

The CAPM is part of a partial equilibrium setting in which investors care about mean and variance and make a two-period decision. It is a general parable, not a practical model.

As long as the market portfolio can be measured over any time horizon, as can be the Rf, any time horizon is appropriate. Just make sure to annualize / scale the returns appropriately.

 

[IntoDarkness]

i just derived this crap for work. too many assumptions i cant even stand it anymore