Hi all,
I was studying RSS and SSE.
To cut things short, I am trying to link the two with risks: systematic and unsysmatic risk; which goes with which?
I assume SSE is the sysmatic risk since it is actual and explained and RSS is unsystmatic since it is based on a prediction with uncertain outcome.
But I am afriad I am wrong.
Can somebody check this?
ANOVA involves analyzing the total variation (TSS) in your dependent variable. Variation in your dependent variable can be explained by your independent variable (RSS-explained variation) but it can also vary due to random factors other than your independent variable (SSE-unexplained variation).
If you want to relate systematic risk with linear regression, you might think of CAPM where beta is your measure of systematic risk. This is the slope (b1) in your linear regression.
Systematic risk in finance equals beta, unsystematic risk is known as diversifiable risk or risk that can be mitigated by having varied assets. RSS and SSE have to do with regression models and is not related to systematic and unsysematic (firm-specific) risk. Total risk = standard deviation.
They are not linked unless you happen to be doing a regression model for beta of an asset
For RSS and SSE when looking at an ANOVA table:
RSS is the Regression Sum of Squares or variation in y that is explained by the factors in the model
SSE (Sum of Squared Errors ) is the unexplained variation within a model that could be reduced if you added another variable or something to that effect.
RSS + SSE = SST which is total variation present
RSS / SST = R^2 for a model
Hope this helps
I think I am just confusiong myself.
Let me think over again because all I see of the CAPM model is just related with regression.
What are the independent variables in your regression model? Unless one of them is the market return, the terms systematic and unsystematic don’t even apply.
When I see S2000 I knew there is an answer.
I am still thinking about it because CAPM line just look like regression from my persepective
To get the one factor in CAPM – beta – you run a (least-squares, best fit) regression of the security’s (or portfolio’s) returns against the market’s returns; beta is the slope of that regression line.
I wrote an article on this: http://financialexamhelp123.com/beta/.