Week 4 Flashcards
Can two things be negatively corrrelated but in the same direction consistently over the long term?
Yes, this is because covariance is given by the sum of (X- its mean)*(y- its mean), the mean in both cases could be the same, but one could be below it’s mean on one day while the other is up, and vice versa.
Why is correlation commonly used? What does it mean?
Covariance can take on crazy values, hence we often scale it to get the correlation, putting the end result between 1 and -1/ It is typically a measure of linear dependence between X and Y.
How can we find the R^2 value using correlation?
R^2 is given by correlation squared.
When will the correlation be equal to the slope between two variables? How come?
When the two variables are one variable regressed on itself with lag. This is because the standard deviation will be basically the same, making the equation almost the same as the slope equation (covariance(x,y) /variance(x)).
What does two fund separation involve?
First we generate a minimum risk frontier by combining stocks into portfolios, this works when the correlation between the stocks is not high, in which case the diversification lowers the standard deviation of the portfolio.
We then include a risk free asset, using it to create a capital allocation line by drawing a tangent to the minimum risk frontier curve to be as north west as possible.
Now, a client can pick different combinations of the risk free asset and risky portfolio , in order to get thte maximum return for a given risk tolerance.
If we do a probability distribution graph for 2 dependent variables, instead of the normal 1, what space will it be? What replaces area in this case?
3D space, the volume underneath the graph will be 1.
