# PCA related Query

I am currently working on a project in grad school where I am using PCA Approach.

I have 4 stocks. I used R to generate Eigen Values, Eigen vectors Eigen Values

Number Value Diff Proportion 1.00 3.51300808 3.18720008 0.8782 2.00 0.325808 0.17528152 0.08145 3.00 0.15052648 0.13986904 0.03763 4.00 0.01065744 0.00266

Eigen Vector
pc(1) pc(2) pc(3) pc(4) Stock 1 0.516215 0.4136083 -0.1234068 -0.7397439 Stock 2 0.5131561 0.31805179 -0.5016014 0.6196046 Stock 3 0.5048276 0.03720224 0.8298851 0.2346397 Stock 4 0.4640495 -0.85228354 -0.2108496 -0.1175298

How to arrive at the Eigen portfolio's. In my paper it says I can arrive at the weights by dividing the eigen vector by the Standard deviation of stocks & then multiply them by Stock returns. For ex, The stdev of stock 1 daily data is 0.0375. When I divide Factor loadings of Stock1, PC(1) I get a 98 something, which cannot be the weight of that stock in the eigen portfolio. Also Can you pls tell me how to interpret the eigen portfolio if i use PC(1) loadings & PC(2) loadings. I am really confused and stuck up. Luckily I came across this site. Any help is appreciated.

-
Maybe this paper can help you: math.nyu.edu/faculty/avellane/AvellanedaLeeStatArb071108.pdf –  neticin Apr 18 '14 at 22:03