Sort stocks into portfolios based on two characteristics

I have 3 matrix with the dimension 235x7690. The 235 represent months, the first month is june 1995, and 7690 are the number of firms. The first matrix represents returns, the second and third matrix represent characteristics of the firms (market equity and the third book-to-market equity). Now I wand to sort all 7690 firms in portfolios. I want to create 6 portfolio (high and low market equity and high middle and low book-to-markt equity). So with the 2nd matrix I want to separete them in high and low and the third matrix in high medium and low (the borders are the 30th percentil and the 70th percentil). The porfoliosort shoud be done every June. The I want to calculate the return of the six porfolios.

Could someone help me? Thank you

[John BG]

would it be possible for you to refine your target into smaller targets?

for instance, leave for now that you want them every June. Once the processing solved then you can solve the frequency to run it, ok?

[FC93]

I don't know if it would be possible. I don't even know how you would do it for 1 "period".

Answer 1

[Prateek Khandelwal]

Hi, To me, it appears you merely want to filter the data into 6 separate buckets and not sort it.

Once you've identified the values distinguishing high vs low market equity value and high vs medium vs low book to market equity values, you can use these to create another table with 'high','low','medium' acting as categorical values. This will be followed by filtering the data out based on these categories.

Have a look at: http://www.mathworks.com/help/matlab/matlab_prog/convert-table-variables-containing-strings-to-categorical.html