Calculating the Portfolio Turnover

Calculating the equal weighted portfolio turnover.

function [p, k_sum]=turnover(old,new)
% PURPOSE: Calculating the equal weighted portfolio turnover
%---------------------------------------------------
% USAGE: [p, k_sum]=turnover(old,new)
% where: old, new are either character arrays, number arrays or cell arrays
% containing the tickers or CUSIPS of two different portfolio. They
% do not have to be the same size, that is, holding the same number
% of securities in both portfolios.
%---------------------------------------------------
% RETURNS: p = turnover percentage, defined as % newly added securites in the
% old portfolio (a % number)
% k_sum = total number of matches in the old portfolio for every cusip/ticker in the
% new portfolio (a vector)
%---------------------------------------------------
% Example: in the old portfolio, there are 10 stocks. If 5 were sold and
% 3 completely different stocks were bought, the turn over ratio would be: 3/10= 30%,
%
% It is not the exact turnover definition in portfolio management, as it
% is not value weighted. It is more accurate when portfolio is equal
% weighted by tickers/Cusips.
%
% The new and old portfolios do not have to be the same format, but have
% to be consistent in terms of CUSIP or ticker handles.
% For example, CUSIP could be in cell array format in the old portfolio,
% while in char array format in the new portfoliio, but they all have to
% be CUSIP.
%
% Both Tickers and Cusips must be in COLUMN vectors!
%
% Cusip is preferrable to ticker, at least in my experience. If you
% download data from FactSet,CompuStat, Cusip is constant through time
% while ticker C changes from Chrysler to Citibank.
%
% written by:
%
% Wei Li
% University of Chicago Graduate School of Business Class of 2003
% [e-mail]
% March, 8, 2004
%