Aaron's Part 2 Code for debugging

clc; close all; clear all;

data = csvread('returns.dat',1,1); % reads in the monthly returns data file for all 19 stock indices

indices = [1, 3, 4 , 6, 8, 9, 10, 14, 16, 18];  % The indicies allocated to me

subset=data(:,indices);  % your subset of the data
training=subset(1:25,:); % take first 25 months as training data to optimise portfolio weights
testing=subset(26:49,:); % last 24 months are test data to evaluate performance

N=10;     % number of stock indices considered in portfolio
T=25;     % months in training data
Ttest=24; % months in testing data
C=1000;   % dollar budget constraint
R=0.005;    % return threshold, to be varied
u = .50;  % maximum percentage an index can hold in each portfolio

%% %%%%%%%%%%%%%%%%%%%%%%%%%%%% QP %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

% Now write some code that defines the matrices and vectors needed to solve
% the quadratic programming problem to minimise variance. You need to
% define H, f, A, b, Aeq, beq, LB, UB in the quadprog function

[leng,wid] = size(training); %get the dimensions of the training set to use in the loop

H = cov(training);                            %H matrix
f = zeros(1,N)';                              %f vector
A = -1 .* sum(training)/leng;                 %inequality constraints
b = [-R*C];                                   %inequality constraints
Aeq = ones(1,N);                              %equality constraints
beq = C;                                      %equailty constraints
LB = zeros(1,N);                              %lower bound is zero
UB = repmat(u*C,1,N);                         %upper bound is mu x C

options = optimset( 'Algorithm', 'interior-point-convex','Display','off');
[xQP,FVAL,EXITFLAG]=quadprog(H,f,A,b,Aeq,beq,LB,UB,[],options);

if EXITFLAG == 1;
    FVAL=round(FVAL/0.01)*0.01;
else
    FVAL=0;
end %report to two decimal places

%Now experiment with R to find the range of values (from -5% to 5% in steps of 0.1) that gives a feasible solution
% with minimised variance. Lock in the largest value of R that minimises variance on the
%training data. Calculate x (the allocations) with that R value.

%initialising variables for data collection in the for loop
Rval = -0.05:0.001:0.05;
x = zeros(10,length(Rval));
FVAL = zeros(length(Rval),1);
i=1;

for R = -0.05:0.001:0.05
    b = -R*C;

    [x(:,i),f_value,EXITFLAG]=quadprog(H,f,A,b,Aeq,beq,LB,UB,[],options);

    if EXITFLAG == 1;
        FVAL(i,1)=round(f_value/0.01)*0.01; %report to two decimal places
    else
        FVAL(i)=0;
    end
    disp(R)
    i=i+1;
end