Momentum Strategy Error - Index exceeds matrix dimensions..?

Hi guys !! I'm working on my master thesis and i really have to solve this algorithm and i'm running out of time. My work cinsists on calculating the Jegadeesh and Titman momentum strategy. I have the matlab code but when i execute it an error shows up "Index exceeds matrix dimensions" at the "idwinner(k,1:ncandidates) = stockid(idx(1:ncandidates));" line. I'm new to Matlab and i dont really know what is wrong with this code.

The code:

>> % load stockdata; stockdata = xlsread('data.xlsx','final','B2:S97'); stockid= xlsread('data.xlsx','final','B1:S1');

 [totalmonths,totalstocks] = size(stockdata);
period1 = [6];
period2 = [6];
startmonth = 14;
%get the number of stocks for each month
for i=1:totalmonths
nbrstocks(i) = totalstocks - sum(isnan(stockdata(i,:)));
end
for i = 1:length(period1)
for j = 1:length(period2)
for n = 1:2
p1 = period1(i);
p2 = period2(j);
% initialize
Rwinner = zeros(totalmonths,1);
Rloser = zeros(totalmonths,1);
idwinner = zeros(totalmonths,round(max(nbrstocks)*0.1));
idloser = zeros(totalmonths,round(max(nbrstocks)*0.1));
for k = startmonth:totalmonths+1-p1-p2-(n-1)
start1 = k;
stop1 = start1+p1-1;
ordermonths = start1:stop1;
start2 = stop1+1+(n-1);
stop2 = start2+p2-1;
holdmonths = start2:stop2;
nstocks = nbrstocks(k);
% order R for the ordering weeks
data1 = stockdata(ordermonths, 1:nstocks);
R1 = ones(1,nstocks);
for m = 1:length(ordermonths)
R1 = R1.*data1(m,:);
end
R1 = R1-1;
% select for the winners and losers
[B,idx] = sort(R1,'descend');
ncandidates = round(nstocks*0.1);
idwinner(k,1:ncandidates) = stockid(idx(1:ncandidates));
idloser(k,1:ncandidates) = stockid(idx(end-ncandidates+1:end));
% caculate the R for winners and losers in holding weeks
data2winner = stockdata(holdmonths, idx(1:ncandidates));
data2loser = stockdata(holdmonths, idx(end-ncandidates+1:end));
R2w = ones(1,ncandidates);
R2l = ones(1,ncandidates);
for m = 1:length(holdmonths)
R2w = R2w.*data2winner(m,:);
R2l = R2l.*data2loser(m,:);
end
R2w = R2w-1;
R2l = R2l - 1;
Rwinner(k) = mean(R2w);
Rloser(k) = mean(R2l);
end
%xlswrite(fullfile(['strategy_' num2str(p1) '_' num2str(p2) '_' num2str(n)
%'.xlsx']),Rwinner,'Rwinner');
%xlswrite(fullfile(['strategy_' num2str(p1) '_' num2str(p2) '_' num2str(n)
%'.xlsx']),Rloser,'Rloser');
%xlswrite(fullfile(['strategy_' num2str(p1) '_' num2str(p2) '_' num2str(n)
%'.xlsx']),idwinner,'idwinner');
%xlswrite(fullfile(['strategy_' num2str(p1) '_' num2str(p2) '_' num2str(n)
%'.xlsx']),idloser,'idloser');
end
end
end

Notice that when i enter a random matlab values the code works very well. Instead of extracting values from the Excel worksheet, i proceed like this:

>> % load stockdata; r=-1+(1-(-1)).*rand(715,215); stockdata=transpose(r); stockdata(stockdata > 0.5) = NaN; stockid=[1:715]; [totalmonths,totalstocks] = size(stockdata); period1 = [6]; period2 = [6]; startmonth = 14; .... Help plz

Ps: here is the Excel file http://www.mediafire.com/?a1gf5auef4qqxo1

[go go]

I met exactly the same problem. Bying usingthe same code I meet the same problem. This error confused me many many days. Have you solved your problem?If you have, could you tell me how to solve the error? I am quite worried about my thesis because the deadline is around the corner. Thank you.

Answer 1

[Geoff Hayes]

The error, Index exceeds matrix dimensions, generally means that the code is attempting to access an element in a matrix (or an array) with an index that is larger than the dimension being used for that index. For example,

 % A is a 12x2 array
 A = zeros(12,2);

 % throws error because 3 is greater than the second dimension size of 2
 A(12,3)

Since your error is being generated/thrown from the line

 idwinner(k,1:ncandidates) = stockid(idx(1:ncandidates));

there is either a problem with idx or with stockid. If you step through the code (with the debugger) you will observe that stockid is an empty matrix. This local variable is initialized at the second line as

 stockid= xlsread('data.xlsx','final','B1:S1');

If you open the Excel spreadsheet, you will see that there is the row from B1 to S1, but it is non-numeric i.e. MSFT, XOM, WMT, etc.

According to xlsread,

num = xlsread(filename) reads data from the first worksheet in the Microsoft® Excel® spreadsheet file named filename and returns the numeric data in array num.

So only numeric data is read in. Since there is none in this row, then an empty matrix is returned.

To get the text, just replace the above line with

[~,stockid] = xlsread('data.xlsx','final','B1:S1');

Now stockid will be populated with the non-numeric stock ids. However, this will require you to change how the idwinner and idloser arrays are initialized. Since the stock ids are varying length strings, these two arrays need to be initialized as cell arrays

 idwinner = cell(totalmonths,round(max(nbrstocks)*0.1));
 idloser = cell(totalmonths,round(max(nbrstocks)*0.1));

Try the above and see what happens!

[go go]

Thank you sooooooooooooooooooo much. The error does no longer appears. What a genius and kind man you are!!!!!!! Thank you for spending so much time in this problem! The feeling that the matlab runs smoothly is fantastic.

[Geoff Hayes]

Glad it worked out!

[go go]

I am so sorry to disturb you again. I found out although this code works well with socio's data. But the figure in the worksheet"Rwinner"and "Rloser" is very strange. The result is not a common return. Besides, When I apply this code to my owndata( my own data have some blank and not a number cells), figures in the "Rwinner" and "Rloser" disapper. I don't know what's wrong with this issue. Could you tell me how to solve this kind of problem? Thank you!

[Geoff Hayes]

go go - this should be posted as a new question as it does not concern socios's question. In this question, please attach only the relevant files, and describe what you mean by the figure...is very strange and the result is not a common return.

[go go]

Thank you! I get it.