Error using fzero. Unable to perform assignment because the size of the left side does not equal the right side

I am writing a function to solve for a steady state for a complex economics problem set. I understand that the high number of globals can cause issues, but after speaking with my professor, I think this is the best method for solving my problem.

I appreciate any help, especially understanding what the error even means.

The globals are defined in another function, but are all parameters from alpha, delta, & beta that have exact values.

function [ zero ] = multshoot(k2)
% multshoot.m
% This m file solves for the terminal level of capital by solving the 2nd
% order difference equation by making a guess for the second starting value
% for output

% The global statement is used to pass some information into the function
% from the main program. The variable ovecms, which contains the time path
% for output is also passed back into the main program

global a b d kstar kvecms T tauh tauk alpha delta beta 

kvecms(2,1) = k2;             % Guess for second-period output
    
% Iterate on difference equation starting from period 3 using starting
% values for the first two periods for output

    for t=3:T     
    kvecms(t,1) = beta.*(1+(1-tauk).*(alpha.*((1-alpha).^a).*(kvecms(t-1,1).^(alpha-1+alpha.*a))-delta)+(((1-alpha).^b))+(1-tauh).*((1-alpha).^(1+d).*kvecms(t-1,1).^(alpha+alpha.*d))).*(((1-alpha).^(1+a).*kvecms(t-1,1).^(alpha+alpha.*a))+(1+kvecms(t-2,1)).*(alpha.*((1-alpha).^a).*(kvecms(t-2,1).^(alpha-1+alpha.*a))-delta)+kvecms(t-1,1))-(((1-alpha).^(1+a).*kvecms(t-1,1).^(alpha+alpha.*a))+(1+kvecms(t-1,1)).*(alpha.*((1-alpha).^a).*(kvecms(t-1,1).^(alpha-1+alpha.*a))-delta));
    % the equation is derived from the non-linear second-order difference
    % equation from the Euler equation, solved for kt+2. Everything is in terms of kvecms
    % The initial starting value is equal to 0
    % Likewise, the second starting value, kvecms(2,1), is equal to the 
    % guess
    end
    
% Examine the difference between the terminal level of output and the
% steady state level of output

zero = kstar - kvecms(T,1);
end

[Tommy]

In the error message, does it say the error is occurring in this function? And if so what line?

[Caitlin Gust]

Thanks for your response. The exact error is:

Error using fzero (line 306)

FZERO cannot continue because user-supplied function_handle ==> multshoot failed with the error below.

Unable to perform assignment because the size of the left side is 1-by-1 and the size of the right side is 0-by-0.

[Tommy]

Thanks! Somewhere in multshoot, you have some line

A = B

where MATLAB is trying to set A equal to B but can't because B is empty (i.e. []) and A isn't. It might be this line,

kvecms(2,1) = k2;

for example, if the k2 passed into multshoot is empty. Or it could be elsewhere (two other possibilities). It's hard to tell which without having the rest of your code. I would recommend running

dbstop if error

in the command window and re-running your code. MATLAB should then stop and enter debug mode when it reaches the error. For more on debugging: https://www.mathworks.com/help/matlab/matlab_prog/debugging-process-and-features.html