# MATLAB MATLAB - implied volatility

#### mts_88

I have found the following function for calculating implied volatility in MATLAB.
When it finds a solution, the solution tends to be correct, but the function does not take into account, that the slope can be zero. If this happens, you divide by zero and the function return inf or -inf, which isn't correct at all. I have tried to take this problem into account, but the result is that the function does not converge. Does anyone have an idea how to implement this in the function?

% % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % %
% -------------------------------------------------------------------
% NEWT_IV.M
% -------------------------------------------------------------------
% The function uses Newton's method to find the root of
% the actual call price minus the Black-Scholes call price, solving
% for volatility.
% -------------------------------------------------------------------
% % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % %

% FUNCTION: newt_iv
% Input(s):
% - C: actual call price expressed in home currency
% - S: price of the underlying expressed in home currency
% - K: strike price of the underlying expressed in home currency
% - r: risk-free rate expressed as a decimal (5%=0.05)
% - T: Tenor expressed in years (6 months=0.5)
% Output(s): the approximate implied volatility using Black-Scholes
% and Newton's method. The approximated implied volatility and
% number of iterations taken to acheive an accurate result are
% displayed.
%
function ivol = newt_iv(C,S,K,r,T)

% Anonymous functions for Black-Scholes
d1 = @(S,K,r,vol,T) (log(S/K)+(r+vol^2/2)*T)/(vol*sqrt(T));
d2 = @(S,K,r,vol,T) d1(S,K,r,vol,T)-vol*sqrt(T);
bscp = @(S,K,r,vol,T) S*normcdf(d1(S,K,r,vol,T),0,1)-K*exp(-r*T)*normcdf(d2(S,K,r,vol,T),0,1);
vega = @(S,K,r,vol,T)S*normpdf(d1(S,K,r,vol,T),0,1)*sqrt(T);
% Initial guess
ivol = 0.2;

% Error tolerance
err = 0.0001;

% Set a limit to the number of iterations
i = 1; % number of iterations
ilimit = 2000000; % limit

cpdiff = err+1; % Cause the first iteration to occur (see next line)
while abs(cpdiff) > err
% Terminate the algorithm if the iteration limit has been reached
if (i > ilimit)
str = sprintf('ERROR: Implied volatility could not be found after %d iterations.', ilimit);
disp(str);
ivol = inf;
break;
% Otherwise continue the iteration
else
i = i+1;
end

% Calculate Black-Scholes call price for current implied
% volatility guess.
ccprice = bscp(S,K,r,ivol,T);

% Find the difference between calculated call price and
% actual call price.
cpdiff = ccprice - C;

% If the calculated call price minus the actual call price is not
% small enough, adjust the guess using Newton's method
if abs(cpdiff) > err
slope = vega(S,K,r,ivol,T);
y_int = cpdiff-slope*ivol;
ivol = -y_int/slope;
end
% If the calculated call price minus the actual call price is zero,
% a root as been found. Unless a root as been round, repeat
% the process using the new upper and lower bounds.
end
ivol;
end

Related Math Software Workshop News on Phys.org

"MATLAB - implied volatility"

### Physics Forums Values

We Value Quality
• Topics based on mainstream science
• Proper English grammar and spelling
We Value Civility
• Positive and compassionate attitudes
• Patience while debating
We Value Productivity
• Disciplined to remain on-topic
• Recognition of own weaknesses
• Solo and co-op problem solving