Use an annealer or a genetic algorithm or other stochastic optimizer to solve the American put problem stated in the caption of Fig. 3.12. (A genetic algorithm is available at the web page for this text, www.math.gatech.edu/∼shenk; it is set up to use the pricing algorithm under the name amerputExerBoundary (for the reader to supply)).

Data given in Algorithm 14     

Figure 3.12

Transcribed Image Text:

inputs: So, K, T, At, r, o, N and the exercise boundary B(t) E = 0; n = T/At i=1,..., N for S = So for t = At, 24t, ..., nAt = T Z~ N(0,1) = St St-1(1+rAt+ o√AtZ) if K-St≥ K * B(T-t) E = E +et(K - St) go to next i endif endfor t E = E + eT(K - St) endfor i E = E/N option price = E