This Code does implement a new pricing engine for continuously sampled arithmetic average price options. The method is taken from the paper published under www.stat.columbia.edu/~vecer/asian-vecer.pdf. The paper describes a one factor pde method to price a greater class of asian options. The method uses a self financing strategie in the underlying to replicate the average.
I included a header and a source file for the engine. I also included an example main routine file that shows the engine code can accuratly reproduce these published prices. The engine files are meant to be put in the ql/experimental/exoticoptions path in the QuantLib sources.
1.) Implement seasoned options (averaging has already begun).
2.) Help User in finding proper grid dimensions. The default does well for a lot of cases. But if volatility is high or the option is deep out or in the money the code should help the user to find proper dimensions and truncations. Trivial cases (negative strike for example) could be trapped more nicely
3.) Implement new options classes like the average strike option. Also rolling future forward curve should be implemented in the future.
4.) Solution surface should be a result object. And should be able to be used in Excel for example