HiI've got a huge problem...I'm trying to model pension fund portfolio optimisation and marry this with a Monti-Klien style banking asset allocation model I've already sorted out. The idea is that I have 3 classes of agents in my model banks, SPVs and pension funds. Banks optimally allocated their assets across differnt loan classes and securitise these loans accordingly. These securitised loans are then passed onto the SPVs for structuring and selling on the OTC marketspension funds then purchase these assets for purposes of their own portfolio optimisation.All this makes sense logically. The problem I'm having is exactly how I structure these assets and get my SPV and pension fund agents to derive their respective bid and ask prices on the OTC market with will be auction based.I assume the standard markov decission process for portfolio selection for the pension funds, however I don't understand quite how to model this. Firstly I need state transission probabilities however returns on the ABS/MBS/CDO/CLN etc are stochastic with the undeerlying loans being effectively American call options with 2-factor diffusion. Secondly, the transaction costs faced by the pension funds will only be realised when the actually enter the OTC market and bid on the ABS/MBS. This transaction cost will therefore change as the model runs.I'm not sure if what I'm propsing in my model makes sense though. If so and you understand what I'm talking about could you please advice me on things like reading or existing models things like that...even better possible hands-on help.many thanks

it sounds like you have a huge problemo.

HiThanks I know unpresidented proportions of HUGE!!! lolI can upload an overview of how the model is structured to work if that will help