How about this paper: http://www.optimization-online.org/DB_FILE/2010/12/2860.pdf

Thanks, but as far as I see does the paper not adress the question of return constraints. 

See this paper, it might  be useful to you. Consumption and Portfolio Decisions When Expected Returns are Time Varying

John Y. Campbell, Luis M. Viceira

NBER Working Paper No. 5857

Issued in December 1996

http://www.nber.org/papers/w5857

You may have to develop a model based on CAPM framework. CAPM has its own drawbacks and usually researchers use annual returns or daily returns in their empirical research. It gets complicated when they use trade by trade data. Your problem might be related to this type of a situation. 

Dear Jelto Borgmann

Greetings,

Please find the three attached files about the literature review about your question related for your topic may be useful for you and good luck

Best Regards,

Dear Prof. Borgmann,

                                       I have to say that if you take a look at my book on bounded Rationality and Incomplete Markets on my RGpage that unless you include the process of learning in Stock Markets and algorithmically estimate Stock Market Structural models  you will not be able to estimate as to whether the market data in question is cointegrated. Or you can use one of the statistical methods of cointegration like Granger causality tests. I hope this makes sense. SKM QC FEPS

Dear all,

thank you very much for your interesting contributions. which ended up in some interesting ideas to make my problem formulations "more" time consistent via constraints.

But so far I see, the study of the impact of expected return constraints on time consistency of the problem formulation is not very well known. If I use for example the nested AVaR as suggested in Shapiro (leading to time consistency for the problem formulation without any expected return constraint), I would expect that the multiperiod problem formulation  minimizing this risk measure under an expected return constraint for the final wealth (E[W_T] \leq d) is not time consistent. Maybe local expected return constraints can mitigate this problem? What would you say?

Thanks!

Jelto