Portfolio Optimization using Ordering

I tried to noodle a bit with the idea of rank ordering for portfolio optimization after reading Satchell & Scowcroft's book on Portfolio Construction. They have two good chapters on the topic, including enough to implement an example. The problem I ran into was when trying to create an efficient frontier in risk/return space was that some of the efficient portfolios from ordered space just weren't efficient, or didn't plot up too nicely. I didn't spend a lot of time trying to solve the problem, just let it sit.

Now I've come across two papers by Almgren and Chriss where they show some math behind the models and run some empirical tests on their model. I'm looking forward to reading their papers a bit more thoroughly and implementing their model to see how it compares to some of the others I've played with