Online portfolio selection

Online portfolio selection (OPS) ^{[1] [2]} is an algorithm-based ^[3] trading strategy that seeks to optimise investment returns. It was first implemented in 2012 by Bin Li and Bin Hoi at Wuhan University.^[4][5][6] OPS is contrasted with more traditional approaches to portfolio optimization: ^[7][8] while mean-variance optimization seeks to balance risk and return, OPS specifically aims to maximize cumulative wealth, drawing on the Kelly approach^[9] to maximizing long-term expected value. Here, OPS sequentially allocates capital among a group of assets, employing, as appropriate, one of several portfolio selection strategies:

• Follow the leader^[10]
• Follow the loser^[11]
• Fuzzy approach
• Risk assessment^[12]
• Follow either winner or loser^[13]
• Pattern recognition based^[14]