Skip to content
Branch: master
Find file History
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
..
Failed to load latest commit information.
data
static
Readme.md
biblio.bib
portfolio_optimisation_using_socp.ipynb
robust_lp.ipynb
simple_SOCP.ipynb

Readme.md

Second Order Cone Programming

Second Order Cone Programming (SOCP) is convex optimization which extends linear programming (LP) with second-order (Lorentz or the ice cream) cones. Search region of the solution is the intersection of an affine linear manifold with the Cartesian product of second-order cones.

The figure below shows an example feasible region of an SOCP problem with 3 variables.

SOCP Example

SOCP appears in a broad range of applications from engineering, control theory and quantitative finance to quadratic programming and robust optimization. It has become an important tool for financial optimization due to its powerful nature. Interior point methods (IPM) are the most popular approaches to solve SOCP problems due to their theoretical polynomial complexity and practical performance.

This directory contains demonstrations using NAG's SOCP solver in Python.

Basic SOCP example

Portfolio Optimisation using Second Order Cone Programming (SOCP)

This demonstration is a walk-through of modelling techniques in portfolio optimization using second-order cone programming in the NAG Library. Models in portfolio optimization include

  • quadratic programming (efficient frontier)
  • quadratically constrained quadratic programming (tev portfolio)
  • optimization with objective of fraction of quadratic and linear (the Sharpe ratio).

General functions are enclosed for users to get the principle idea on SOCP reformulation. They provide one of the ways to build and solve their problems using NAG's SOCP solver and could be copy and paste into a model and reuse repeatedly.

Robust linear programming in portfolio optimization using the NAG Library

A mean-variance model with probability constraint using randomly generated data.

Data

  • stock_price.pkl - pickled data file contains daily prices of 30 stocks in DJIA from March 2018 to March 2019. It is used to estimate out-of-sample expected return and covariance matrix.
  • djia_close_price.csv - CSV version of daily prices of 30 stocks in DJIA from March 2018 to March 2019.

Poster

A 2019 poster discussing NAG's SOCP functionality is available on the NAG website

You can’t perform that action at this time.