LMM adopts Modern Portfolio Theory (MPT) to build well diversified and optimized portfolios. Covariance, correlation, the Capital Asset Pricing Model (CAPM), the Security Market Line (SML), Sharpe Ratio along with other quant methods will be utilized to assess portfolios. The intent here is not to take sides in the passive versus active portfolio management debate but to use cutting edge techniques and algorithms to reduce risk and maximize return. The core belief is that it is possible to find significant alpha and to beat the market over defined time intervals, if not necessarily in the long run. This is in line with the LMM goal to use of Machine Learning algorithms in the financial vertical.
2 thoughts on “LMM Goes Algo for Alpha”
Comments are closed.
To see the algorithms that LMM Labs has tested and to learn how to run your own backtests in Quantopian, go to the following URL: https://github.com/LMMLabs/Algorithmic-Trading.git or click on Algorithmic Trading under GitHub Repositories in the side panel.
You can run backtests on Quantopian algorithms by doing the following:
1. Create a Quantopian account at URL: https://www.quantopian.com/home
2. Login and go to the Research menu
3. From the research menu select Algorithms
4. You will see the default algorithms listed here, click on the New Algorithm button (Upper Right)
5. Once the Algorithms IDE opens, cut and paste your algorithm into the code window (Left Portion of Screen) or create new one from sctatch
6. You can now Build Algorithm to test it
7. Once tested you can Save the algorithm and Run Full Backtest
8. After the Full Backtest completes, go to the Notebook tab and run the Notebook
Running the Notebook on the backtest gives you a wealth of statistics and graphs related to the portfolio you are testing. The Notebook format and structure is almost identical to that of Jupyter Notebook.