Can we use machine learning to understand stock movements?

When designing a diversified portfolio, it sometimes helps not to look at individual holdings in isolation.

While Orbis Investments uses fundamental, bottom-up research to select its stocks, there’s another overlay the investment manager uses that help it to determine how said stocks are behaving in relation to one another.

If you read this, you will learn:

how using quantitative analysis and machine learning, Orbis can visualise relationships between different stocks
- how the ‘clustering’ of different stocks can reflect their relative price movements
  - - - why using this approach can help determine that a portfolio is sufficiently diversified by virtue of not taking ‘one big bet’, but rather a series of smaller ones

The opinions, advice, or views expressed in this content are those of the author or the presenter alone and do not represent the opinions, advice or views of No More Practice Education Pty Ltd. Our contents are prepared by our own staff and third parties who are responsible for their own contents. Any advice in this content is general advice only without reference to your financial objectives, situation or needs. You should consider any general advice considering these matters and relevant product disclosure statements. You should also obtain your own independent advice before making financial decisions. Please also refer to our FSG available here: http://www.nmpeducation.com.au/financial-services-guide/.

Can we use machine learning to understand stock movements?

When designing a diversified portfolio, it sometimes helps not to look at individual holdings in isolation.

Contact