STREET DOGS: Knowing total supply and demand that move share prices is hard

What moves share prices in the short term is not mystical, in our view. It is simply supply and demand. The problem is that it is virtually impossible to precisely know the total supply and demand.

In some jurisdictions, like the US, institutions are permitted to purchase insight to order flows from retail brokerage firms such as Robinhood. While retail order flows may not comprise the total market, if an institution aggregates this retail flow with institutional flow, as well as place markers on momentum and quant strategies, this is often adequate insight or an edge as to how security prices will move.

There are numerous machine-learning algorithms trading markets. One example is Kensho. Kensho’s systems continuously gather text, information, and data from filings of publicly traded companies. It then organises this data and categorises which companies would benefit or suffer as a result of a change in a certain variable expected to be announced in the future.

For example, its systems will gather text, information, and data on say healthcare companies and predetermine that cuts in Medicare spending would benefit certain companies, while hurt others. As soon as these policy changes are announced, Kensho will commence buying securities in companies it believes will rise, while selling securities in companies it believes will fall. And it will do all of this before any one of us can finish reading three words in the news release.

The goal for any fund manager is to execute successfully despite these forces at play. It is impossible to be correct 100% of the time over the short term. (Short term, in this context means two days or seven days). To be right more than 51% of the time over such a short time horizon is already meaningful.

• From Andaz Private Investment