Information for Algorithmic Trading: Perspective from Jeff Bezos and Wilbur Ross

Written by Michael Bigger. Follow me on Twitter.

 

 

Recently, Jeff Bezos and renowned investor Wilbur Ross discussed information. Their views about information made us think about how to incorporate information into our algorithmic trading strategy. Let’s examine what they said and then examine the meaning of this for algorithmic traders. 

 

In this blog post, I discuss "Innovate the Amazon.com's Way". In the Harvard interview, Jeff Bezos made the following observations about information: 

 

1. Information perfection is on the rise.

2. Information costs are going down.

 

In the March Issue of Fortune Magazine, Wilbur Ross, in addition to reinforcing the points made previously, adds his own observations:

 

1. Market information is timely.

2. Market information is abundant and overwhelming.

3. Market participants have not demonstrated more ability to gain meaning from more information.

4. Therefore, the value of the expertise and ability to interpret this mountain of information goes to infinity.

 

The implications these statements have for algorithmic trading are profound. Embedding the expertise and ability to input and interpret a vast amount of information on a timely basis into an algorithm will be a significant source of value creation for algorithmic trading. It has never been cheaper and easier to access the internet information reservoir. The big value will be derived from developing some astute ways to interpret it. We are working on this. Are you?

 

Comments

[ed (faustroll)]

Thanks Michael,

I especially like those points made by Ross, who is quite a shrewd guy. I can point to two domains within the Ross comments that are both highly relevant and in high demand. One is the real time or timeliness or HFT aspect which is a huge subject and an area in which quite a bit of innovation is taking place. The other is in "doing interesting things with data," which is also a huge subject.

I'm keen on leveraging open source in both these areas. I should post about the route to developing an open source HFT platform when I get a chance.

[Michael Bigger]

Thanks Ed. If you have some articles (link) about the points you mentioned I would like to read them. Another area of interest to me, is to mash-up some Twitter content into my algo. Let me give you an example of what I am thinking about. I follow this guy who is very good with charts. He posts many charts on Twitter via Stocktwits. I need some standardization in his messages so that my computers can pick them up and mash them up into my algo. I told him to start rating his chart from 1 to 5 as a bullishness indicator (his perspective). So much useful content is posted but most of the originators do not metric their information. I think that could be a big area also.Your thoughts?