Algorithmic Trading: Innovate the Amazon.com Way
Monday, March 15, 2010 at 7:26AM
Michael Bigger
Written by Michael Bigger. Follow me on Twitter.
 
 
Some recently published articles and a video have shed some light on what fosters creativity and innovation at Amazon.com. The lessons learned from Amazon.com’s strategy can help all of us in our algorithmic trading activities. The articles and the video can be found at the following links:
 
1. Harvard’s interview of Jeff Bezos 
2. Jeff Bezos discussing Amazon.com’s acquisition of Zappos
3. How Amazon Innovates: Lessons in Strategy for Microsoft and Others
 
This is what I have learned from Jeff Bezos:
 
1. Be patient; think long-term. Start early and compound knowledge.
2. Learn from your mistakes. Iterate!
3. Learn and invent.
4. “It is always day 1 for innovation.” The same is true of algorithmic trading innovation.
5. “We are willing to plant seeds and wait a long time for them to turn into trees.” 
6. “Is it big enough to be meaningful?”
7. “Important to focus on what’s not going to change in the next five to ten years." What's not going to change with the market you are trading? You might want to build around that.
8. “Information perfection is on the rise.” Can you exploit that?
9. Information costs are going down. Can you exploit that?
10. “Close following doesn’t work as well in a fast-changing environment.” Come up with your own algorithm recipes.
11. “Ask ‘Why not?’” Why not develop your own algorithm? Why not develop a second one?
12. “Maximize the number of experiments you can do per given unit of time.” You never know what you will discover.
13. “Reduce the cost of the experiments.” Therefore, you can do many more experiments.
 
We have applied all these concepts in our trading activities. As an example, items 8 and 9 explain why we have spent so much effort on Twitter and its application programming interface (API). We want to mine information and mine it cheaply.
 
What say you?

 

Article originally appeared on (http://biggercapital.squarespace.com/).
See website for complete article licensing information.