Date Tags blog

This is the first of several blog posts that we will be doing, one post for each of our home page headlines. Today's post is how to turn your hindsight market knowledge into foresight.

Regardless of how hard you try to incorporate hindsight into foresight there always will be unpredictable events.

"Reality looks much more obvious in hindsight than in foresight. People who experience hindsight bias misapply current hindsight to past foresight." They perceive events that occurred to have been more predictable before the fact than was actually the case.

—HERSH SHEFRIN, Finance and the Psychology of Investing

And we take Hersh Shefrin's advice quite seriously and try not to play Monday morning quarterback.

What we learn from our experience we incorporate into our tools. All our applications are in the process of constant improvement. But we don't apply blindly freshly acquired hindsight knowledge as we don't believe in one size fits all, each situation is different.

There are many factors that affect the market and politics is one of them, which complicate predicting what happens next and how the market will react; and by the market we don't mean US market only, as EU and Asia play an important role too. We are relying more and more on the data and number crunching; and we always remind themselves that data mining results should be considered as a good starting point but not as a final result, just ask all the numerous pollsters that have all the data and resources that money can buy and still were so wrong. A big part of what we are doing is separating noise from data. Along with stock market data, we are also looking at SEC filings and economic data.

It will not eliminate all of the mistakes, but we are getting further and further from the guessing game and our goal is to make foresight as close as possible to hindsight. Does it involve elements of AI ...? The answer to this question deserves a separate blog entry.

We are about to release our screening application that identifies patterns on historical data, currently in alpha. There is a possibility that parts of this application will be released as an open source, stay tuned.


comments powered by Disqus