Forecasting Weather, Markets and Everything Else

Splitting a week between the sun on the west coast and storms on the east coast provided a good reminder of the most important part of making a good forecast: choosing the topic and the timeframe.  If you are forecasting weather a week in advance, you want to predict for a day rather than a specific hour, and it's easier to do in San Diego than in New York. 

I was fortunate to be out in Coronado last week with 200+ members of the StockTwits community for a discussion of investing principles and possibilities when this point was made in a variety of ways.  Scott Ingraham shared his perspective on the past, present and future for internet startups in the real estate space.  JC Parets charted technical insights to futures and forex markets.  Bill Gurtin’s fixed income talk revealed what an expert and true lifelong student really looks like.  Committing to particular markets and approaches (with at least a tendency toward certain timeframes) is a great foundation for successful forecasting – which, after all, is the core of successful trading. 

Since the person sitting on the other side of a transaction effectively turns most trades into a forecasting contest, the other factor to consider is the relative strength and weakness of your competition.  Josh Brown gave a clear and succinct example of this notion when discussing how to compete with high-frequency traders: low-frequency trading.  Other speakers had additional ideas on the road less taken.  Estimize is doing “wisdom of crowds” work that is going to make sell-side estimates a thing of the past.  Lucky Sort and ChartIQ are doing amazing things to put otherwise-unwieldy data sets into the hands of everyday traders.  There was enough material to do an entire conference on “the advent of unstructured data mining for alpha capture” (which was the title of a talk by Dan Mirkin of Trade Ideas), but we could have done an entire conference on at least ten other topics that were touched on, too.

Those of us working at Array were naturally energized to see smart and ambitious people in the alternative data and wisdom of crowds space (all much more likely to be collaborators than competitors, I believe).  And we embrace the idea of choosing both timeframe and topic carefully; we have, for example, been supporting a very successful fund for a while that emphasizes mid-range turnover of small and mid-cap equities.  But mostly right now we’re just excited to be part of a large, diverse and growing community of people working on a variety of issues in this space – one forecast at a time.