Hedge Funds as Software Companies? Nah. And You Don't Need a Ph.D to Rock It

My Infectiously Greedy friend Paul Kedrosky, major domo of the upcoming Money:Tech conference, ripped out a little ditty where he implied that developers were becoming even more important and prevalent at hedge funds than at software companies. Now Paul knows that I agree with him much of the time, but this does not happen to be one of those times. I think Mr. Greed is mixing up two separate and distinct concepts: models and data. Sure, one can benefit from multi-lettered, Ph.D-laden development teams to design new statistical arbitrage models supported by ultra-low latency execution platforms. This is the province of Renaissance, the quant part of DE Shaw (which is now a multi-strategy mega-firm), AQR and other like-minded nerdos. But this is not - repeat NOT - what makes Web 2.0/3.0 so exciting. It's the data, stupid.

And you don't need an army of topologists or theoretical physicists to either get the data or extract value from it. Depending upon how wide one wants to cast the net or how specialized one's needs might be, there are lots of free tools out there to harvest and aggregate Internet data that can add real value to an investment process. Tools that didn't exist even 12 months ago. Sure, if you want a customized, tickerized feed of alternative data, stitching together memes across multiple data types, neat, tidy, spam-free and delivered exactly how you want it, that's different. But plenty of folks are just scratching the surface of all the data that's out there on the Internet and don't need this level of customization and service level to extract value and meaning.

I think the other big thing besides the data itself is how to extract meaning from the data given your perspective on trading and investment. Long-term investor looking to fill in the investment mosaic around a specific idea or theme? Trader looking for indications of when it is best to exit a position in light of volatile market conditions? Quant seeking to develop an analytical framework for a data type around which they've got a theory for extracting statistically-significant levels of alpha? The potential uses of data and related analytics are virtually limitless. And it's not about hard-core coding, it really isn't. It is about the way to incorporate alternative data into one's investment process. Broadening one's horizons, and being forced to use new tools and techniques for tapping into this new river of data and meaning is hard. But ultimately, the payoff will be new perspectives, tools and approaches for making money that didn't exist before. And this, my friends, is what investing is all about.