Khosla Backs Kurzweil the Quant

In addition to living forever, Ray Kurzweil has also apparently created a quantitative-investing tools company called FatKat. While there is undoubtedly a quant explosion going on in the market — driven by cheaper processing, widespread data, and better algorithms — it will be interesting see where Ray goes with this.

So far, so good though. Because according to PE Week, Ray has raised money from my friend Vinod of Khosla Ventures and KPCB fame. You know … he’s not a bad guy to have involved.

More here and here and here.

Related posts:

  1. Solo & Seeding with Vinod Khosla
  2. Khosla Ventures
  3. Vinod Khosla Presenting on Biofuels
  4. Vinod Khosla on the the Risk-Money Connection
  5. An Ode to Quant Investing


  1. mike says:

    I’ve never understood this business model. If you have a “black box” that can beat the market can’t you make a lot more money using it yourself rather than selling it?

  2. Not really, if you’re realistic. If you can beat the market by e.g. 3 percentage points a year, that’s fantastic, but I won’t make you a huge amount if you don’t have a big pile spare capital to invest.
    If you can get 50 million in money to manage, and take one point of that 3 point advantage, you’ll make a fortune.
    Even better if you can get that money to manage and flip some coins … :-) [heads I win, tails you lose ...]

  3. David says:

    More on the “quant explosion” and blackbox trading in yesterday’s FT:

  4. worth says:

    Can you program a system to make money for you? If it was an algorithm based on numeric data (which I assume it must be), it’s own activities would influence the system and make the system less effective. Also, as others mimic the performance, it would become less able to produce better-than-average returns, because everyone’s average would go up. But what makes my skepticism irrelevant anyway is that a successful predictive system/algorithm can’t be based purely on numbers, since numbers are always historical and cannot predict future performance. So can an algorithm be based on the words in reports and press quotes and analyses in addition to numbers?