If you're a hedge fund looking to crunch massive quantities of data, it's generally cheaper to pay for space a la carte on Amazon's cloud than invest in million-dollar hardware.
That's the premise behind a spate of new finance-focused data shops turning out software that runs on the cloud. Ufora, a company profiled in Bloomberg Business, designs software that can process a trillion data points in minutes for the cost of a sandwich.
The technology is complex and involves a type of machine learning, or artificial intelligence, but computing power has become cheap enough that Ufora founder Braxton McKee can analyze a big market data model using only $10 worth of capacity on Amazon Web Services.
Ufora's hedge fund clients—like all hedge funds today—have good cause to want to keep costs low.
These privately-offered investments, which typically court only those who can invest at least $1 million, are having a tough time holding investors' interest these days.
That's partly because their high fees have become harder to justify given that recent returns have actually trailed those of cheap index fund-based portfolios, and performance is increasingly in step with that of benchmarks, meaning that mangers aren't adding as much value or diversification.