“How many of you believe that Sharpe ratio is bull**? How many of you use portfolio theory, optimal portfolio, and all that cr*p?…If someone starts talking about Sharpe, VaR, and all that, throw them out of your office.” –Nassim Taleb
Financial technology – something we all thought was complete – has been upended. Fundamental assumptions have been exposed as faulty. And now we have the opportunity to recreate our finance industry from the bottom up. We have a choice: a path of openness and information sharing, or more opacity and secrecy.
We will show you the potential of an open approach to assessing corporate credit risk by supplying corporate financial information that can be openly analyzed, avoiding conflicts that have been exposed in the current market.
Historically, the financial industry has valued opacity and closed solutions, with credit rating agencies functioning in a government-regulated oligopoly, enjoying immense protection from competitors. These firms haven’t been obligated to disclose their methods, creating a powerful incentive to alter the ratings based on favoritism, coercion, and bribery, and offering few means of convincing customers otherwise.
As a result of these market failures the state-of-the-art in credit-rating generation is woefully archaic and in need of retooling; however, the difficulty in assembling the data and the infrastructure needed to experiment with these scores prevents simple experimentation by developers.
With our latest project, FreeRisk, we aggregate accurate, accredited risk data, enabling users to generate crowd-sourced algorithms to analyze credit risk and allowing anyone to view the results of these algorithms. FreeRisk aggregates both all standardized XBRL data and public-domain financial data, as well as user-generated content incorporating unstructured data released in financial reports like footnotes, critical to accurate risk assessment. This system allows credit evaluators to focus exclusively on creating and applying risk analytics, instead of working through the complex data management tasks traditionally required to solve these problems or relying on black-box credit ratings.
Systems like FreeRisk can become a catalyst for innovation in financial technology, and by enforcing the same open source principles that generated so much new value in software, we can be assured that this knowledge will be shared, audit-able, and clear. More data, freer data, and easier data solutions will enable the rapid experimentation needed to create the accurate risk assessment we’ve been missing.
Toby Segaran is the author of the O’Reilly titles, “Programming Collective Intelligence” and “Programming the Semantic Web” and a contributing editor of “Beautiful Data” . He frequently speaks on the subjects of machine learning, collective intelligence and freedom of data at conferences worldwide.
Toby previous worked as a Senior Data Scientist at Metaweb before it was acquired by Google in 2010. He now works on large-scale data reconciliation problems at Google. Prior to Metaweb he founded Incellico, a biotechnology software company which was acquired in 2003.
Toby holds a B.Sc in Computer Science from MIT and is deemed a “Person of Exceptional Ability” by the USCIS. He loves applying data-analysis algorithms to everything ranging from pharmaceutical trials to online dating to financial risk models.
Jesper develops experimental online services designed to introduce emotional contexts into online relationships, creating more authentic experiences. He is the co-founder of Bloom Studios, developing novel data interface applications for web and tablet platforms. He is also an accomplished data scientist, working on problems including home valuations for Trulia, credit card fraud for Visa, and social network analysis for Visible Path. Jesper speaks frequently at international technology and design conferences and has appeared in print and broadcast media for projects like Avoidr, Freerisk, and his Foursquare privacy hack. He holds a B.Sc. in Physics from Haverford College and an M.B.A. in Econometrics from University of Chicago.
Comments on this page are now closed.