Tuesday, November 11, 2008

Distance-Trust Technologies: Risk Modeling

A few Rust Belt Bloggers have already posted their Neighborhood Walk. Our idea is to communicate the splendors of our respective communities to people who have never have the opportunity to visit. The heart of the challenge is what I call the "proximity problem." Information can stream almost instantly to your desktop, but knowledge and trust don't come along for the ride.

International collaboration is a bit like maintaining a long distance love affair in that it rarely works. However, certain technological advances (e.g. e-mail) can improve the relationship, making it a bit more functional. GlobalHigherEd points to HUBzero as just such a tool, allowing "geographically dispersed researchers" to build knowledge. The international migration of scholars is big business, as well as the most visible part of brain drain hysteria. That Purdue University, ground zero for Rust Belt talent out-migration, deveoped HUBzero is likely no coincidence. HUBzero could allow the Postindustrial Heartland reclaim some of its investment in human capital because knowledge is becoming de-localized.

More difficult than exporting/importing knowledge without exporting/importing people is the long distance tranfer of trust. This is why the First National Bank of Orwell (Vermont) is booming while bigger financial institutions are on the rocks:

First National, whose main branch is next to a village store where local residents sit on the porch, drink coffee and chew the fat, has stayed successful because its employees know the area and their customers.

“Having the borrower sit in front of us is very meaningful,” [Bank vice president Bryan Young] said. “We’re not massive brokers, trying to underwrite a loan from the 55th floor of an office somewhere. There’s serious value in looking someone in the eye and understanding what their drive is, where they’re coming from and how serious they are about the project.”

The employees at the branches here and in neighboring Shoreham also know where a building project is and what it looks like.

“We benefit from knowing what house they’re talking about, what shape it’s in and what neighborhood,” Mr. Young said. “We benefit from a very detailed knowledge of our community, its people and its geography.”

In other words, a local presence improves risk assessment. This should sound familiar to any venture capitalist or entrepreneur. Startups do not fall far from the money tree.

Today's financial crisis can be understood as a failure of distance-trust technologies that allow large banks to assess the risk of lending money to a client whom you can't look in the eye. Risk models failed to capture the kind of knowledge that seems to only come from face-to-face interaction:

“Complexity, transparency, liquidity and leverage have all played a huge role in this crisis,” said Leslie Rahl, president of Capital Market Risk Advisors, a risk-management consulting firm. “And these are things that are not generally modeled as a quantifiable risk.”

Math, statistics and computer modeling, it seems, also fell short in calibrating the lending risk on individual mortgage loans. In recent years, the securitization of the mortgage market, with loans sold off and mixed into large pools of mortgage securities, has prompted lenders to move increasingly to automated underwriting systems, relying mainly on computerized credit-scoring models instead of human judgment.

So lenders had scant incentive to spend much time scrutinizing the creditworthiness of individual borrowers. “If the incentives and the systems change, the hard data can mean less than it did or something else than it did,” said Raghuram G. Rajan, a professor at the University of Chicago. “The danger is that the modeling becomes too mechanical.”

The above shortcomings bring me back to Theodore Porter's book, "Trust in Numbers: The Pursuit of Objectivity." One of the case studies is the story of actuaries in 19th century England. Life insurance agents based in London would have a difficult time competing with a small company in a village, who knew the local families intimately. Statistical innovations were used to empower the economies of scale advantage, putting the parochial institutions out of business. The system was far from perfect and these distance-trust technologies often failed.

Any scholar (such as Porter) dedicated to the sociology or history of science wasn't all too surprised by the subprime market meltdown. All of them studied how the lack of transparency often frustrated the transfer of knowledge. In fact, market deregulation stems from the realization that we can't know enough to effectively tweak transactions in a productive manner. Our risk models are not sophisticated enough.

I contend that there is a middle ground between universal risk modeling and a world with silos of local knowledge. The Tech Belt Initiative is a functional geography that can combine the best of social media technologies with the occasional face-to-face meeting to build the necessary trust. Bringing this post back full circle, the Neighborhood Walk is the first of the Rust Belt Bloggers to scale up local knowledge.

No comments: