Rather than communing with my laptop, tonight I had the pleasure of discussing the challenges of emerging technology with the inaugural meeting of the SCL Junior Lawyers Group.
The overriding questions seem to be: where will the key trends take the law and lawyers over the next 100 or 10,000 years? And how do we minimise our exposure to the downside - and maximise our exposure to the upside - of the next Black Swan?
- Fewer houses come to market, thereby propping up prices (or slowing their decline)
- Therefore fewer people go underwater on their mortgages
- Foreclosures have a devastating impact on the prices in the surrounding neighborhood
- Households preserve more wealth and are therefore more likely to spend rather than save
- Consumer confidence in housing increases
- More loan modifications (though how many successful ones is unclear)
- Time is bought for the rest of the economy to recover
On the other hand, [they said] this might not be good for the economy because:
- The problems in the housing market have simply been put on hold, not solved
- The excess inventory in the market won’t clear unless prices fall to a more natural level, and the sooner the inventory is cleared, the sooner the housing sector recovers and builders can get started again
- It’s unlikely that loan modifications will ever work on a large enough scale to make a difference
- Foreclosure delays are a distorting incentive on mortgage borrowers, who will be more likely to strategically default
[But] we have a nagging feeling that there are unintended consequences (or even straightforward expected consequences) that we simply haven’t thought of..."
Which encourages the aforesaid aggressive provisioning and selling practices:
"In its offer for the $1.5bn stock sale of privately held social-networking company Facebook, Goldman Sachs disclosed that it might sell or hedge its own $375m investment without warning clients. Under the deal, private wealth-management clients would be subject to “significant restrictions” limiting their ability to sell stakes while Goldman Sachs own holding can be sold or hedged at any time, and without warning."
It's fascinating how much data is publicly available. Here, for instance, is a summary of key data that describe the UK electricity market, including demand and generation by fuel type.
So what?
Well, apart from putting various fuel types into perspective, and maybe settling a few arguments, it's worth reflecting that the Hawthorne Effect was named after the electricity plant in which it was first documented. Henry Landsberger found that workers' productivity improved when he measured it to study the impact of light levels on their work, but declined again when his experiments ended. That suggests that when people know you're measuring their activity, it improves.
Alternative energy-generation measurement widget, anyone?
Amazingly, UK water companies actually produce a report that shows the water quality in your area. Just plug in your post code and tick the content you want to see. This is designed to tell you what 'contaminants' the water companies have been able to filter out, and allow you to decide what you might need or want to filter out yourself for whatever reason.
Unfortunately, there seems to be no mash-up that enables you to directly compare the remaining contaminants in your area with the capabilities of all the available water filters on the market - a "water quality/filter filter".
Something for the Open Data community, using API's and data from the various utilities and filter providers?
Derman's paper helps one get to grips with the financial crisis by succinctly explaining the shortcomings of financial models. Importantly, he points out that:
"Models are analogies, and always describe something relative to something else. Theories, in contrast are the real thing. They don't compare; they describe the essence, without reference."
While Derman gives examples of various theories that can be expressed in mathematical equations, he shows that finance is not capable of such expression. "There are no genuine theories in finance... Only imperfect models remain."
Derman suggests that we "use models as little as possible, and to replicate making as little (sic) assumptions as [we] can," and that we adhere to five rules:
While every financial axiom is wrong, the question is "how wrong, and can you still make use of it?"
"Build vulgar models in a sophisticated way", "using variables the crowd uses... to describe the phenomena they observe."
"A user should know what has been assumed when he uses the model, and... exactly what has been swept out of view."
Models can't be truly right. "You are always trying to shoe-horn the real world into one of the models to see how useful an approximation that is."
"To confuse the model with a theory is to embrace a future disaster driven by the belief that humans obey mathematical rules."
Of course, such limitations could also be said to extend to non-financial models deployed in and around the financial markets, demonstrating the enormous challenge inherent in the regulation of markets for complex products.
Financial models don't operate in a vacuum. The debt markets comprise at least as many models with inherent assumptions about how various aspects of those markets should operate as there are roles, functions, systems and controls, whether they be related to accounting, regulation, underwriting, collections, rating, marketing or audit. Everyone is operating on models - rating models, asset pricing and valuation models, accounting models that assume a company's health is reflected in its financial statements, regulatory models that may be either 'light touch' or heavily prescriptive. And everyone is operating on his or her own model of how these models work together.
The shortcomings of financial models apply equally to all of them.
However, all these models only ultimately 'bite' when a transaction occurs. And since transactions only occur between buyers and sellers (or their agents), only their beliefs about how models 'work' affect each transaction - capitalism keeps the authorities and everyone else on the sidelines. So the 'protective' models deployed by support functions and external actors can only be effective if they are properly deployed and fully understood by market participants. This seems impracticable, given that the likes of lawyers, accountants, ratings agency managers and bond traders have very different views of the same market, and differing attitudes to their employers, clients and so on.
So it's no real surprise that the narrative of the current financial crisis (e.g. "The Big Short") demonstrates the deficiencies in all these models and the manner in which they were deployed, as well as the (sometimes willful) lack of understanding of them amongst virtually all sub-prime debt market participants, regulators, intermediaries and advisers.
This poses an enormous challenge for the future development of markets for complex products. Better financial models, and better use of those models, won't avoid future financial crises. More rules and regulations cannot really be the answer, at least while they remain external to market participants and their transactions - and weakly enforced. Ultimately, we must either improve the knowledge of market participants relative to the complexity of products (through better education and training and/or by reducing the complexity of the products) or give regulators, or some independent creature - a more active role in transactions, if not as outright participants or potential participants.
Regulatory participation in transactions - or the threat of it - could be achieved partly through real-time transaction reporting from all significant financial markets, as is currently proposed in various initiatives around the globe. But that begs the question what the regulators will actually do with the transaction data.
As suggested in my previous post, perhaps adding short-selling to the regulatory repertoire would not only improve transparency and timeliness in dealing with market misconduct, but also provide regulators with a better feel for the limits in the models deployed in and around the financial markets.
