The role of the database in the financial crisis

By Felix Salmon
November 23, 2009
this paper, by Norm Cimon, is still very interesting for where it places the blame for a large part of the financial meltdown: the rise of computing power generally, and the relational database in particular.

" data-share-img="" data-share="twitter,facebook,linkedin,reddit,google" data-share-count="true">

It’s over a year old now, but this paper, by Norm Cimon, is still very interesting for where it places the blame for a large part of the financial meltdown: the rise of computing power generally, and the relational database in particular.

All securitizations, after all, are based on computer analysis of a database of underlying assets. And Cimon explains:

The development of a good relational database design relies on the sort of cooperative effort that only exists in a handful of companies. That’s because extracting all the information from the involved parties is a daunting task. The rule of thumb is that it can represent upwards of 85% of the effort involved in re-engineering a corporation’s data assets, while the actual code to accomplish the job accounts for less than one-fifth of the work…

Now imagine coordinating the collection of such information across something as vast as the global mortgage-backed security industry which developed over the last 5-10 years. It never happened.

The investment banks and ratings agencies who were putting this stuff together simply never stopped to dot their i’s and cross their t’s — they were originating bonds at such a torrid pace that they didn’t have the time to make sure that reality was well represented in the way their models were constructed. Anybody who tried to create the kind of well-formed database that Ted Codd would be proud of would fall far behind the pack, and never sell anything.

Cimon concludes:

Weizenbaum’s book contains a very powerful argument for exerting human control over important business processes. He understood intuitively what we would end up doing to our accounting practices, our banking concerns, our investment houses, and all the other institutions when easily automated tasks could thoughtlessly be driven by limitless computing power. With networked computers now cast by all organizations, including the financial sector, into the role of wizard-behind-the-curtain, we all live in Oz. It’s long past time we pull back the veil and call a halt to the mindless application of this supreme and supremely dangerous creation before the damage gets any greater.

In the age of zero-price computing power, it’s far too easy to cut corners and trust in black boxes. What’s more, it’s incredibly difficult for any regulator to audit such things: I don’t think the Fed or the SEC has a department of relational-database integrity. Maybe they should think about starting one up.

More From Felix Salmon
Post Felix
The Piketty pessimist
The most expensive lottery ticket in the world
The problems of HFT, Joe Stiglitz edition
Private equity math, Nuveen edition
Five explanations for Greece’s bond yield
9 comments so far

You could end this with enforcement of a single regulation.A ratings agency is reviewed based on the 5-year result of their rating, not the dollar figure handed them at origination.I have no doubt we’ll never see this, because that would mean Goldman Sachs couldn’t game the system quite so well!

Posted by Unsympathetic | Report as abusive

Ack…I thought this was going to be about Al-Quaeda

Yes, and I heard that the rise in deaths due to automobile accidents increased with the number of roads built, because it enabled more people to drive in more places.The computer, and its software, like a relational database, are just tools. If you don’t know how to use the tool (e.g., you create an invalid model of a complex system that is rife with positive feedback), you’re going to run the risk of breaking something.You’re joking about the department of relational-database integrity thing, right?

It is the lack of a central database that is at the heart of the financial crisis. There is a significant information gap in the structured finance marketplace. Wall Street, through its servicing subsidiaries, has timely, accurate information. All the other market participants have out of date information.As predicted by Joseph Stiglitz in his Nobel Prize winning work on information asymmetry, when the market participants without the information believe they have been taken advantage of by the party with the information, they stop participating in the market. In short, investors stopped buying structured finance securities when they realized that Wall Street had used its informational advantage to its benefit (11/2007 WSJ Heard on Street column describing how Morgan Stanley used its informational advantage to short sub-prime mortgage securities). After all, what investor thinks they can make money trading with Goldman of the long run when Goldman has an informational advantage on each transaction.As for whether or not this central database can be put together in a timely manner, the simple answer is yes it can. My firm developed and patented a simple, low cost solution for doing this. The initial structured finance asset type was medical receivables. Medical receivables are far more complicated than mortgages.As for will this central database be put together again the answer is yes. Lloyd Blankfein in a Financial Times column last month said that best risk management practices are to look at every position every day. For structured finance, this means that you have to look at the underlying collateral performance every day. Congress is moving forward through the systemic risk regulator to make this database a reality.For more on my firm’s solution, please visit more on the role that information will play in restoring confidence in the financial system, please see

Isn’t it better just to get rid of securitization? Or *at least* simplify it back to the old german model? Of course far fewer people would make money from such a paired-down apparatus, but society would probably like them better anyway if they made chairs.

It is much better to keep securitization. Banks are actually bad holders of the assets they originate. Just look at the various lending debacles like loans to less developed countries where they blew up their balance sheet. Securitization allows them to remove credit and interest rate risk from their balance sheet and transfer these risks to investors who are willing for a price to hold them. In addition, banks don’t have adequate capital to support all the lending of modern economies.What the database does is to keep open the lending spigot throughout the course of the economic cycle.Investors need daily data on the underlying collateral performance in order to know what they own and to make portfolio management decisions. Daily performance data is the starting point for putting through the analytic and pricing models of choice for the investor and developing an independent value for the securities. This value is then used to compare to the prices shown by Wall Street to make buy, hold and sell decisions.With this type of data, investors can easily see if the actual performance matches expected performance. If an issuers loans aren’t performing, investors can either not buy the next deal from the issuer or charge a much higher cost to reflect the quality of the loans being originated. It is this type of feedback mechanism that would have prevented the worst of the sub-prime problems (selling garbage as if it were gold).

“It is much better to keep securitization. Banks are actually bad holders of the assets they originate.”You will, I trust, forgive me if I see that as an argument in favor of firing the bankers and closing the banks instead of an argument in favor of securitization.If the banks aren’t doing their jobs properly, why are we supposed to support them?

Banks are in a unique position for originating the assets (loans). To keep a modern economy functioning properly, you need to be able to originate loans over the entire economic cycle.You might recall the Savings and Loan crisis in the late 1980s. Savings and Loan were restricted by charter to making mortgages. No one accused them of not doing their job properly. In a rising interest rate environment, holding fixed rate mortgages is a great way to lose money. In fact, you lose money in two ways. First, you lose money from the interest rate mismatch between the cost of funds, which increases, and revenue from the mortgages, which stays flat. Second, you lose money from higher defaults as borrowers lose jobs because of the slow down in the economy caused by the higher interest rates. Was it the Savings and Loan’s fault that rates went up?Securitization solves the problem faced by financial institutions as holders of the assets. So long as the securitization market is functioning, banks can continue to originate the loans that their customers want and need.

First, I want to thank Felix for asking if he could post this after I sent it off to him. I’d come across the Wired article he’d written a few years back ( e/17-03/wp_quant?currentPage=all) about the equation used to measure risk. The same issue had a companion piece about a coding system for tracking investments. It was obvious that both dovetailed with what I’d put together so I forwarded him a copy.I wrote it to make a point that still escapes many analysts: the way to make sense of mortgage-backed securities and other such instruments is not through the use of ever more sophisticated equations for risk but simply by tracking what’s in them. We’ve known how to do that for a long time. That we chose not to says some very disturbing things about the knowledge and/or honesty of the people who steered us into this train wreck.As for knowledge, books such as The Black Swan only hint at the reality of the world we’ve created. Statistical theory and it’s attempt to model the after-the-fact distribution of market investments is useful but only to a point. The likelihood is that the trading systems we’ve created are non-linear. If so, it’s entirely possible that feedback can force them onto trajectories that don’t look remotely like the ones typically modeled using statistics.It’s easy to get fooled. Such systems can look stochastic even when all the variables are strictly determined. So this is not about rational versus irrational actors (see Justin Fox’s Myth of the Rational Market for example: review/Krugman-t.html). This can happen even as everyone acts quite rationally. It’s a property of the system itself. That may be what’s been built and what we’re watching operate.What’s deeply disturbing is that the quants seem to have no understanding of these systems, even though the major developments in the field have all come within the last 50 years. This isn’t about equations for risk, it’s about the possibility of slipping into a different set of trajectories or “orbits”, a different portion of the attractor for the system.Or do they know and not care? If so it’s dishonest. The money that’s been made in commissions and bonuses is almost beyond comprehension. A tacit admission that the system is prone to this sort of jolt may not be one they are willing to accept, or they can accept it to make public.Again, we can solve this by tracking what’s in those investment vehicles. That we can’t or won’t do it means one of two things. The financial institutions participating in these markets may be unable to organize itself well enough to do it. That’s a market failure. On the other hand they may be unwilling to do this because of the money that would be lost. Near-instantaneous feedback about the performance of the components would remove the latency in the current system. That would flatten those fees real quickly. That’s a moral failure and it means the system should be taken apart.

Posted by Norm Cimon | Report as abusive
Post Your Comment

We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see