Learning from Baseball: A Bottom-up Approach to Systemic Risk

Reactions to the Obama administration’s plans to reform financial services regulation have been playing out true to form. Some critics want more sweeping change to prevent the kind of systemic risk that nearly brought us a repeat of the Great Depression. Others, worried that too much new regulation will stifle economic growth and innovation, argue the plan concentrates too much power in the hands of government regulators.

Nobody, however, is talking about baseball. But they should. What the reformers of Wall Street could use is the financial equivalent of Moneyball, an approach to fielding a winning baseball team based on statistical analysis and empirical evidence. The approach was deployed by the major-league Californian team Oakland Athletics (Oakland A’s) and made famous in a 2003 best-selling book by Michael Lewis, who also knows a thing or two about Wall Street.

The idea is that baseball experts were making decisions based on statistics like batting averages, batting statistics and stolen bases. The smart guys in Oakland dived into the data and saw that other measures, like on-base percentages (how often a batter reaches a base for a reason other than opposition error), were actually more reliable indicators of player success.

Like traditional coaches and scouts, financial regulators have been putting too much emphasis on the wrong metrics. Instead, a bottom-up, data driven and analytical approach to regulatory reform may do more to identify systemic risks in the financial system than an army of bank examiners in the most carefully calibrated multi-agency organisational structure.

Consider Tier 1 capital ratios, a traditional measure of bank balance sheet strength. Based on this metric, Bear Stearns, Merrill Lynch, Lehman Brothers, Washington Mutual and Wachovia were all adequately capitalised shortly before they failed or were forced to sell to other institutions. Similarly, too little attention has been given to liquidity, which is ultimately where all financial risk ends up. The administration’s plan addresses these issues head on.

Predictably, however, much of the attention on the new plan has centered on the zero-sum game of bureaucratic power inside Washington: who’s up and who’s down? Was this a win for the Federal Reserve? A push for the Securities and Exchange Commission (SEC)?

A Creative Approach

There can be no doubt that creating a new financial services regulatory structure is a necessary pre-condition for effectively managing systemic risk. But it’s not sufficient. That’s because successfully managing systemic risk in the financial sector in the 21st Century requires more creative approaches than traditional bureaucratic approaches typically foster. Identifying systemic risk involves detecting troubling patterns early. It entails connecting dots across different parts of the system. And it involves projecting how different variables can influence each other in unexpected ways. Just as important, it involves analysing data in ways that produce compelling enough findings to drive policy change – even as more traditional indicators are telling a different story.

What regulators need – however their agencies are constituted – are new tools and capabilities to understand and evaluate risk. To use another baseball term, they need their own version of Sabermetrics, the term coined by the highly influential baseball writer and statistician Bill James.

Here’s the good news: a lot of this data already exists in the information technology systems used by banks. It just isn’t being harnessed or analysed in way that regulators can use in a timely way. Policy makers would be wise to look closely at the vast data streams that could provide faster and more complete information. Such a bottom-up approach could yield important insights that could help shape the new regulatory regime in a more efficient and expedient way.

Now the bad news: there are barriers making it difficult for regulators to tap into the data. Today, for example, companies that service mortgage loans have data about borrower payment patterns and other information that provides real-time signals about loans that could go into default. Information like this is extremely valuable in designing loan modification programs aimed at keeping families in their homes, which is a national priority. But servicers aren’t able to share this real-time data without explicit permission.

Another barrier is that different banks use a maze of homegrown, customised systems that are difficult to access because they can’t communicate with each other. This makes it difficult to aggregate and analyse transaction information or trends across different institutions in real time.

And given the existing black-box mentality, where risk has been evaluated for individual firms and not across the industry, there have been few incentives for information technology providers to develop tools that measure and analyse systemic risk.

Moneyball, of course, is about incentives and data and doing more with less. The Oakland A’s, after all, embraced the approach because they needed to find a way to win without the budget to support the high salaries paid by wealthier teams. So they found a new way to identify talent in promising players that others didn’t see.

Conclusion

To anticipate and prevent the next systemic threat to our economy, regulators will need to see risks that others don’t. In some respects, the deck is stacked against them, for in a free-market economy, even a heavily regulated one, there will always be innovators and risk takers.

But as the recent economic crisis demonstrated, we don’t want to allow excessive risk on the part of big institutions that can bring down our financial system. Yet we also need a system with enough borrowing, investment and risk taking to generate growth, create jobs and compete in the global economy. The challenge for regulators is to make monetary policy and provide oversight that prevents systemic failure – but not the failures and successes that are a normal part of the business cycle. This is not an easy task, and reasonable people disagree on how to approach these questions. Some economists, for example, do not believe it is the role of the Fed to let the air out of financial bubbles. Wherever you stand on questions like this, regulators will have to wrestle with determining whether the next economic euphoria is a run of the mill bubble – or one that can bring down our financial system.

Dealing with these difficult questions, of course, requires an effective regulatory structure and appropriate regulatory powers. But the exercise of that power will be strengthened if regulators have access to more complete and timely financial data from banks and processors. This will give them greater ability to run different kinds of scenarios and better anticipate new risks. What we learn from this information can enable regulators to issues warnings, or put on the brakes, before things get out of hand. Again.

9 views

Related reading