Confusion on Complexity
Ninth District Highlights - March 2013
Published March 11, 2013 | March 2013 issue
All essential truths are ultimately defined by their contradictions. The best strategy is to let a competitor come to you. Except when you must take the action to the competitor. The virtues of simplicity in bank management and bank supervision and regulation have been extolled of late. Observers of banking and regulation advise us to avoid the complex. I concur. In many cases, community banks and supervisors do not need formal statistical models, to choose one example, to be successful in identifying and responding to risk.
But that concurrence is hardly a vote against complexity. Complexity is often required, for both banks and supervisors. Indeed, what is purported to be basic or simple risk identification or management actually requires a fairly sophisticated approach. The complexity of the analytical modeling, or lack thereof, does not define the sophistication of the underlying thinking. I offer a few examples across several major types of risk to make this point.
Very few banks need a credit risk management system that measures the probability that an obligor will default and the likely loss given default. Indeed, for commercial credits, most community banks do not typically need an econometrically derived measure of credit risk. Credit risk management systems based on accounting data and basic descriptive information typically are sufficient.
At least one major credit risk facing community banks, however, requires a sophisticated approach even if not a highly mathematical one. A concentration of credit risk can pose a significant potential risk to bank earnings and even to viability if a large enough “hit” to that risk arises. Many community banks cannot avoid concentrations given the economic locale in which they operate. However, they must manage that concentration to avoid the worst potential outcomes. When funds become available, bank management must decide if the bank should make another loan that would add to concentration, offset the concentration risk by buying a security not tied to the area of concentration, hold additional capital or reserves against the concentration or sell loan participations. Regardless, management and the board of directors must perform some kind of risk analysis to determine how much risk to accept and/or what steps to take to manage that risk.
The appropriate level of concentration cannot be determined without considering the major events that could make a concentration a major source of loss. Management must ask: Are the major potential sources of loss a fall in value of collateral or a fall in the income of the industry to which a bank is exposed? The bank must then determine how it will manage those potential losses. Bank management can and does carry out this type of analysis routinely. However, such analytics are hardly simple and, if taken seriously, this necessary process requires complex thinking.
The same points apply to supervisors as well. We must take care not to push down empirical analysis required of the largest institutions—that must use automated and empirical means to oversee their vast portfolios—to smaller institutions that can provide oversight without the same infrastructure. That cascading approach will not make the smaller banks safer. At the same time, we must ensure that the key analysis of the risk of concentrations occurs in a sufficiently robust manner at all institutions. For community banks, this means examiners must conduct in-depth discussions with management and the board of directors regarding the thought process behind their decision-making and review materials such as board minutes that are descriptive of the analysis rather than assessing complex models.
Community banks almost by definition have a strong sense for business opportunities in their areas of operation, and even in potential areas of operation. For many such banks, expansion into new products or new geographies does not require reams of data gathered by surveys, mapping software or focus groups. Banks often get timely observations directly from customers and potential customers on potential gaps in the market that the bank might fill. And most community banks know what they do well and do not do well.
At the same time, understanding the comparative advantages of a bank relative to potential competitors and the competitive dynamics of a new market is not so easy. How will competitors respond to the entry of a new bank into a market? How will the strategies that a bank uses with its current set of products serve it when it offers a new product? Do the dynamics of the new market differ from the market in which the bank currently operates? What appears to be a step to diversification can end up being a costly expansion followed by a costly retrenchment.
Again, my message is hardly that the analysis needed to expand is somehow beyond the ability of a typical community bank. Certainly, there are hundreds of contrary examples across the Ninth District. Rather, what may sound simple—expand by building on what a bank already does well—requires analysis at a sophisticated level to serve a bank well.
Interest Rate and Market Risk
The world of interest rate and market risk has built complexities on top of complexities. There is hardly a path of future rates that quants will not offer up for modeling. The likelihood that a given path may actually occur or that the path will actually matter for the portfolio of a bank does not always alter the opportunity to quantify its potential. And once the modeling has occurred, the risks associated with the modeling exercise itself arise. Yet more room for complexity.
I can easily see how the discipline of interest and market risk management may seem a sideline to some in banking, delegated to a vendor and checked in a rather cursory way. This approach, however, would be a big mistake.
Experience shows that interest rate and market risks are real and have led to many financial institution failures. Most community banks are heavily reliant on deposit funding. Management must consider the behavior of depositors in the face of a change in rates, particularly when the direction of rates changes relatively quickly and by a decent amount. Determining depositor response requires both data and analysis, which may be direct but still complex in some respects. The need for sophisticated analysis grows if a bank decides to buy a security whose value could change in possibly unexpected ways with a change in rate.
The lesson applies to supervisors as well when it comes to market risks. In the face of a crisis caused by credit risk, we appear to have an excuse for not putting rate risk at the top of our list. However, continued deprioritization of risk associated with a change in rates—particularly given the state of rates today relative to where they may be in the future—would not serve us well.
There has been some backlash against complexity in banking regulation. Some of the concern is valid, and supervisors need to be alert to situations where complexity is being imposed needlessly. However, that should not obscure the fact that risk management at community banks needs to be highly analytical and requires what can fairly be called “deep thinking,” even if it is not captured by formulas.