Branch Closures Round Two?

Branch closures are never easy work, but U.S. banks had an obvious place to start as they tightened up following the recession. By focusing on hopeless units, the industry closed more than 7,000 branches between mid-2009 and mid-2012.

Now, however, it appears that continuing expense pressures will require still more closures, raising the question of how to choose among the stragglers. In making the next round of cuts, banks will have to sort through branches that are either hovering at breakeven or, by simple measures, even showing nominal profits. The winnowing process is destined to touch off turf battles as well, with regional managers coming to the defense of their territories.

To deal with this complex scenario, regional banks will need much more information than has been commonly used to evaluate a network. Standalone measures of current and historical branch profitability will not get the job done.

There are three major questions that should be analytically considered. The first is the customer consequence of a potential branch closure, both the odds of losing valuable relationships and the impact on future sales in the area. The second is the network consequence, including market presence and the impact on public perceptions. The third is the potential for a fallback presence, such as automated teller machines or a thinly-staffed “express office,” which avoids a total retreat from a locale.

While these common-sense questions are familiar to many executives, they often need to be addressed with greater analytical rigor. The typical closure review is heavily influenced by a system-wide branch performance ranking based on current and historical financial accounting metrics, with management judgment then used to interpret the findings and consider the options. Winning banks bring robust market and customer context to the deliberations as well, better assuring that the dynamics of each local network are taken into account.

Why More Closures?

Typically as banks tighten down following a recession, they cut back to a level of capacity that strikes a balance between the goal of near-term savings and the goal of preserving long-term growth. Then they hunker down and wait, relying on the critical assumption of an eventual revival in market conditions and renewed expansion.

The problem this time around is that rebound assumptions are not working out. Many aspects of the pre-recession U.S. banking market either will never return or will take two to three times longer than normal to recover. So, while banks have largely completed a first round of cuts in branch outlets and staffing, some are still carrying way too much network capacity relative to regional market potential.

Based on a recent Novantas analysis of the top U.S. banking companies, the problem of network overcapacity is concentrated among a group of large to mid-size regional banks that are heavily burdened by “zombie branches,” projected to barely get by even under the most favorable of future circumstances. While banks in the top quartile of network health have only a 2% average concentration of marginal branches, second-tier banks have a 5% concentration; third-tier banks have a 10% concentration; and banks in the bottom quartile have a 17% average concentration.

Although the situation can be improved as banks continue to refine staffing models and productivity, that avenue of expense reduction can’t be expected to provide a full safety net. For many marginal branches, there will be no escape from the challenges of weak demand, flat margins, fee-crunching regulations and steady customer migration to alternative channels.

Even for less troubled networks, the pressure for cost reduction and branch closures will remain high. The fact is that the pace of branch growth has far outstripped population growth for decades. As of 2011, there was an average of one branch for roughly every 1,400 U.S. households. By contrast, the average branch covered about 1,600 households in 2,000; about 1,800 households in 1990; and about 2,900 households way back in 1970. This trend is unsustainable, especially considering how customers are downplaying branch usage in favor of remote channels.

Three Questions

Traditional branch review methods are insufficient for nuanced closure decisions, either limiting the ability to find savings or identifying the wrong branches for closure. What is needed is a new framework for pruning and reconfiguring the network.

Future Value. In considering the balances and revenues that might be affected by potential closures, banks often are tempted to take a historical view of the customer base. This orientation is reflected in efforts to allocate deposits and loans to various branches, based on where people opened accounts or transacted recently. But to make better closure decisions, regional banks need a clear view of the current and future potential revenue that a branch represents to the network as a whole.

This includes understanding the lifetime value of current customer relationships and ideally the defection potential for various categories of customers under different network situations and closure scenarios. It also includes a forward view of product sales, including the value of potential new business placed at risk by closure and the likelihood that nearby branches may be able to pick up some of the slack.

Realistically, it can be difficult to build company-wide consensus on measures of customer relationship value. However, the payoffs can extend well beyond improved branch closure decisions. Such value analyses can also be used in decisions about branch maintenance, upgrades and re-formatting; customer incentives and marketing campaigns; and in goal-setting and in re-allocating sales staff for maximum productivity.

Network Economics. The closure model of the future does not just look at individual branches; it looks at sets of branches to understand how simultaneous actions can magnify the impact on customers. One local branch closure may be viewed as an inconvenience; three or four may influence public perceptions about the overall company and set off a ripple effect of defections that range out of proportion to the local actions taken. Few closure models consider these differences.

A companion question is how cuts in network density may affect overall growth in account balances and household relationships in a local market. Although two local branches may be clear candidates for closure and no spillover effect on public perception is anticipated, it still may be better to preserve one unit and avoid a much larger collective impact on future revenues. Also, intuition is often not the best guide in choosing which branch out of several candidates should be closed in a given area. The network impact could be quite different even though the individual performance characteristics are roughly comparable.

Close vs. Downsize. Network pruning is often approached as a series of keep-or-close decisions, which overlooks options to scale back an installation. The goal is to preserve sales and service functionality at much lower cost, and the options can be as simple as leaving behind an imaging-enabled ATM to take deposits or as complex as overhauling the physical format for tightly focused operations.

Banks already know about express outlets given their experience with in-store branches. Especially given the declining need for teller-based transaction services, we see growing opportunity for small footprint installations that provide a valuable supplement to the local network. This is an example of how innovation can provide a tangible benefit to customers while boosting efficiency and flexibility for the provider.

Mr. Teas is a principal and Mr. Larson is a manager at Novantas LLC, a management consultancy. Mr. Teas can be reached at [email protected], and Mr. Larson at [email protected].