What do Google, banks and chicken salad have in common?

In the last two months, the Department of Justice (DOJ) has filed a lawsuit against Google and initiated a process to revise its 1995 bank merger guidelines. The Google case is sure to be fought somewhere between the real world, cyberspace and the rules of behavioral science. And outside-the-box arguments and defenses will likely reshape antitrust enforcement for years to come.

Relative to that reshaping, some commenters suggest that antitrust enforcement efforts focused on technology companies may spill over onto banks. Perhaps, but that won’t change the bottom line that bank mergers are hard to classify as anticompetitive given that entities like the U.S. Post Office and Facebook are itching to join an already vast array of non-banks vying for consumer and commercial dollars.

Remember that when the antitrust laws were passed, banks controlled approximately 95 percent of consumer and business dollars. Today that number is less than 40 percent.

Moreover, the Dodd-Frank Act dealt with some of those very issues in 2010. It limited any one bank from having more than 10 percent of the nation’s deposits, and required that their acquisitions pass stringent tests regarding the potential impact on systemic stability. Not surprisingly, since then, none of the country’s trillion-dollar asset banks have completed a bank acquisition. 

DOJ’s 1995 current merger guidelines were published in prehistoric technological times, shortly after the first online banking site was launched in 1994. Historically, those guidelines have relied in part on the calculation of Herfindahl-Hirschman Indices (HHIs)(essentially the sum of the squares of deposit market shares) tethered to data from limited geographic and product markets considered to be reasonable proxies for the competitive nature of the market. Unfortunately, the guidelines often create a parallel statistical universe that does not correlate to the actual competitiveness of the market. 

In the bank merger cases that DOJ has lost (which is most of them), the banks have proved to the satisfaction of the courts that notwithstanding HHIs and other numbers presented by the DOJ, markets were actually behaving in a competitive manner. I worked on the approvals and then litigated several of the last bank merger cases brought by the DOJ in the early 1980s as a lawyer at the Office of the Comptroller of the Currency (OCC). In those cases, we argued that static HHI deposit calculations were unreliable, mechanical proxies for actual market competition.

A colleague of mine at the OCC who had litigated and won most of the previous cases for the agency brought in the 1970s told me that the DOJ would likely present the court with chicken salad when all it really wanted was a story about a chicken. He was correct. We were able to demonstrate that the various numbers and tables presented by the DOJ about banks in the market did not reflect the actual competition being created for example by money market funds, which were attracting billions of what would have normally been bank deposits. 

Assertions of anticompetitive behavior based on mathematical market share data have continued to be difficult to prove. That’s particularly true as banks have evolved from “locally-limited” businesses that can no longer be measured by discreet geographic and product markets as the Supreme Court described in the Philadelphia National Bank (1963) and Phillipsburg (1970) cases. DOJ’s guidelines became even more misfocused as competition changed. This history of events effectively evolved into a truce between a government that couldn’t prove bank merger violations and the banks that found it easier to divest branches than litigate with it. No cases have been filed in 35 years. 

In 1995, perhaps anticipating future concerns, Anne K. Bingaman, an assistant attorney general in DOJ’s antitrust division, recognized that “the great majority of bank mergers do not cause antitrust concerns” and that the DOJ is “familiar with the types of efficiencies that may be produced by bank mergers where a bank merger achieves significant economies of scale or scope.” Bingaman also found that consumers “benefit from lower costs and/or improved services” and confirmed that DOJ’s competitive analysis was taking all those factors into account.

Some of those concerns arrived on October 16, 2020, in the form of a letter to the DOJ from Sen. Elizabeth Warren (D-Mass.). Arguing that too many bank mergers are being approved, the senator warned against any changes that would “weaken the already insufficient process currently in place.” 

Warren’s letter goes on to urge the DOJ to ensure that the application of the antitrust laws to bank mergers should protect the interests of low and moderate income communities to prevent higher consumer costs. That goal is not expressly found in Section 7 of the Clayton Antitrust Act or the Bank Merger Act. The senator’s letter probably should have been directed to her colleagues in the Senate. 

The absence of bank merger challenges since the 1980s is in large part due to five factors.

First, the DOJ had been unsuccessful in proving violations of the Clayton Act using the statistical proxies that it employs.

Second, the continuous evolution in financial technology and exponential increase in the number of financial competitors upends many of the conventional principles necessary to prove that a market is non-competitive. Local, regional and global banks now have so many competitors that their mergers do not generally create antitrust issues.

Third, prefiling meetings with bank regulators weed out mergers that adversely impact competition that are not likely to receive approval.

Fourth, the divestiture of branches in concentrated “bank markets” can eliminate any micro-anticompetitive concerns by introducing new entrants into those markets. 

And finally, the increasing threats created by massive technology companies have soaked up much of the government’s resources and focus.

To make it even more challenging for the DOJ, the product and geographic markets that banks today compete in include cryptocurrency companies and exchanges, money market funds, mutual funds, insurance companies, investment banks, broker-dealers, online non-bank and peer-to-peer lenders, investment advisors, deposit brokers, hedge funds and private equity firms. 

If the government had thought that a bank merger violated the Clayton Act, I am sure it would have brought a lawsuit. The facts on the competitive ground belie the proof needed to make a case. In the future, to better evaluate the anticompetitive impacts of bank mergers, accurate competitive market analysis must include the use of artificial intelligence rather than maps; sophisticated algorithms rather than HHI ratios; technologically-driven measurements of relevant products and markets rather than artificial statistical proxies; and rigorous empirical data and cost/benefit analyses of the macroeconomic impact of mergers. It all must add up to a clear picture, not chicken salad.

This is where the Google case comes in. Much like the decisions in the last bank merger cases decided in the early 1980s, as the Google case unfolds, it is likely to reveal the keys to decoding the next generation of antitrust standards that relate to a digital and borderless world dominated by data. Proponents and opponents of bank mergers should pay close attention.

Thomas P. Vartanian, formerly a bank regulator at two different federal agencies and then a private practitioner for four decades, is the executive director and professor of Law at George Mason University’s Antonin Scalia Law School’s Program on Financial Regulation & Technology. He is the author of “200 Years of American Financial Panics,” which will be published in early 2021.