A gaggle of state and federal officials just announced a $16.7 billion settlement with Bank of America. The federal government and participating states will take part of that hefty sum, and the rest will go to select underwater homeowners and nonprofits. This record-breaking settlement stems from the crisis-era mortgage activities by Bank of America and the two companies it probably now wishes it had not acquired, Countrywide and Merrill Lynch. The various settlement documents present a smattering of examples of bad mortgage practices and email snippets, speculate that these might have violated federal or state law, and lay out a complicated formula for allocating the settlement loot. But it is difficult to assess whether the settlement is a proper response to the underlying conduct.


The settlement stems largely from Countrywide's and Merrill Lynch's practice of originating and packaging mortgages with high default risk into securities and selling those securities without revealing the poor quality of the underlying mortgages. For example, the statement of facts details how Countrywide made sure to securitize — rather than hold in its own portfolio — default-prone mortgages taken out by speculators (as opposed to owner-occupiers). And Merrill securitized loans that raised red flags in its due diligence processes. The statement of facts also cites an instance in which Bank of America, in a departure from its past practices, did not hire a third party to conduct due diligence of the mortgages underlying one of its securitizations.

Whether these practices constituted legal violations is unclear, but the existence of contractual remedies to address some of the conduct at issue raises questions about whether state and federal legal cases based on the same conduct were necessary. Some investors, using contractual representations and warranties, successfully forced Bank of America to repurchase the loans. In fact, one portion of the Securities and Exchange Commission's settlement is based on Bank of America's failure to adequately disclose to investors the potential magnitude of its contractual repurchase obligations.

The settlement agreement's greatest complexity comes in doling out the money. The Department of Justice's share is $8 billion, but the amount is allocated for the settlement of several different claims, only some of which are explicitly identified. Each participating state gets cash ranging (without explanation) from $23 million to $300 million. Some of the states pledge to use the money to address problems arising from the alleged misconduct. Even if the misconduct were precisely defined, the settlement document offers the states great leeway to distribute the money as they wish. Maryland, for example, will distribute some of the money to "certain investors," and New York will use some of the money on "anti-blight projects."

The consumer relief portion of the settlement is $7 billion (plus an additional $490 million to cover some of the taxes that consumers who get relief will have to pay). A complicated appendix explains what that $7 billion actually means. For example, forgiving $1 of a borrower's principal generally counts as $1 of the settlement amount. But if Bank of America services the loan rather than owns it, forgiving $1 of the loan counts only as $0.50 toward the settlement. On the other hand, if the loan is insured by the Federal Housing Administration, it counts as $1.75 of the settlement. A $1 donation to a nonprofit that rehabilitates abandoned houses or a legal aid association counts as $2 toward the settlement. The settlement prescribes a minimum amount of relief that must derive from relief actions in each settling state. The link between the elaborate relief scheme and the allegedly offending conduct is far from clear.

Settlements like this one are not very satisfying because they leave so many questions unanswered. Which specific state and federal laws were allegedly violated? (A few violations are clearly spelled out, but we are left guessing about other potential violations.) Is the amount of the settlement properly calibrated to these violations? Is the allocation among the settling state and federal government bodies appropriate? Will the beneficiaries of the settlement be people who were harmed by the alleged misconduct?

Peirce is a senior research fellow with the Mercatus Center at George Mason University.