The Federal Reserve
Federal Reserve Governor Dan Tarullo is reportedly set to reveal to the Senate Banking Committee today that the Fed is planning to boost capital requirements for too big to fail (TBTF) banks. Wall Street objects, saying it’ll place US banks at a competitive disadvantage to their European counterparts. This is said to be causing US banks to consider moving some trading operations out of the US. I may be missing something, but to me this is an unsatisfying debate, from both an epistemic and policy standpoint.
The postcrisis drive to boost capital requirements—in the main, an effort to require TBTF banks to hold more cash relative to their riskiest liabilities, so as to reduce the chance that they will go under—has been a mixed success. Basel III (not in force till 2019) is the leading international effort in this regard, but it has been widely attacked as weak, including for imposing too-low capital requirements. And so when some agency or commentator calls for new capital requirements in the US or another jurisdiction that exceed Basel minima, banks often say the proposed rule will place US banks at a disadvantage relative to European banks and hint that it will be self-defeating (since many regulations can be dodged by financial engineering or by moving the regulated activity overseas, though the latter can actually be quite difficult).
As regulators, finance professionals, and scholars begin to think about the new proposed rule, let’s stop and define our terms. If I were working for the Senate Banking Committee, I’d be curious about the following question today:
Do we know, in fact, that if this rule is adopted, banks in the US will be required to hold more capital than their European counterparts?
On the surface, the answer appears to be “yes,” because the proposed ratios in the Fed rule are higher than those in Basel III. But by itself this is not terribly revealing. The Fed’s new rule will presumably determine capital ratios on the basis of Generally Accepted Accounting Principles, or GAAP (which is mainly used in the US). Europe and most of the rest of the world use the International Financial Reporting Standards accounting system, or IFRS. This distinction can have systemic consequences when it comes to calculating capital ratios, because apples are not always being compared to apples. As this note from Ernst & Young explains (PDF):
IFRS and US GAAP differ substantially in their treatment of what is included in the exposure measure (i.e., the denominator of the leverage ratio [of assets to liabilities]).
This is largely because of differences in GAAP and IFRS netting rules, which in turn mean the volume of liabilities that need to be backed by a given percentage of assets differs. The International Swaps and Derivatives Association further explains (PDF):
The different offsetting requirements [imposed by IFRS and GAAP] result in a significant difference between amounts . . . [calculated] in accordance with IFRS and amounts [calculated] in accordance with U.S. GAAP, particularly for entities that have large derivative activities.
According to these FDIC figures, GAAP is overstating US bank capitalization ratios relative to IFRS, sometimes by 50% or more. In other words, differences in accounting treatment may be painting an unduly rosy view of the health of US banks, and it could be that new rules like the one the Fed is proposing are a corrective, or at least are not going to make Europe as appealing as opponents contend, because of IFRS.
Public and political debates over bank capital requirements should address the role of these accounting differentials. The banks themselves and the organizations that have a role in overseeing the implementation of capital requirements, like the Bank for International Settlements, clearly appreciate the challenges of measuring assets using two different metrics (see discussion of “calibration”). The political debate over the possible flight of trading operations should likewise consider these crucial accounting differences rather than assuming that a given ratio in Europe is automatically equivalent to that ratio in the US.