I Don’t Think the LCR Means What You Think It Means

The Liquidity Coverage Ratio does not mean what I thought it meant and, in some cases, may not mean much at all.  As a reminder, the LCR is defined as the ratio of high-quality liquid assets (HQLA) to 30-day projected net cash outflow under stress.  Cash outflows are “projected” by applying fixed outflow and inflow assumptions to the bank’s liabilities and assets.  For example, 30 percent of the undrawn amount of committed liquidity facilities is assumed to be drawn. 

You might think, therefore, like I did, that a bank holding company with an LCR of 110 percent has enough HQLA to cover net cash outflows that were 10 percent higher than its projected net outflows and no higher.  You’d be wrong, though.  We’ve both been tripped up by the way that HQLA at the bank is handled when calculating the LCR for the consolidated holding company.

Specifically, when calculating the LCR of the holding company, excess HQLA at the bank is excluded.  “Excess HQLA” is HQLA in excess of projected net cash outflows.  The reason for the exclusion is that there are legal limits on the ability of a bank to provide support to affiliated entities, such as an affiliated broker-dealer.  Consequently, you don’t want the bank holding company to pass the LCR requirement if the broker-dealer is counting on excess HQLA at the bank to meet its liquidity needs.

While that makes sense, consider the following example.  The example is based on the 2021Q3 balance sheet of a large U.S. bank, and it is not contrived or atypical.  The bank has an LCR of 171 percent and the nonbank parts of the holding company have an LCR of 182 percent, but the holding company has an LCR of 110 percent.  Why?  Because only the bank’s HQLA up to the value of the bank’s projected net cash outflow counts toward the consolidated LCR, the bank is treated as having an LCR of 100 percent.  The bank is a lot bigger than the rest of the holding company, so when you combine that effective LCR of 100 percent with the remaining holdco’s LCR of 182 percent, you get 110 percent.

Table showing LCR Example-Based on a Real Bank’s Current Balance Sheet

There are some unfortunate implications of this result. 

  • Even though the bank accounts for the vast majority of the holding company, if the bank’s net cash outflow were to increase by 70 percent—nearly double—the holding company’s LCR would be unchanged.
  • Conversely, if the bank were to replace 41 percent of its HQLA with illiquid assets, the holding company’s LCR would not change.
  • Even though the holding company has an LCR that indicates its HQLA is 10 percent larger than its projected net cash outflows, the net cash outflows of both the bank and the rest of the company’s net cash outflows could simultaneously increase by 70 percent and the holding company would be able to pay the outflows using its HQLA.
  • If the holding company owned nothing other than the bank, whose LCR is 171 percent, and was financed entirely with equity, the holding company’s LCR would be 100 percent.
  • If a holding company just owned a small bank and had a large broker-dealer and the bank’s LCR really was 100 percent and the broker-dealer’s LCR was 120 percent – that is, a holding company with a vastly inferior liquidity situation than the bank described in the table — the holding company’s LCR could be above the LCR of the example bank.

What does the 10 percent in the holding company’s LCR of 110 percent mean?  It means that if the holding company’s total combined net cash outflow were to increase 10 percent, and that dollar increase were to occur exclusively at the nonbank part of the bank holding company (an 82 percent increase), then the nonbank part of the holding company could meet the outflow using its HQLA.  That is, the measure is only telling you something about the nonbank part of the holding company, but it is doing so in an essentially meaningless way.  As long as the bank’s LCR is 100 percent or higher, it would be more accurate to simply report the nonbank affiliates’ LCR of 182 percent.  Alternatively, the holding company LCR could be calculated as the weighted average LCR of the bank and the nonbank parts of the holding company (as long as the nonbank affiliates’ LCR was above 100 percent) or the minimum LCR of the two parts of the holding company.  All these alternative measures have flaws, but they are superior to the approach being used now.