Summary
The need for large commercial banks and their holding companies to satisfy the liquidity coverage ratio (LCR), a key liquidity requirement established as part of the post-crisis regulatory reforms, creates foreseeable consequences for market interest rates and presents challenges for monetary policy implementation. Actions by banks to comply with the LCR will push short-term interest rates lower and deposit rates and medium-term interest rate higher. As the Fed reduces the size of its securities portfolio, demand by banks for excess reserves to comply with the LCR may push market rates temporarily near or above the interest rate the Fed pays on excess reserves (IOER), with pressures increasing at quarter ends. The Fed could ease these pressures by ceasing to require banks to hold part of their liquidity buffers as excess reserves, by allowing banks to include required reserves in their buffers, or by establishing voluntary required reserves that can be included in the buffers.
Background
The LCR is defined as the ratio of a bank’s hiqh-quality liquid assets (HQLA) to its projected net cash outflows under severe liquidity stress. HQLA consists of the bank’s deposits at the Federal Reserve in excess of its reserve requirements (excess reserves), Treasury securities, and, to a more limited extent, agency mortgage-backed securities, agency debt, and certain other securities.
There are important interplays between the LCR and monetary policy, particularly as the Federal Reserve continues its gradual reduction of the large portfolio of securities that it acquired during the financial crisis. As explained in one of my recent blog posts, that decline also reduces the quantity of excess reserves available for banks to use as HQLA to satisfy the LCR. Recent shifts in certain money market interest rates have suggested that the quantity of excess reserves may be becoming insufficient for banks to satisfy the LCR without taking steps to reduce their projected net cash outflow. This blog post considers in greater detail the implications of the LCR for the relationships between interest rates (1) in the steady state, (2) while the Fed’s balance sheet declines, and (3) at month and quarter ends.
Steady state
Suppose we are in a steady state with reserve balances constant and banks and asset prices fully adjusted so that the quantity of excess reserves supplied by the Fed equals the quantity demanded by the banks at prevailing market prices. It is easier to understand the relationships between interest rates that must hold in this equilibrium by rewriting the LCR requirement as
If a bank is out of compliance, it can satisfy its LCR by holding more HQLA, increasing its projected cash inflow over 30 days, or reducing its projected cash outflow over 30 days. Banks will choose to comply in the least costly means possible.
Interest rates on HQLA or assets that provide a 30-day projected cash inflow will be lower by an “LCR premium”
As a result, discussed here in a 2016 TCH research note, overnight interest rates (such as repo and fed funds) and the rate on HQLA (excess reserves and T-bills) should all be approximately equal after adjusting for risk and any other characteristics, because they are all equivalent ways to satisfy the LCR. Because each asset provides the utility of helping to satisfy the LCR, those rates should be below the risk-adjusted rates on instruments that don’t provide a projected cash inflow within 30 days and aren’t HQLA. That “LCR premium” is similar to the money premium enjoyed by money-like assets because they provide money-like services (see (Stein et al 2014)).
Interest rates on liabilities (including deposits) with little or no 30-day projected cash outflow will be higher by the LCR premium
Moreover, the LCR premium should equal the spread between overnight rates and rates beyond 30 days because reducing your HQLA requirement by terming out your borrowing and therefore lowering projected net cash outflow should cost the same as satisfying your requirement by holding additional HQLA.
Retail deposit rates should also be higher by the LCR premium. The 30-day outflow rate assumed in the LCR for insured retail deposits is only 3 percent, so they are an attractive source of funding from an LCR perspective. As a result, banks subject to the LCR can be expected to compete more aggressively for deposits. Consequently, as the supply of excess reserves declines, and as market rates rise further from the zero lower bound, deposit rates should settle at a spread that is closer to market rates, and deposits should be more concentrated at larger banks, than was true historically.
The LCR premium will begin to boost term rates at about the one-week maturity
The LCR premium will not kick in precisely at the 30-day maturity point. Borrowing with a 31-day maturity only reduces a bank’s HQLA requirement for a day, so borrowing beyond 31 days (or using an instrument that has an evergreen 31-day maturity) is more valuable. On the other hand, in the United States, banks are required to hold enough HQLA to make it to 30 days, not simply the amount they would need at 30 days. In general, banks’ deepest liquidity deficit occurs at a horizon of just a few days. Consequently, term borrowing that extends for just a few days helps reduce a bank’s HQLA need. In sum, the LCR premium likely begins to apply for borrowings that extend at least a few days and then increases to a constant level as the maturity extends beyond 30 days.
In the steady state, the LCR will not push market rates above IOER…
Because the LCR premium should apply equally to all assets that can be used to satisfy the LCR, in the steady state, LCR scarcity is not a reason why IOER should be below other overnight rates, in particular, the fed funds rate. IOER would only be below other overnight interest rates when demand for excess reserves for the thing that excess reserves can uniquely do—prevent an overnight overdraft—comes into play. By this reasoning, if the Fed wishes to conduct monetary policy so that the fed funds rate is above IOER (aka a “corridor system”), excess reserves may need to decline to very low levels in the steady state (for a discussion of the decision the Fed must make about how to implement monetary policy, go here).
…even though supervisors require some HQLA must be excess reserves
While the LCR regulation treats excess reserves and Treasury securities the same, Fed supervisors have told banks that they must hold a certain, not publicly known, fraction of their HQLA as excess reserves. The Fed’s “LCR reserve requirement” creates an added function that only excess reserves will be able to satisfy. Even with the LCR reserve requirement, however, scarcity caused by the LCR cannot drive IOER permanently below other money market rates, including the fed funds rate, in the steady state. If IOER is below other interest rates, and HQLA must include excess reserves, then the average return on HQLA must be below the fed funds rate. But, if that were true, a bank could always satisfy its LCR more cheaply by replacing HQLA with a fed funds loan (and the associated cash inflow), so IOER must not be below the fed funds rate in equilibrium. (In the next section we describe how IOER could end up below the fed funds rate for considerable periods as the Fed’s balance sheet continues to decline.)
In the steady state, excess reserve scarcity will boost the LCR premium
As long as the Fed requires banks to hold a minimum fraction of their HQLA as excess reserves, the effective total aggregate amount of eligible HQLA will equal the supply of excess reserves divided by the LCR reserve requirement. For example, if the Fed requires one fourth of HQLA be excess reserves, then HQLA can be no more than four times excess reserves. As the Fed shrinks its balance sheet, asset prices will need to adjust to lead banks to demand a steadily declining amount of HQLA. In the steady state, banks’ demand for excess reserves to satisfy the LCR reserve requirement will be equilibrated to supply by a wider LCR premium (rather than pushing market rates above IOER). Increases in the LCR premium make it more expensive for banks to hold HQLA, so they will term out their borrowing and shorten the maturity of their lending. Those shifts, in turn, will boost term and deposit rates.
If the Fed requires a fraction of HQLA be excess reserves, it may not be able to adopt an implementation frame with IOER below the fed funds rate
But that suggests there could be a serious impediment to the Fed’s conducting policy so that IOER is below the fed funds rate. As I discussed in my 2016 TCH research note, if the Fed keeps shrinking reserves until it reaches a steady state where the fed funds rate is above IOER, but it requires that banks hold a material fraction of their HQLA as excess reserves, it could end up in a situation where HQLA is extremely scarce while the LCR premium is extremely wide. It is precisely such a conflict between the LCR and monetary policy implementation that Morten Bech of the BIS and Todd Keister of Rutgers have highlighted for several years (for a high-level explanation of their work, go here).
Transitional dynamics
But the above analysis only holds in the steady state. As the Fed shrinks its balance sheet and level of excess reserves declines, the relationships between interest rates may be different. Specifically, the steady state relationships assume that banks will have fully adjusted their balance sheets to satisfy the LCR in the least costly way possible. In reality, it will take time for banks to adjust their funding and liquidity profiles. At any given point in time, each bank will likely have established a specific amount of excess reserves that it needs to satisfy the LCR reserve requirement as well as its own account management needs.
As the Fed’s balance sheet shrinks, the fed funds rate could rise above IOER
As the quantity of excess reserves declines, banks may find themselves consistently falling short of their desired levels, driving up market interest rates, including the fed funds rate, relative to IOER. I say “temporarily,” but with reserves constantly declining, the system could be outside of its steady state for years.
The Fed may get confusing signals about when to stop shrinking its portfolio of securities
As discussed at pop up symposium on the issues we sponsored in May (for a summary of the discussion, go here), the Fed may want to shrink its portfolio of securities until it reaches the minimum size consistent with a monetary policy implementation framework in which the fed funds rate roughly equals IOER in equilibrium (aka a “floor system”). But because of the time it will take banks’ to adjust to declining levels of excess reserves, the Fed may find that market interest rates rise relative to IOER as its balance sheet declines, even though it has not actually reached the minimum level of excess reserves compatible with a floor system. In particular, if the Fed paused its portfolio-reduction for a period, market rates may move back down to IOER.
Period-end effects
The fed funds rate could rise or fall relative to IOER around quarter ends
There may also be some predictable patterns over reporting periods. Banks are expected to meet their LCR every day, but banks (and only the largest banks) only report their LCRs publicly on a quarter-average basis. If supervisors assess the sufficiency of banks’ holdings of excess reserves on a quarterly-average or quarter-end basis, then pressures on overnight rates could become more acute near quarter ends (respectively). Just as was true at the end of the two-week maintenance period over which required reserves are calculated, the fed funds rate could move up or down as a period average shortage or excess becomes evident. With excess reserves trending down, it could become typical that the fed funds rate would increase relative to IOER toward quarter end but then edge back down, as happened over the most recent quarter end.
The LCR premium could rise at quarter, or year ends
As discussed in my 2016 research note, the LCR’s impact on relative interest rates will also depend on bank’s capital requirements. Banks bound by risk-based requirements may prefer to hold additional HQLA (with zero or very low risk weights) than make short-term loans (with higher risk weights). Banks bound by a leverage ratio may prefer to meet their LCR by replacing short-term borrowing with longer-term borrowing or deposits rather than holding HQLA. Leverage ratio/LCR effects on term spreads could intensify quarter ends because European and Japanese banks’ leverage ratios are evaluated on a quarter-end basis. Similarly, at year end, GSIB (global systemically important banks) surcharges are calculated, which provides banks a strong incentive to both be smaller. As a consequence, large banks will have a further incentive to reduce HQLA by terming out their borrowing, putting further upward-pressure on term spreads.
Policy implications
The implications of LCR compliance for relative interest rates have implications in turn for the Federal Reserve’s QE portfolio winddown and eventual monetary policy implementation framework.
• If the Fed wishes to return to a relatively small balance sheet, then it cannot require that banks hold a material fraction of their HQLA as excess reserves. Banks hold about $2¾ trillion in HQLA, and if a material fraction of that amount must be held as excess reserves, then the Fed must have a correspondingly large portfolio of securities.
• The Fed could address this issue in part by classifying banks’ required reserve balances as HQLA and counting all reserves toward the LCR requirement. Not only would the change allow the Fed to shrink its balance sheet further, as my colleague Jeremy Newell and I argue in a recent blog post, it is both a sensible change and a legally required one.
• If the Fed wished to operate a corridor system with IOER below market rates, but also wanted banks to hold large amount of reserves for liquidity management, it could allow banks to establish voluntary reserve balances compensated at the target federal funds rate (as explained here.)
• As the Fed shrinks its portfolio it may find that the transitory increases in the fed funds rate could give false signals of structural shortages in excess reserves. On the other hand, the elevated demand for excess reserves to satisfy the LCR (or simply for cautions post-crisis liquidity management) could imply the balance sheet winddown will end sooner than expected.
