The minutes from the June 18-19 Fed meeting show that the Fed is considering allowing banks to use collateral, such as T-Bills for excess reserves. They are considering setting up a repo facility that essentially results in banks simply posting T-Bills instead of cash for excess reserves. The minutes reveal a number of pros and cons with the approach. It would be wise to have banks only post T-Bills for excess reserves above $20 billion.
There has been a lot of discussion regarding how much excess reserves is desirable given that, pre-crisis, excess reserves were small. Essentially there was only "required reserves” and banks with a little extra were lending it to those that wanted or needed a little more. The overnight rate charged between banks was kept in line with the Fed’s target fed funds rate by injecting or removing liquidity as necessary.
With the current large supply of excess reserves, the actual fed funds rate would plummet toward zero if the Fed was not propping up the rate by making excess reserves valuable by paying banks interest on those reserves. Since the financial system was awash with liquidity from QE, there was little need for lending between banks and the quoted fed funds rate stayed exactly the same as the rate being paid on excess reserves.
Recently, the fed funds rate has moved slightly higher than the rate paid by the Fed. The assumption can be made that this means that there are finally some banks that see growth opportunities and want or need additional reserves. And to borrow money from another bank, they would obviously need to pay a rate that is higher than what the Fed is paying. With there still being $1.4 trillion in excess reserves, it might seem a little surprising that banks are in any need for liquidity. However, excess reserves are concentrated among just a few large banks, forcing the rest of the banking industry to borrow from those banks.
One advantage of having the banks with reserves over $20 billion buy T-Bills to use as collateral for excess reserves is that it keeps the actual fed funds rate from popping above the target rate anytime any bank looks to borrow money from another banks since they are no longer competing against the rate the Fed is paying on excess reserves. The rate banks would need to borrow at would need to compete with (be higher than) the yield of T-Bills.
Other advantages of banks buying T-Bills with excess reserves cash is that it lowers the yield on T-Bills and encourages banks to earn higher yields by lending the money rather than parking cash at the Fed resulting in a stimulus for the economy. Lower T-Bill yields also helps to steepen the yield curve and reduce the (mistaken?) perception of a looming recession due to the inverted yield curve. Reducing the amount of excess reserves that the Fed pays interest on also essentially saves taxpayers a few billion dollars and reduces the optics of the Fed subsidizing the profits of large banks (i.e., reverse Robin Hood).
The $20 billion cut off for how much in excess reserves receives interest from the Fed versus how much of the reserves consists of T-Bills can be adjusted to keep the actual fed funds rate within the target range. But the market will likely self-correct some as well. Specifically, if the T-Bill rate falls very far below the interest paid on excess reserves, banks using T-Bills for collateral will likely significantly reduce excess reserves, which would push the actual fed fund rate higher. And, of course, the Fed would ask the handful of banks affected how they might adjust their excess reserve balances under this policy to confirm this most likely result. However, if the actual rate did stay below the target range for too long, the cut off level could simply be raised to as high as it needed to be. Indeed the policy could be implemented slowly beginning with a very high threshold that only moves a couple hundred billion into T-Bills initially.
Lastly, using a blended policy of paying interest on excess reserves and requiring banks to use T-Bills as collateral for a portion of excess reserves will help give the Fed further insight into the optimal level of reserves needed within the financial system by observing an actual fed funds rate that can move both above and below the interest paid on excess reserves.
