Forget Tapering and Exit via Reserve Requirements
July 2, 2013
The market turmoil that followed the last FOMC meeting suggests it may be time for a primer on the relationships among the Federal Reserve’s asset purchase program, its Federal Funds rate target policy, and its unconventional policy at the zero bound. What are these programs and how do they fit together? It turns out that they are more closely intertwined than the discussions so far have suggested.
First, here are some basics. Classical supply and demand theory tells us that in a competitive market place policy makers can fix either a price or a quantity, but not both. Fixing a market’s price means the quantities supplied and demanded will vary over time. Conversely, fixing the supply of a good in a market means that the market-clearing price will not be controlled.
Fed policy to set the Federal Funds rate is designed to fix the price (interest rate) at which excess reserve deposits held at the Fed trade in the wholesale money market. This means that the Fed must purchase or sell assets, or engage in short term repurchase and reverse repurchase transactions, in order to add or subtract whatever reserves are needed to keep the short term Fed Funds rate within the FOMC’s target range of 0–.25%. There is one minor technicality here: the Fed’s price-setting mechanism isn’t perfect because it doesn’t stand ready either to buy or to sell reserves at its target rate throughout the day. But we should not let that small technical caveat distract us from the main points to be made.
Now, what about the Fed’s $85 billion/month asset-purchase program that is supposedly an unconventional policy tool? In one sense it is, and in another sense it isn’t. When the Fed purchases long-term Treasuries or MBS in the marketplace, it pays for those assets by creating high-powered money in the form of deposits at Federal Reserve Banks. Therefore, as is true of the implementation of its Federal Funds target policy, the Fed is simply adding to the quantity of bank reserves. What is different is that the addition of reserves is not associated with its short-term Federal Funds target, but is instead focused on asset purchases of longer maturities in the Treasury and MBS markets. These purchases are continually adding reserves to the banking system and not adjusting them up or down on a day-to-day basis the way the Fed does in the overnight Federal Funds market.
To date, most of the Fed’s purchases have been almost exclusively in over-five-year maturities. By taking longer-maturity Treasuries and MBS out of the private-sector market, the prices of those securities have been increased, placing downward pressure on their interest rates. As distinct from the Fed’s specific target range for the short-term Federal Funds rate, the asset purchase program doesn’t have a set target schedule of rates for the assets the Fed is buying; it is just attempting to put downward pressure on those rates and let them settle down at the then-new market-clearing rate. Considering the two policies together, the Fed is attempting to anchor the term structure near zero on the short end while forcing longer-term rates lower. The result is a highly distorted term structure at the moment.
The Fed’s asset purchase program is simply a flow component of the broader strategy it initiated much earlier in the financial crisis, when the Fed quickly built up its asset portfolio in an attempt to flood the market with liquidity and to stimulate lending. Notice, however, that the three strategies have a common feature in that they all involve the injection of high-powered money in the form of excess bank reserves. These excess reserves are the same funds that can exchange hands in the wholesale Fed Funds market. So what the Fed is doing through its asset purchase program is converting longer-term assets held by the private sector – Treasuries and MBS – into bank reserves that are short-term payable upon demand.
This raises the following question: How can the Federal Reserve System Open Market Desk simultaneously adjust the volume of Federal Funds traded in the overnight money market to keep the target rate between 0 and .25% while at the same time continuously adding to excess reserve balances that could be traded in the wholesale money market generated through its asset purchase program? Shouldn’t this continued flow of Fed Funds push the rate to zero? This is where the payment of interest on reserves comes in.
Virtually none of the excess reserves available are actually traded in the overnight market, because they are effectively being sterilized by the fact that the Fed is currently paying interest on reserves at 25 basis points. That rate exceeds, by at least 15 basis points, the 10 basis points or so that a bank could earn by lending those funds in the overnight Federal Funds market.
So where do the funds come from that are trading in the overnight market? We do know that two important sources are Freddie Mac and Fannie Mae. These institutions accumulate payments on the mortgages that are behind the MBS they have issued, and those payments temporarily are warehoused until they are disbursed each month to the holders of those MBS. Since Freddie and Fannie are not banks, they are not permitted to hold deposits at the Fed and cannot earn the 25 basis points available to banks. Instead, they lend those funds out in the overnight market at rates below the 25 basis points paid on bank reserves. It is far better for them to earn 10 basis points by lending accumulated deposits into the overnight market than to let those deposits sit idle earning zero.
Who might the borrowers be? The answer is that any commercial bank would gladly borrow at 10 basis points in the overnight market and earn a 15 basis-point spread by holding reserves paying 25 basis points at the Fed. What we are observing is a risk-free arbitrage to the borrowing institutions and a wealth transfer from the Fed – and ultimately from US taxpayers – to the borrowing banks.
The Fed seems convinced that its asset purchase program of $85 billion in injection of high-powered money is largely imperceptible at this point, in terms of its effects on interest rates. Hence, in the press conference following the last FOMC meeting it saw little risk in beginning to talk about scaling back the program. And its messages suggest that the Committee envisioned that a phased tapering in the program could proceed smoothly, with little market disruption. But this view ignores the interrelationship between asset purchases, bank reserves, the near-zero Fed Funds target, interest on reserves, and the risks facing bond managers who have a fiduciary responsibility to protect their clients’ net worth.
Simply scaling back on the program in phases will start to ease upward pressure on asset prices as the largest marginal participant – the Federal Reserve – begins to exit the market. But the market doesn’t know where on the term structure the effects will be most important, since these will depend on where the cutbacks occur relative to Treasury issuance of new securities. This lack of information and the uncertainty it creates is what was so disruptive to markets. Federal Reserve Bank of Richmond President Jeffrey Lacker put it correctly last week when he argued that volatility and uncertainty will be with markets for some time due to this possible change in Fed purchases.
Extrapolating the exit problem forward to the time when the Fed will begin asset sales, the problems become even more difficult. Selling assets will clearly increase the supply on the market, lower asset prices, and raise yields. And how attractive those assets will be to potential buyers, who are mainly the holders of excess reserve balances, will depend upon the prices and maturities offered and on the attractiveness of those prices relative to the risk-free return of earning 25 basis points on reserves.
How sales will be distributed across the term structure will radically affect both its shape and any price risks that potential buyers will face. Clearly any rational buyer will require a significant price risk premium, especially at the start of the program, to compensate for the eventual continual flow of asset sales and hence the serial decline in prices of any accumulated holdings of Treasuries. A decline in the price of any security when it is sold will also reduce the market value of any similar outstanding securities, whether they are sold or not. The Fed could couple asset sales with a change in the interest rate it offers on reserves, but it isn’t clear that increasing the opportunity cost of holding excess reserves, for example, by reducing the rate paid on reserves in order to make Treasury purchases more attractive will be sufficient to induce institutions and buyers to lower the price risk premia they will demand.
Finally, we haven’t talked about how private-sector holders of outstanding Treasuries will respond to potential asset sales by the Fed. Clearly, the selloff that occurred last week at even the mere hint of a change in any part of the Fed’s policy is a clue and should send a powerful signal to policy makers that their assumption that a smooth path exists is questionable.
So what might the Fed do? Instead of spooking fixed-income markets with uncertain proposals for asset sales, they might opt for an obvious alternative, which we have put forth before. Raising reserve requirements to idle at least 90% or some other very high percentage of the excess reserves would eliminate the need for the Fed to sell assets. The Fed could couple that increase with a commitment (forward guidance) to lower reserve requirements as bank lending increases and the economy grows, but in a phased way that provides flexibility and accommodation to lending institutions, even as it eliminates the inflation risks associated with an unanticipated increase in the money supply and acceleration in bank lending and aggregate demand that increases the Fed’s inflation forecasts. And doing so even before tapering its asset purchase program would send a strong message about what the exit path would look like. The increase in reserve requirements would be a nonevent for the affected institutions, since the reserves are not being used to support lending, and banks would continue to earn a 25-basis-point return on their total reserve positions. The Fed might even commit to increase that payment in line with movements in market rates.
Critics might argue that an across-the-board increase in reserve requirements would unfairly impact smaller community banks, but the Fed has the legal authority to impose differential reserve requirements and has done so in the past. It could tailor the reserve requirement structure to avoid unduly constraining certain classes of institutions. Finally, by going the reserve requirement route, the Fed would buy the needed time to allow its existing asset holdings to run off, thereby eliminating the threat of asset sales and undue price risks to debt markets. It could then gradually return to its interest-rate targeting regime.
Bob Eisenbeis, Vice Chairman & Chief Monetary Economist
Bob Eisenbeis is Cumberland’s Vice Chairman & Chief Monetary Economist. Prior to joining Cumberland Advisors he was the Executive Vice President and Director of Research at the Federal Reserve Bank of Atlanta. Bob is presently a member of the U.S. Shadow Financial Regulatory Committee and the Financial Economist Roundtable. His bio is found at www.cumber.com. He may be reached at Bob.Eisenbeis-at-cumber-dot-com.
Please use the comments to demonstrate your own ignorance, unfamiliarity with empirical data and lack of respect for scientific knowledge. Be sure to create straw men and argue against things I have neither said nor implied. If you could repeat previously discredited memes or steer the conversation into irrelevant, off topic discussions, it would be appreciated. Lastly, kindly forgo all civility in your discourse . . . you are, after all, anonymous.