Operation Twist: The Next Tool of Unconventional Monetary Policy

The Federal Reserve has announced plans to implement Operation Twist over the next 10 months. During that time the Fed will use proceeds from sales of Treasury debt with less than three years to maturity to re-invest in longer maturity Treasury bonds. Additionally, to add a twist to its holdings of non-Treasury liabilities, the funds from maturing mortgage-backed securities will be invested in similar assets but of especially long duration.

The Fed's latest action can be interpreted as a logical extension of QE2. In the latter exercise there was explicit balance sheet expansion as additions to the monetary base were used to purchase Treasury debt. A widely recognized feature of QE interventions is that the financial and economic consequences depend not only on the quantity of reserves injected into the system, but also the kind of asset purchases that they are used to fund. In a world in which there was no segmentation of the yield curve and no commitment problems affected forward guidance for the policy interest rate, the kind of Treasury debt purchased would be of no consequence. However, theory and evidence suggest that both conditions fail.

Firstly, investors have preferred space on the yield curve, meaning that when the yield on an n-period asset is forced down through the central bank purchasing that asset at a reduced yield, there is only imperfect substitution to other assets along the curve so that the slope of the yield curve changes in addition to its height. By concentrating QE funds on long-dated debt the Fed can twist the yield curve. Accepting the view that declines in long rates relative to short rates are a form of stimulus given the importance of long rates for pricing US mortgages and corporate lending, Operation Twist gives extra bite to the stock of QE announced to date. This interpretation of Operation Twist is often described as reducing the term premium on long bonds. The Fed willingly accepts Treasuries with little compensation for the risk associated with future fluctuations in yields and, since the Fed is a big player  in the market, the market average valuation of risk declines so that the term/risk premium on long assets is forced down.

What is the role of QE and Operation Twist in relation to policy rate guidance and its impact on long rates? A conditional commitment to keep the Fed Funds rate at current levels for at least a couple of years can reduce long yields through either (i) reducing market expectations for short rates over the lifetime of long assets so that long yield fall via a simple Expectations Hypothesis of the Term Structure effect, or (ii) reducing the amount of uncertainty over market expectations of future short rates in a way that reduces the riskiness of long assets and shrinks the risk/term premium (matching the outcome from a reduced market valuation of risk arising from Fed actions and described above). However, both of these effects may be limited if guidance over future interest rates lacks credibility due to the interest rate path being at odds with central bank inflation targets -- markets attribute less than unit probability to future interest rates in line with the guidance because a future inflation problem (related or unrelated to monetary policy) will necessitate higher interest rates). How can QE and Operation Twist help to address this problem? Both may play a role as commitment technologies that help to increase the credibility of interest rate projections and ensure that they are reflected in current long rates. The logic here is as follows:

-- QE involves reserve injections beyond a critical level X necessary to drive the Fed Funds rate to a lower bound

-- so a pre-requisite for normalization of the Fed Funds rate is that reserve injections be reversed until reserves are forced below X

-- if the Fed then wishes to deviate from past Fed Funds guidance, it cannot do so immediately; it must wait for that length of time necessary to drain reserves below X

-- QE on a significant scale increases the time needed to eliminate excess reserves and therefore commits the Fed to a lower bound for the Fed Funds rate for a longer period, addressing the time inconsistency associated with interest rate projections

-- so long yields are determined based on a longer duration for short rates at the lower bound and less risk of a deviation from that level, i.e. long rates fall and monetary stimulus is imparted to the economy

-- what is the role for Operation Twist? QE is always subject to natural attrition if assets mature and the proceeds are not re-invested, and sufficient attrition could force reserves below the critical level such that the Fed Funds rate increases and deviates from previous interest rate guidance; but obviously the rate of attrition and the potential for a rising Fed Funds rate will be reduced if Operation Twist has been used to increase the maturity of the Fed's portfolio, so the above monetary stimulus is enhanced when QE is backed up by Operation Twist interventions.

An important assumption in this analysis is that the ability of the Fed to drain reserves and raise the Fed Funds rate at a rate in excess of that implied by natural attrition of the balance sheet is limited. Why might this be the case? An obvious option for draining reserves from the overnight market is reverse asset purchases. But dramatic reverse asset purchases may not be feasible in a short space of time, e.g. if the market price of Treasuries is low the Fed may wish to delay sales in order to allow a price recovery (if the Fed is averse to low bond prices then large and dramatic sales are to be avoided as they will force down the price).

Limits to the speed at which asset sales can be effected by the central bank have forced central banks to look at other options. One possibility now open to the Fed is that they use their new powers, conferred by the legislature, to pay interest on reserve deposits at the Fed. By offering a sufficiently high deposit rate, reserves in the market would be immobilized (the Fed converts a cash liability into a deposit liability). In theory, the draining of reserves from the market necessary to normalize the Fed Funds rate could occur via this mechanism, in which case QE and Operation Twist would not serve as effective commitment technologies. However, there are limits to how high the deposit rate on reserves can be set and therefore to the extent that it can be used to facilitate upward adjustment of the Fed Funds rate. Firstly, paying interest on reserves is a fiscal action (a point made by Goodfriend, JME 2011). As the Fed is not well placed to make a return on deposit liabilities (purchasing assets or making loans would push the reserves back to the market) it must rely on a fiscal transfer to cover the interest cost. Even if politicians are willing to provide this support, it may not be immediately forthcoming, or forthcoming on the scale needed to fund a deposit rate high enough to drain excess reserves from the market and allow the Fed Funds rate to rise. This is especially likely to be the case given that many politicians are opposed to paying interest on reserves -- the Congress already passed the TARP to bail-out financial institutions and is unlikely to want to pay those institutions a risk free return on the liquidity they have received in exchange for distressed assets.

If these constraints are sufficient to ensure that the Fed cannot effect rapid elimination of excess liquidity on any scale required, then QE and Operation Twist will help to address any time inconsistency associated with interest rate projections, and will in turn promote maximum monetary stimulus for economic recovery.