This commentary is jointly written by Bob Eisenbeis and David Kotok; their bios may be found at www.cumber.com.

Note to Readers:  This is the first of a two-part commentary motivated by speeches and editorials from Federal Reserve officials about possible exit strategies from its current quantitative easing policies.  We comment on some problems that the strategies may pose.  We also identify subsidies in the Fed’s current policies.  In part two we comment on the Fed’s own operating policies that may have played an important role in creating the too-big-to-fail problem.  This last issue was overlooked by the Dallas Fed’s Fisher and Rosenblum in their WSJ op-ed piece of September 27, 2009.  They lamented the bottleneck that the concentration of banking resources now creates as the Fed attempts to exit its QE strategy.  They fail to mention how the Fed’s determination of primary-dealer status has contributed to the problem

It is becoming increasingly clear from recent information coming from the Fed that its exit strategy from the crisis-induced injection of liquidity will rely upon two mechanisms.  First, the Fed will try to sop up excess reserves by engaging in reverse repurchase (RP) agreements using accumulated mortgage and Treasury debt.  Second, the Fed will attempt to adjust interest rates upward and perhaps sharply, as Governor Warsh recently suggested. 

The Fed will attempt to sterilize the excess reserves that it has created and that have accumulated by raising the interest rate it pays on such funds that are placed with the Fed by banks.  The Fed could also raise reserve requirements, although there is no indication they will pursue the reserve requirement course. 

If the Fed follows the mechanism that is now used by many other central banks, the rate paid on excess reserves will set a floor, the discount rate will set a ceiling, and the targeted Federal Funds (FF) rate will be in the middle.  The actual transaction-driven effective FF rate will fluctuate within that floor-ceiling corridor.  At least that is the theoretical construction.

Presently, an apparent anomaly exists in the FF market.  The interest rate the Fed is paying on excess reserves is 25 basis points, the desired target for the FF rate is defined to be a range of between 0 and 25 basis points (less than or equal to the excess reserve rate), and the discount rate is set at 50 basis points.  The effective FF rate is trading roughly in the range of 14 to 16 basis points.  The 25 basis points the Fed is paying on excess reserves, while consistent with the target, sets up a riskless arbitrage possibility for banks.  They can borrow at 14-16 basis points in the FF market and immediately lend to the Fed at 25 basis points – they make 9-11 basis points risk-free.

Part of the reason the effective FF rate is below the upper range of the target is explained by the actions of Freddie and Fannie, who accumulate large volumes of cash payments from mortgages until required disbursements are made on mortgage-backed securities and they must deploy those funds on a short-term basis.  The GSEs can’t hold deposits at the Fed or earn the interest on excess reserves that the banks are able to earn. 

Because GSEs are not banks, they are faced with either earning a zero return on those funds by simply sitting on idle balances or they can lend the funds in the FF market, which they are doing at rates below what the Fed is paying on excess reserves.  This arbitrage is a direct subsidy from the government (Freddie and Fannie are now under government conservatorship) and from the Federal Reserve to the banks, because it enables them to improve earnings and build their capital. 

As long as the Fed wants to subsidize the banks, it will be difficult for it to raise interest rates with the FF rate trading below the rate paid on reserves.  This pattern is particularly likely as long as Freddie and Fannie continue to pump funds into the market. 

Interestingly, Sweden’s Riksbank has actually been charging a negative rate on excess reserve deposits held with it, in an effort to induce lending.  But with so much liquidity in the market, this isn’t a feasible strategy for the Fed, since it would stimulate a rapid increase in the money supply and risk inflation.  For a discussion of Riksbank’s policy see: “Interest Rates go Negative: Compare Riksbank (Sweden) with our Federal Reserve” http://www.cumber.com/commentary.aspx?file=070909.asp .

As for the reverse RPs, some reports suggest that the Fed is exploring the establishment of relationships with money-market mutual funds (MMFs) to engage in reverse repos.  The Fed would essentially sell securities overnight or perhaps on a term basis, to the MMFs, with an agreement to buy them back at a fixed price.  The MMF pays cash, which has the intended effect of draining a portion of the excess reserves that had been pumped into the banking system.  Those reserves would then be unavailable for credit expansion should the economy gain speed. 

From the MMFs’ perspective, this would be a risk-free transaction that would use the Fed’s long-term securities as collateral and create a short-term asset for the MMFs, consistent with their investment strategy.  This would be the equivalent, from a rate perspective, of a short-term Treasury, and superior to short-term commercial paper. 

There are several implications of this strategy for interest rates and markets.  First, given the large volume of securities that are involved, nearly all short-term interest rates would rise.  This includes the rates on Fed Funds, Treasuries, and short-term commercial paper.  Unless the Fed is extremely adept in coordination of this strategy within its interest-rate corridor strategy, the effective FF rate could push against and even exceed the discount rate (the supposed upper bound on interest rates).  If this occurs, another risk-free arbitrage situation will be created and the discount-rate policy will have to be changed.  The FOMC will find itself chasing its reverse RP policy as it sets the target Fed Funds rate.  This “chasing the market” problem will likely be heightened the moment the Fed changes the language in its policy statement.  Right now that statement says that rates will remain low for an “extended period.”

Rational investors will assume, especially now that Governor Warsh has raised the possibility (although Governor Kohn expressed another view on Sept. 30th), that the FOMC might raise rates rapidly.  That would be a departure from the Greenspan gradualism that characterized the last period of FOMC tightening. 

Eventually, rates will begin to rise from their present near-zero level to what is expected to be an equilibrium rate.  There is no present way to determine what that new equilibrium rate will be.  And the start of a rate-hike sequence will be without regard to the timing or the Fed’s intended path for the FF rate.  The point is that it is impossible to pursue a reverse RP liquidity mop-up strategy easily, without also affecting many other interest rates.  Markets know that the rate movements might be quite large. 

Secondly, there is also a potential important consequence for the commercial paper market.  If the reverse RPs become more attractive investments to MMFs than commercial paper, because of their zero risk and guaranteed liquidity, then the supply of short-term money to corporations may dry up.  This would likely occur just as the business expansion takes off, unless the Fed rejuvenates its special commercial paper facility.  But of course, the objective in a tightening regime is to reduce liquidity, not increase it. 

The third implication concerns the structure of the markets for FOMC day-to-day policy actions.  The idea of engaging in reverse RPs with MMFs is rooted in the fact that the current group of primary dealers is capital-constrained.  They have been contracting their balance sheets for almost two years.  They have limited capacity to invest in short-term reverse RPs and may not have enough to meet the Fed’s policy needs. 

This problem of the primary dealers is the subject of part two of this commentary.  The issue has, in large part, been created by the NY Fed’s Open Market Desk operating procedures and the Federal Reserve’s rescue policies for large, complex financial institutions.  We note that these facts were overlooked in the recent op-ed by the Dallas Fed’s Fisher and Rosenblum, as they tussled with the problem of “too big to fail.”  We will discuss this in part 2.