… the Fed continues to quietly sterilise away.
Case in point, the Fed’s continuing 28-day term deposit facilities.
Readers may recall, these were originally introduced over a year ago and billed at the time as a type of “exit strategy” testing exercise, alongside reverse repos and outright asset fails (the latter arguably turning into Operation Twist).
But the concept of the term deposit facilitywas originally aired as far back as December 2009.
As Reuters noted back then:
Dec 28 (Reuters) – The U.S. Federal Reserve on Monday proposed the creation of a new mechanism, a “term deposit facility,” to help the central bank’s policy-makers withdraw money from the banking system when they decide to tighten monetary policy. The program would allow financial institutions to earn interest on loans of longer maturities to the central bank. The Fed already pays interest on banks’ overnight reserves. The Fed said it would like to offer term deposits at auction.
If that strikes you as very similar to the ECB’s term deposit facilities — which have always been communicated as the ECB’s main sterilisation mechanism for asset purchases — that’s because it is.
The only difference is that the quantities being offered by the Fed are much smaller — $5bn to be exact.
But from the Fed’s point of view, that’s all it really needs — the facility’s main objective is, after all, to mop up ‘high powered money’ which has not been absorbed by its interest on excess reserve (IOER) facility — which, contrary to popular understanding, is actually a sterilisation mechanism in its own right.
The ECB offers its own version in the form of an overnight deposit facility which currently pays out 50 basis points.
This point could not have been better explained by Marvin Goodfriend of Carnegie Mellon University, in an ECB paper here:
Pure monetary policy consists of open market operations that expand or contract bank reserves or currency by buying or selling Treasury securities. Pure monetary policy works by varying the scarcity of bank reserves to manage the spread between the interbank interest rate and interest paid on reserves, whether or not interest is paid on (excess) reserves. In the United States, the Federal Reserve chooses the scarcity of bank reserves to manage the spread between the federal funds rate and interest on reserves; in the euro area, the Eurosystem chooses the scarcity of reserves to manage the spread between EONIA and the deposit rate.
And it’s the issue of scarcity that’s key.
Central banks depend on there being a) desire for government bonds and b) the ability to create a scarcity in government bond markets (when it suits them) in order to control monetary policy effectively.
Buying bonds = more scarcity = ‘risk free’ rates go down (you are prepared to overpay to obtain a risk-free asset).
Selling bonds = less scarcity = ‘risk free’ rates go up (you have to be comepensated for holding a risk-free asset).
That’s easy enough in a central bank environment which focuses on only one type of quality collateral. But in a market where there are multiple credit options, central banks can lose control if they treat all bonds as equal even if the market doesn’t.
First, there is the risk that quality bonds “go special” — become so scarce that no matter how much the central bank offers on deposit it can’t stop rates turning negative since participants simply can’t get hold of enough collateral to create the reserves in the first place.
Second, there is the risk of lesser quality bonds becoming so abundant that the market depends on the central bank to offer deposits just to stop rates spiking.
If the divide between the lesser quality bonds and the good quality bonds gets too wide, the central bank can lose complete control of the spread between its key market rate (the Eonia rate in the case of the ECB, and Fed funds in the case of the Fed) and the deposit rate (or interest on reserve) it uses to place a floor on rates.
But rather than the rate zooming higher (as in a credit crunch) this time the real problem will come if the rate shoots lower — as has been happening in the US.
As we’ve discussed, in Europe (until recently) scarcity of quality collateral was never an issue because the ECB was happy to accept what Willem Buiter affectionately termed “rubbish” assets for liquidity operations.
If anything, the ECB had to work much harder at sterilising asset purchases with larger amounts of fixed-term deposits than the Fed (purchases which were mostly designed more to maintain the illusion that all the collateral was equal).
But that has now led to a situation where only the best collateral (core eurozone bonds) will be used comfortably by banks for interbank collateralised borrowing arrangements. And every time a previously acceptable market becomes a no-go area — that only goes to limit the acceptable collateral pool all the more.
In that circumstance the ECB’s term deposits become increasingly ineffective at propping up real market rates as reflected by the interbank core collateralised rates.
So while overnight deposit sums are rising on the surface of things (because banks are stuffing as much spare liquidity sourced by disposing of low quality collateral into the safe interest-dispensing hands of the ECB)…
… the overall amount of spare liquidity in the market is dying:
You can think of the first chart as a reflection of the number of banks trying to capture the last remaining free arbitrage in town. Which is understandable, given as Goodfriend explains, this was part of the point of the ECB’s strategy in the first place:
At the peak of the credit turmoil the Eurosystem announced refinancing operations in which it would accommodate whatever funding banks asked at various fixed interest rates and terms up to one full year. For instance, by offering banks unlimited credit for one year at 1% together with a 0.25% rate at the deposit facility, the Eurosystem set the “term spread” between liquid reserves and one-year credit. A bank could acquire whatever euro reserves it demanded for a year at a net pecuniary cost of 0.75%. The Eurosystem presented banks with an arbitrage opportunity – subject to meeting collateral requirements, banks could borrow reserves until they drove their marginal implicit liquidity convenience yield on reserves down to 0.75%.
Panel B of Chart 6 in the paper by Cassola, Durre and Holthausen shows that EONIA was pressed nearly down to the deposit rate (interest on reserves) floor throughout the first period of long-term, fixed-rate, full allotment refinancing operations. The perfectly elastic demand for reserves was fully accommodated at just above the 0.25% deposit rate floor during the period. In other words, banks took advantage of the arbitrage opportunity offered by the Eurosystem to satiate themselves with reserves.
Which brings us back to the subject of sterilisation and the scope for “real QE” in the Eurozone.
On that matter, our point is simple. “Real QE” is arguably something we’ve already seen, while the debate over sterilisation is nothing more than a distraction from the real point.
The question shouldn’t be whether the ECB stops sterilising asset purchases or not.
Rather, we should be asking if it’s correct for the ECB to continue asset switching at all. After all, by removing the risk of holding lesser quality collateral (to maturity) by giving banks high-powered money to keep at the ECB’s deposit facility at a perpetual “risk-free rate” instead, the ECB is effectively de-risking the market rather than helping to solve the problem (which is, of course, that the collateral is bad and always has been). It’s the asset switching, however, which is exacerbating the two-tier market. It’s the asset-switching which is killing price discovery. And it’s also the prospect of new asset-switching ops that’s causing mayhem, uncertainty and volatility.
And to what end? Mainly for the purpose of creating a toxic dumpster of dark inventory that can be held back from the market on the mistaken hope that one day the market will realise this is only temporary glitch in the matrix rather than a genuine re-pricing of credit.
Sterlisation or no sterilisation won’t change that.
Meanwhile, in the event the ECB eventually runs out of lesser quality bonds to soak up, the demand for good collateral (presumably German bonds) would be such, only German repo rates would truly be indicative of borrowing costs in Europe.
And that would be very much situation Swiss (a.k.a at a negative rate).
It would also render the ECB pretty much useless in monetary policy– since the bank would no longer have a float of desirable collateral the market gave a damn about, scarce or not scarce.
Arguably the answer is to stop asset-switching now, and to let the market re-price credit logically.
After all the problem with subprime was never the subprime tranch themselves, it was the stuff that was disguised as triple-A — when it wasn’t.
Goodbye risk-free, hello French spreads at 21-year high – FT Alphaville
Coverage of eurozone debt crisis – FT Alphaville
When a government bond becomes a Giffen good – FT Alphaville
Sterilising Silvio – FT Alphaville