How long does a long-term refinancing operation (LTRO) have to be to become quantitative easing (QE)?
By Stephen Lewis, Monument Securities
LONDON (Dow Jones)--How long does a long-term refinancing operation (LTRO)
have to be to become quantitative easing (QE)? This is a question for all those
market commentators who propagate the notion that the ECB was engaging in QE by
the backdoor when it conducted its three-year LTRO yesterday.
After all, this was not the first time the ECB had resorted to LTROs.
In May 2009 the ECB's Governing Council decided to provide liquidity through a
one-year LTRO. This action was not hailed then as the ECB's version of QE,
though QE policies were highly topical since both the US Federal Reserve and
the Bank of England had recently introduced them with much fanfare. Perhaps we are expected to believe
that because this week's operation was for a three-year term, rather than for
one year only, it was more like QE than the ECB's previous liquidity-supplying
actions. Where, then, is the dividing-line to be drawn between LTROs
that are not QE and those that are? Or could it be that when
the ECB's Anglo-Saxon critics spent years castigating it for its failure to
adopt QE, it had all along been conducting what was, in effect, QE? Whatever
the answers, it
can hardly represent a game-changing revolution in ECB thinking
that it has extended the term over which it is prepared to undertake
refinancing operations.
This discussion, in fact, prompts a deeper question regarding the
appropriate criteria by which to judge whether a central bank is engaging in QE. The phrase 'quantitative easing' arose in
recognition of the direct effects that measures falling under this heading have
on the quantity of money, as opposed to the cost of credit. For a central bank action to
constitute QE, there ought, therefore, to be an observable impact on a monetary
aggregate. Unfortunately, in practice, it is well-nigh impossible to distinguish
the effects on money supply of the central bank's QE from those of a multitude
of other influences affecting the outcome, even in those countries where central banks are avowedly pursuing
a QE policy. In any case, the baseline, that is, what would have happened in
the absence of the central bank action, is purely hypothetical. Because of the
practical difficulty in applying a results-based test for QE, central banks have taken
to using the impact of their actions on their own balance sheets as the
criterion for determining whether those actions constitute QE. Thus, the Bank of England refers to
its QE as 'the programme of asset purchases financed by the issuance of central
bank reserves'. If a rise in the central bank's reserves is the result of its
operations, then those operations are QE. For the markets, however, QE has come to mean
central bank buying of government bonds. This may well be because
such purchases have figured prominently in the QE of the Fed and the Bank of
England. But a central bank's buying of government bonds is not necessarily
financed by an increase in its reserves. In the Fed's current 'Operation Twist', for example,
the central bank is financing its purchases of long-dated government securities
by selling some of its holdings of shorter-dated government notes and bonds.
These transactions are not regarded as QE because it is obvious that the Fed's operations are not aiming to
increase the aggregate volume of US Treasuries it holds or the size of its
balance sheet. However, the original 'Operation Twist' in the 1960s also
failed the QE test, though less obviously because the Fed was
then financing its purchases of longer-dated government securities by selling
down its holdings of Treasury bills.
The ECB
will not be financing its latest LTRO by selling any of its recently-acquired
government bonds but there could yet be substantial shifts in
the composition of its liabilities, without any increase in their volume, to
accommodate the extension of liquidity through the LTRO. The maturity of a
€141.9bn short-term loan from the ECB to lenders will offset part of the
increase in reserves that would otherwise stem from the three-year operation.
But that is unlikely to be the only consequential adjustment in the ECB's
balance sheet. Until the full picture emerges with publication of the ECB's next weekly
report, it is hard to judge what the balance sheet impact of the LTRO might have
been.
Whether the banks which have entered into LTRO arrangements with the ECB
take advantage of their enhanced liquidity to buy government bonds or not will
have no bearing on the effect the LTRO has on the ECB's liabilities. Further, Mr Draghi has never
endorsed the view that the reason why the ECB decided to conduct the LTRO was
to provide banks with funds to purchase government bonds. The notion that the ECB had been
trying to deliver support to peripheral government bond markets, via the
commercial banks, gained credence largely from comments from Mr Noyer, of the
Bank of France, in the weekend following the EU summit. That was a critical
juncture for French policymakers, struggling to defend France's 'triple-A'
credit rating. Market judgment on the summit's outcome was in the balance. The
governance reforms had been positively received but the meeting had been thin
on proposals to stabilise euro zone markets in the near term. In such
circumstances, it is understandable that Mr Noyer should have been eager to
emphasise the possibilities opened up by the ECB's 8 December decisions aimed
at strengthening banks' liquidity.
Certainly, the financial markets were disposed to believe the euro zone
authorities were delivering what investors most wanted, support for the
peripheral government bond markets. So the story evolved that the
ECB had dramatically changed tack by embracing QE and that this
would unleash substantial bank buying of hitherto shunned government bonds. In fact, what the banks
will do with the liquidity they have gained through the LTRO is far from clear. Mr Nicastro of Unicredit today
declared his bank would boost lending to industry and households. The ECB will
be pleased to hear that but, recently, the correlation has been low between
what banks say they are doing and what shows up in the lending data.
