The ECB’s second three-year longer-term refinancing operation (LTRO) is similar in gross terms to the first but will have a significantly larger impact on net lending to the banking system and bank reserves, suggesting further liquidity support for markets.

The €529.5 billion lent in today’s three-year LTRO compares with €489.2 billion in the first such operation conducted on 21 December. Banks, however, repaid €275.1 billion of other “monetary policy” loans in the week of the December operation, resulting in a net increase of €214.1 billion.

This week, banks have repaid €218.9 billion of shorter-term loans, implying a net rise in monetary policy lending of €310.6 billion. Some banks previously borrowing under “emergency liquidity assistance” programmes, however, probably took advantage of looser collateral rules to switch into the cheaper three-year LTRO. The net increase in lending, therefore, should be less than €310.6 billion but still significantly higher than in December.

The rise of €214.1 billion in monetary policy lending in the week of the December operation was reflected in a €164.6 billion increase in banks’ current account and deposit facility balances at the ECB. Assuming the same percentage pass-through this week, these reserves may grow by about €240 billion from their level of €571 billion at the end of last week, a 42% gain. Such a rise would push G7 bank reserves well above their December record – see chart.

A provisional verdict based on today’s money supply numbers for January is that the Draghi bazooka has succeeded in stabilising bank balance sheets but has yet to lay the monetary foundation for an economic recovery. Peripheral monetary indicators, indeed, have deteriorated further, suggesting a faster contraction of economic activity.

The stabilising effect of the bazooka is evidenced by a €68 billion, or 0.7%, rise in broad money M3 in January, reversing falls in November and December. This reflected the sovereign carry trade: banks bought €52.3 billion of Eurozone government bonds, the most since October 2010 and up from €2.7 billion and €3.7 billion respectively in December and November – see first chart. (These numbers exclude the ECB’s purchases under the securities markets programme and are not seasonally adjusted.) As expected, the biggest buyers were Italian and Spanish banks, whose holdings rose by €28.4 billion and €24.4 billion respectively, presumably reflecting purchases of local debt.

The M3 rebound, however, was not mirrored by narrow money M1, which rose by only 0.1% last month following a 0.2% December decline. M1 is a better economic leading indicator than M3, probably because consumers and firms shift money into more liquid forms before increasing spending. The six-month rate of change of real M1 slipped further to -0.5% (not annualised) in January from -0.3% in December – second chart. This measure had turned positive between August and November last year, signalling less negative economic news in early 2012, allowing for the usual six-month lag; the recent decline implies deteriorating prospects for mid 2012.

M1 comprises currency in circulation and overnight deposits. The ECB publishes a country breakdown of deposits but not currency. A 1.0% (not annualised) fall in Eurozone real M1 deposits in the six months to January conceals a solid 2.2% rise in “core” economies (defined here as Austria, Belgium, France, Germany, Luxembourg and the Netherlands) offset by a 6.1% plunge in the “periphery” (i.e. Greece, Ireland, Italy, Portugal and Spain) – third chart. The latter represents a new low, suggesting that the rate of peripheral GDP contraction will accelerate in mid 2012.

The peripheral country decomposition, as expected, shows the largest decline in Greece (-12.9%, not annualised) but Ireland is now as weak as Portugal (-9.2% versus -9.0%), with Italy not far behind (-8.0%); Spain’s contraction, oddly, is milder (-1.5%) – fourth chart. In the former cases, these are depression-scale declines, although normal monetary relationships may be breaking down in the context of rising EMU dissolution risk.

The Office for National Statistics (ONS) left its estimate of the change in GDP in the fourth quarter unrevised at -0.2% in today’s second release. The suggestion that the economy contracted is at odds with an above 50 reading of the purchasing managers’ composite output index as well as resilient labour market indicators and could be explained by a larger-than-usual fall in working days between the third and fourth quarters – see previous post. (The ONS does not apply a working day adjustment to quarterly GDP, although some of the monthly input variables are corrected.)

The apparent conflict between the GDP number and PMI / labour market data is reminiscent of the third quarter of 2009 – the quarterly GDP change was initially estimated at -0.4% but has subsequently been revised to +0.2%.

A monthly GDP estimate calculated from data on services, industrial and construction output (99% of the economy) was 0.05% above its fourth-quarter average in December, implying marginally positive carry-over into 2012 – see chart. The services sector ended the year solidly, with December output 0.3% above the fourth-quarter average and “only” 1.7% below a peak reached in November 2007.

The expenditure breakdown reveals a “surprise” 0.5% rise in household consumption last quarter after a 0.1% third-quarter decline. This confirms, belatedly, an assessment here last May that consumer prospects were improving, in contrast to media and MPC gloom at the time. The recovery appears to have continued in early 2012, with January retail sales volume 1.5% above the fourth-quarter average.

GDP weakness reflected a 5.6% fall in business fixed investment and lower stockbuilding. Investment figures are often revised heavily and the scale of the decline is difficult to square with a 1.7% rise in import volumes of capital goods last quarter.

The view here remains that the economy is pulling out of a “soft patch” that has been exaggerated by GDP data and media reporting, with a recession unlikely barring a negative external shock.

The Bank of Japan’s 14 February decision to expand its asset purchase program (APP) by ¥10 trillion to ¥65 trillion appears to represent a serious effort to loosen monetary conditions and may contribute to a further weakening of the yen over coming months.

The APP was introduced in October 2010 and stood at ¥44.5 trillion on 20 February. The aim is to reach the new ¥65 trillion ceiling by the end of 2012, suggesting a monetary boost of about ¥20 trillion over the next 10 months, equivalent to $250 billion or 4.25% of Japanese annual GDP.

Sceptics, however, argue that the BoJ is, once again, using an expansion of the headline APP target to deflect political pressure and is unlikely to follow through with a significant liquidity injection. They point out that the rise in the APP to ¥44.5 trillion since October 2010 has not been fully reflected in the size of the BoJ’s balance sheet and the monetary base (i.e. notes in circulation plus bank reserves) – BoJ assets have increased by ¥23 trillion and the base by only ¥16.5 trillion.

Less than 40% of the APP “injection”, in other words, has fed through to the monetary base. On this basis, the planned ¥20 trillion expansion by the end of 2012 may boost the base by only about ¥8 trillion.

To understand why this is probably too pessimistic, it is necessary to examine the reasons for the limited impact to date. The current ¥44.5 trillion APP total comprises “fund-supplying operations against pooled collateral” of ¥33.5 trillion and securities holdings of ¥11 trillion. Banks have used a large proportion of the funds accessed under the former facility to repay existing borrowing from the BoJ. Total BoJ loans have risen by only ¥11 trillion since October 2010, explaining the slippage between the APP and balance sheet expansion.

The even smaller impact on the monetary base reflects, in addition, an increase of ¥8 trillion in BoJ repo borrowing, which drains reserves from the banking system.

Looking ahead, the planned ¥20 trillion APP expansion by the end of 2012 is intended to be achieved mainly via a ¥19 trillion expansion of securities holdings, implying an equivalent boost to the size of the balance sheet. The BoJ’s new commitment, meanwhile, to target inflation at 1% over the medium to long term would seem to preclude sterilising this liquidity injection by increasing repo borrowing further. The APP rise, therefore, should be broadly matched by the monetary base.

The chart shows straight-line projections for BoJ assets and bank reserves assuming that 1) the planned increase in securities holdings is achieved, 2) there are no other influences on the size of the central bank’s balance sheet or the monetary base and 3) notes in circulation are stable so the rise in the base is reflected in reserves. The suggestion is that BoJ assets will expand to about ¥159 trillion by end-2012 versus a previous record of ¥150 trillion reached after the March 2011 earthquake / tsunami, with bank reserves rising by nearly three-quarters from their current level.

The Markit Chinese manufacturing purchasing managers’ survey seems to be less reliable than its official counterpart but today’s flash results for February are consistent with the view here that downside economic risks have increased, warranting policy easing beyond the 0.5 percentage point cut in reserve requirements announced over the weekend.

The key new orders index was flat at 49.1 in February despite a positive impact from an earlier-than-normal New Year holiday, while the finished goods inventories index rose – a probable dampener on production plans.

The official survey, released next week, deserves more weight and may confirm weakness, judging from a leading indicator derived from the OECD’s Chinese leading index – see chart.

The Eurozone flash PMIs also released this morning did not, as suggested in a post yesterday, surprise positively but the manufacturing new orders and services new business indices continued to recover, reaching their highest levels for seven and six months respectively – the message from the earnings revisions ratio was directionally correct.

As previously discussed, the key purchasing managers’ manufacturing new orders indices correlate well with “revisions ratios” of equity analysts – the net proportion of company earnings estimates revised higher each month. February ratios suggest that new orders will top out soon in the US (consistent with evidence presented in yesterday’s post) but continue to recover solidly in Euroland – the “flash” PMIs are released tomorrow.