The May reserve period started with an improvement in overall liquidity conditions as the surplus moved up €27bn to €113bn after this week’s ECB operations. Indeed, at the MRO, total borrowing and the number of bidders both increased (by €8bn and 22 banks, respectively). The STRO posted similar results (+€4bn and +20 banks, respectively).
At the weekly sterilisation auction, the ECB failed again to sterilise the amount announced (€167.5bn), and demand for the 1-week term deposit was less than the previous week (-€21bn). Given the tighter liquidity conditions stemming from the fall in the surplus to about €80bn, banks have adjusted their demand for borrowing and weekly deposits at the ECB.
This “automatic” adjustment mechanism is favoured by the full allotment procedure at the ECB’s operations and is triggered any time Eurosystem liquidity conditions tighten. The absence of any liquidity-absorbing shock (ie, abrupt increase in autonomous factors) and improvement in overall liquidity conditions has favoured the decline in Eonia fixing to 17.7bp on the first day of the new reserve period.
The latest Eurozone real economy data have not disappointed and have increased the chance of ECB action(s) in June. Inflation was confirmed at 0.7% in April, just 0.2pp higher than in March, defying initial consensus expectations of a rebound to 0.9%. Q1 GDP growth was very weak at a mere 0.2% q/q in the eurozone (below our and consensus expectations of +0.4% q/q). More significantly, growth was flat in France and contracted by 0.1% in Italy.
At the May meeting, ECB President Mario Draghi repeated that euro strength was a concern and stated that the GC was comfortable acting in June, owing to a consensual view that the
path of inflation was unsatisfactory. We think the lower-than-expected inflation over the
past few months will lead the ECB to lower its current inflation forecasts (currently 1% in 2014, 1.3% in 2015 and 1.5% in 2016), likely triggering a package of measures (as the latest ECB’s rhetoric seems to suggest).
Our economists expect the ECB to cut official rates at its June meeting, bringing the refi rate to 10bp and the deposit facility rate to -10bp. The launch of a targeted liquidity injection to support bank lending is another possibility. In this respect, it is worth noting that as Reuters reported on 12 May, the Bank of Italy is working on some measures to broaden the collateral pool within the Eurosystem collateral framework to include a range of assets related to SME lending.
Some of them could target current account facilities provided by banks to SMEs that allow them to borrow up to an agreed amount while they wait for payments from their customers. Although a large-scale asset purchase program (QE) cannot be ruled out for June, the fact that medium- to long-dated inflation expectations are still anchored leads us to believe that a more gradual approach via rate cuts is the most likely action at this stage.
Dovish rhetoric from the ECB, along with the latest disappointing data, have reinforced market expectations of some ECB actions at the June meeting (or in the next few months), not only in the form of a policy rate cut (5bp cut in the refi rate fully priced in June) but also of a liquidity injection via large-scale asset purchases.
Indeed, current pricing of 1y Eonia at 7bp, and of Eonia fixing at 5bp in the December reserve period, suggest expectations not only of a refi rate cut but also of a rise in excess liquidity (possibly because of a liquidity injection via asset purchases, as the drop in the 10y Bund yield to about 1.30% could suggest).
Interestingly, the part of the curve beyond mid-July flattened sharply with the 1y1y Eonia forward declining to 11bp. Such a move could be interpreted as a sign of rising expectations of extension of the full allotment and/or injection of long-term liquidity.
In the context of tightening Eurosystem liquidity conditions, the Eonia normalisation toward the refi rate should continue. We expect 3y LTRO repayments to continue, lowering excess liquidity even if part of the repayment is likely to be switched to the ECB’s shorter refinancing operations.
Therefore, any reduction in the depo rate – even into negative territory – would not be significant for the Eonia fixing. In contrast, the reduction in the refi rate would bring down the fair level of Eonia, probably to 8-10bp in the case of a refi rate at 10bp. Such action would reduce the Eonia level but not its volatility during phases of tighter market liquidity (eg, at the turn of the month).
Therefore, the reduction of the MLF rate would be very effective in containing Eonia volatility. Of course, any liquidity measures that could increase excess liquidity well above €200bn would re-establish the relationship between Eonia and the depo facility rate, thus making the cost of the overnight liquidity sensitive again to the depo rate level. Thus, a negative depo rate might create the conditions for a negative Eonia rate.
However, it is important to note that Eonia is the weighted average of rates on overnight lending of unsecured liquidity in several eurozone jurisdictions. Therefore, a negative fixing would imply a negative credit premium, which would not be sustainable.
Thus, although an initial period of volatility and negative printings for Eonia would be likely in the case of a negative depo rate and liquidity injections, we would expect Eonia to subsequently stabilise at about zero.
We do not see the case for further liquidity measures to be implemented, as liquidity conditions have improved and the Eonia volatility increase has mainly been the result of
geographical fragmentation of the markets. Full allotment is already a powerful tool to prevent liquidity problems, and its extension beyond the current expiration of mid-2015 (which could be part of the measures announced at the June meeting) would, in our view, effectively reinforce forward guidance and reduce the term premium at the back end of the money market curve. Note that its extension for another year (for example, until 2016) would be similar to another VLTRO in terms of liquidity availability.
Suspension of the weekly sterilization auction would increase the liquidity surplus above €200bn for a time, but it would not reduce Eonia volatility. Indeed, the main bidders at this operation are banks in core countries that are likely to keep depositing at the ECB or investing in other core assets in case the auction is suspended and not lend in the liquidity market to other banks.
In case of any conditional liquidity operations, the allotment is unlikely to be as big as the 3y LTROs in terms of net liquidity injections; therefore, the effect in terms of the increase in liquidity surplus would be minimal, and insignificant in terms of the effect on Eonia.
Large-scale asset purchases have the potential to increase excess liquidity in the system for a long period, thereby pushing Eonia close to the deposit facility rate with a flattening of the money market curve even to long maturities (eg, blue and gold contracts on the Euribor strip). However, at this stage, we believe the bar for such bold measures remains high.
So, based on our expectations of ECB actions in June in the form of a policy rates cut (with negative depo rate) and possibly some conditional liquidity measures and the extension of the full allotment, we expect the Eonia fixing to decline to 8-10bp, starting with the June reserve period.
The weak inflation outlook should reinforce the ECB’s forward guidance, thus favouring the stabilisation of the Eonia curve at about 10-11bp approximately for the period of the extension of the full allotment. A possible trading range for the 1y Eonia is likely to be 7-13bp. Indeed, the current level of the 1y Eonia at 7bp probably reflects that some likelihood of asset purchases is currently priced in.
This is also evident in the shape of the Eonia forward curve, with the December Eonia priced at 5bp. Although we do not rule out such option, we believe that it is too early for the ECB to embark on QE. Therefore, we see limited room for a further decline in the 1y Eonia and at the current level we suggest paying it tactically.
Similar monetary policy expectations explain the decline of the 1y1y Eonia forward to 11bp. In our scenario of a policy rates cut and no massive liquidity injection, the 1y1y Eonia forward rate is likely to move to 10-20bp.
Thus, we regard any further decline from the current level as an opportunity to pay. Across countries, the different monetary policy outlooks suggest going long 1y1y Eur OIS forward vs 1y1y USD OIS forward.
Looking at the Euribor fixing, with the 3m Eonia likely stabilising at about 10bp and the FRA/Eonia staying at 14bp, we put the fair level for the 3m Euribor at about 24-25bp. Importantly, the accommodative monetary policy stance should favour some tightening of the basis.
However, the higher sensitivity of the Eonia compared with the 3m Euribor fixing to a change in monetary policy conditions should cause an initial widening of the FRA/Eonia with a gradual adjustment afterward.