Draghi between Scylla and Charybdis

Draghi will need all his rhetorical ability to tread a narrow path at the next Press Conference. On one hand, he will need not to disappoint market expectations, which grew after his recent speeches in Frankfurt and at the European Parliament. On the other hand, I do not think he will have something precise to offer in terms of additional measures, with two possible minor exceptions I will present below. This is not so much because of the utterances by Weidmann, who keeps repeating the usual misgivings about further measures, but rather because of statements by Draghi himself and other Council and Board members stressing that, before introducing new measures, one has to see whether the ones that are being implemented right now will be sufficient. In my sense, this includes not only the monetary measures as such but also the consequences of the Comprehensive Balance Sheet Assessment. I expressed a cautious assessment in this last respect in a previous post , but it is clearly to early to pass a definitive judgment. 1Constancio even ventured a period of one quarter to discern possible effects of the measures already taken before considering new moves. In a way, one could see these statements as a sort of expectations management.

There are two reasons to expect a pause, albeit an active one, in additional strong measures, one is more political, the other is more substantial.

The political reason is that in the Governing Council there is right now no “comfortable majority”  for new measures. For confortable majority I mean something more than the Kuroda majority but also quite less than unanimity. I think the Governing Council would be loath to decide against a minority of 4 members or more, especially if these members would all belong to the creditor group.

The substantial reason is that there is a lot of uncertainty about the possible effects of the outstanding measures. I think this applies in particular to the Asset Backed Securities (ABS) purchases, in addition to the  Balance Sheet Assessment, which I mentioned already. These effects will only become clear over time, especially taking into account that improvements will have to be realized in an inertial variable like inflation prospects.

In order to avoid both Scylla and Charybdis, Draghi could be a bit more specific about the measures that the Eurosystem Committees are studying, in addition to repeating the mantra about the two unanimous conclusion of the Council – willingness to do more unconventional measures in the case of need and the expectation that the balance sheet will go back to the level reached at the end of March 2012. By saying that no decision has been taken as yet but that preparatory work for possible decisions has progressed he may manage to tread the narrow path that I mentioned at the beginning.

As I said above, there is also a chance, not a high one though, that he could announce two minor adaptations to current policies.

First,  he could announce the extension of the ABS and Covered Bond (CB) purchases to other asset classes, in particular corporate bonds. This should not be too difficult (politically and technically), notwithstanding the recent article by the Bundesbank  fearing a bubble in this area. It is not clear to me, however, why these were not added from the start to the purchases of ABS and CB. In addition, purchasing them now would make unavailable the fig leaf that could be used when deciding Quantitative Easing as including also (as was the case for the FED, the Bank of England and the Bank of Japan) private paper in addition to sovereign one.

Second, he could make the Targeted Longer Term Refinancing Operations (TLTRO) more attractive, say by eliminating the 10 basis points surcharge or increasing the amounts available or even lengthening the maturity. This could also help deal with an issue raised by some observers, i.e. whether the negative remuneration of ECB deposits is an obstacle for QE, which is the next topic I deal with in this post.

The fear that the negative remuneration of ECB deposits may hinder QE starts from the assumption that the negative deposit rate is the only rate determining the willingness of banks to hold deposits with the ECB. To me, the deposit rate cannot be looked at in isolation and has to be considered, in particular, together with the rate at which the ECB lends money, i.e. the MRO rate, currently at 5 bp, and the TLTRO rate, currently at 15 bp. In a way this is a specific application of the general point first made by Tobin that the demand on any asset does not depend only on its return but on the returns on all available assets. The difference between the cost of ECB lending and the (negative) remuneration of deposits gives the cost of central bank intermediation and this has not changed with the latest rate cut but was, over time, substantially lowered as the width of the corridor came down. This is the “intermediation margin” (similar to a mark-up between a bank lending rate and deposit rate) that determines the willingness of banks to borrow from the ECB and re-deposit the excess liquidity with it.

Consistently with this point, one cannot look at the deposit rate and at the deposit amounts in isolation. It is true that it is mostly core banks that deposit the money with the ECB but it is not them that borrow it, it is rather the peripheral banks, which use the proceeds to pay interbank liabilities to core banks and deposit money with them, at negative rates corresponding to EONIA and OIS rates. Sure, the banking system overall pays an intermediation margin to the ECB, but this is a kind of insurance premium it is willing to pay given the fragmentation of the money market. Some banks can borrow more cheaply from the ECB than from the market, while some other banks lend money cheaply (indeed at negative rates) to the ECB but not to peripheral banks. It is thus risk considerations that justify a positive intermediation margin for the ECB. Indeed the ECB intermediation margin can also be seen as a liquidity  insurance premium.

The fear that the negative rates and QE instead of being complementary and consistent measures to reach certain ECB objectives, such as a weaker currency, are conflicting measures is not warranted in my view. At any point in time and for any “ECB intermediation margin”, there is a demand for ECB intermediation, as I have shown in a previous post. 2 If the ECB tries to provide more liquidity at the same price through outright purchases, banks will answer reducing the amount of liquidity that they take from the ECB through repo operations. The way to give incentive to banks to take and hold more liquidity from the ECB (either through repo or outright operations), and thus increase outstanding liquidity, is to lower the price of ECB intermediation. The lengthening of the lending through the TLTRO is one way to implicitly do this, as the ECB now lends at 4 years for the same price at which it lends at 3 month in the normal LTROs. Further increasing the maturity or eliminating the 10 bp spread over Main Refinancing Operations for TLTRO operations would be other ways to directly reduce the “intermediation spread”. Exercising a lowering impact on the yield of purchased securities either through the ABSPP, the CBPP3 or QE is another way to indirectly achieve the same result: holding liquidity with the ECB would cost less relative to the lower yield of purchased securities. Just to see this more graphically: if QE makes the return on more assets negative, the negative rate on ECB deposit is less of a disincentive to hold liquidity. This also means that the ECB has to pay dear for its purchases. This is not in the sense that the ECB must pay prices higher than the current ones, but in the sense that by its own action the ECB increases the prices (and lowers the yields) of purchased securities, thus making its intermediation cheaper relative to the yields of purchased assets. Indeed already the expectation of QE has lowered the yield of securities that could be bought in QE. This last observation also sheds light on the question of the effectiveness of QE. The point is often made that QE cannot work when yields are already so low. But yields are so low also because of the expectation of QE, so their low level is proof of the effectiveness of QE even before it was launched, rather than a reason of its lack of effect.

In the end, if the banks incentives to borrow more from the ECB given by the longer TLTRO and the lower yields on purchased assets would turn out not to be enough, direct action on the ECB lending margin, i.e. the narrowing of the corridor, one way or the other, would be needed. But in current conditions this would be better done by eliminating the spread on the TLTRO or bringing the MRO rate down to 0 rather than by increasing the deposit rate towards 0.

Let me slightly revise the expectations that I gave some time ago:
1. If inflation prospects will not improve there is a 55 per cent probability of QE around the end of the first quarter of 2015.
2. If inflation prospects keep worsening there is an 85 per cent probability of QE within the first quarter of 2015,
3. If inflation prospects improve the probability of QE falls down towards 0.

4. Corporate bond purchases are about equally probable as a late/early addition to the private purchases (late because decided later than the other purchases/early because taking place before QE) or as side dish of QE.

5. There is a significant but not high probability, say 30%, of a revision of the TLTRO conditions to make it more attractive, in particular the elimination of the spread over the MRO.

6. I regard as low (less than 25%) the probability of purchases of “agencies” (CDC, CDP, KfW, ICO, ESM, EFSF, EIB,…) and international institutions (WB, EBRD, IADB,..). As regards agencies, they are a bit too close to sovereign to really provide a political alternative and too small to provide a practical alternative for large amounts. As regards international institutions, it is unclear what the effect on the € area would be, in addition to the limited amounts available in €.

  1. Assessing the Assessment[]
  2. Liquidity supply and demand in the €-area: a schematic approach[]