by Gavyn Davies, Department of Macroeconomic Research, Fulcrum Asset Management.
Investors have been debating whether the central banks are “out of ammunition” following recent excursions into negative territory for interest rates by the Bank of Japan, ECB and several other European central banks. These negative rates have largely proven unsuccessful. The markets have (rightly) taken the view that negative policy rates will prove very damaging to the profitability of the financial sector, despite efforts by the central banks to alleviate these effects. Although the Bank of Japan still seems pointed towards a “tiered” form of negative rates, this will not fully protect the banking sector, and is likely to be abandoned. By contrast, the ECB has understood the problem very well, and has decided to go back to an earlier form of “alchemy”, ie a very rapid expansion in its balance sheet, via purchases of private sector assets and a massive liquidity injection in the form of a TLTRO with favourable rates. This method is more likely to work than negative rates, which can even be counter-productive.
For many years, investors have been in thrall to the central banks. But recently this has started to change. In particular, the excursion into negative interest rates has caused alarm in the markets.
There is much talk that monetary policy has run out of ammunition. This talk surfaces as fre- quently in discussions with central bankers them- selves as it does with investors. To quote the title of Mervyn King’s riveting new book, is this “The End of Alchemy?”
Last Thursday, the ECB announced a new pack- age that included a rate cut deeper into negative territory. After initial doubts, the equity markets were impressed, because Mr Draghi had learned the lessons of past failures (see Davies, 2016). The pack- age cleverly protected the banks against the effects of negative rates, drew a line under further rate cuts, and instead included a boost to the ECB’s balance sheet that is likely to be much larger than markets initially realised (see Figure 1).
At the ECB, negative rates are probably now dead, but other forms of “alchemy” are still very much alive.
Why do risk assets apparently abhor negative interest rates? After all, bond yields have fallen in response to negative policy rates, and in the past other asset prices have usually benefited from a re- duction in the discount rate to be applied to future corporate earnings.
In theory, there should be no abrupt change in the effectiveness of monetary policy when rates pass through zero, assuming there is no stampede into paper currency as banks and their customers seek to avoid the levy from negative rates. But theory has not worked too well in this instance.
Households seem to believe very strongly that the natural state of affairs is for nominal interest rates to stay positive. A sense that something must be very wrong if you have to pay to lend money could damage economic confidence. That is exactly what has happened in Japan1, where consumer confidence has imploded since negative rates were announced.
But the real problem is the impact of negative rates on the profits of banks and financial companies. Mr Kuroda still seems to be under the impression that his tiered system of rates on bank reserves held at the BoJ will prevent any damage to profits in the financial system, but the markets profoundly disagree with him.
This is the crux of the matter, and it is still widely misunderstood, so it is worth spelling out the mechanism.
At the simplest level, banks basically have two types of assets: reserve holdings at the central bank, and higher yielding loans and bonds. The BoJ’s tiered mechanism will prevent the payment of nega- tive interest rates on the bulk of bank reserves, so there will be little loss of income on that score. But money market rates have gone negative, and have taken bond yields down with them. There is also likely to be a sustained drop in bank lending rates, so banks interest income will drop significantly.
That would be fine as long as banks were able to maintain their interest margins by paying neg- ative rates on deposits. But that is very difficult, because households and some corporates would pre- fer to hold physical banknotes rather than see their bank accounts shrink in nominal terms. Further- more, central banks do not want to hear a “giant sucking sound” as the banks see a major drain on their deposit base, which could lead to a drying up of credit availability in the economy. And politicians do not want the opprobrium from voters that a levy on bank deposits might bring.
This means that interest rates on bank deposits are unlikely to go substantially negative, though banks are trying to increase revenue from fees on deposits and other services, which would have a similar effect until customers rebel. It is the inability to drive deposit rates far into negative territory that is the real problem for banks when the authorities force bond yields below zero by “going negative” on policy rates. The fact that they maintain positive interest rates on the bulk of the banks’ reserve hold- ings at the central bank is somewhat helpful, but is not enough to protect bank profits overall.
A recent research paper by Claudio Borio at the Bank for International Settlements argues that bank profitability is damaged in a non linear way when interest rates fall and yield curves flatten, and that applies in spades when rates go negative. Borio says that negative rates “could cripple banks’ mar- gins, profitability and resilience” (see Borio et al., 2015). The markets cannot be expected to ignore such profound effects on the financial system.
Negative rates should therefore be shelved by the central banks as a policy tool. They might even be counter-productive.
However, that does not mean that there is noth- ing left in the locker, as Mr Draghi demonstrated on Thursday. The announcement included new pur- chases of corporate bonds, opening the way for fu- ture purchases of private sector assets, where the potential is very large2.
More important in the immediate term, the pack- age used a massive new TLTRO to inject liquidity and reduce banks’ funding costs, thus protecting the profitability particularly of the weakest parts of the banking sector in the face of negative rates.
It will simultaneously increase the likely size of the ECB’s balance sheet to 40-50 per cent of GDP, depending on the take up of the TLTRO (see the Methodological Note below for the analysis of the measures). This is much larger than the size of the ECB balance sheet at the previous peak in the euro crisis of 2012 (see Figure 1).
The effectiveness of this method to protect the banks was reflected in the jump in their share prices after the ECB meeting. But it cannot be used again and again, because the maximum size of the TL- TROs is limited. Mr Draghi therefore conceded that the ECB is probably now at the effective zero lower bound. Central bankers have been searching for it, and now they have found it.
The ECB’s massive shift back to an earlier form of “alchemy” will be an acid test for the future of un- conventional monetary easing. There are certainly legitimate doubts about its effectiveness. But at least it is not going down a path that is actually counter-productive.
Methodological note on estimating the impact of the ECBs TLTRO-II operations3
On 10 March 2016, the ECB introduced a new se- ries of TLTROs4. The rate for these new TLTROs will be the MRO rate at time of take-up (currently 0%), plus a reduction depending on each bank’s net lending vs. a benchmark (based on its past lending trajectory), which could take the overall rate as low as the deposit rate (currently -0.4%). Based on the current pace of loan growth, which has started to recover in the Euro area, it is likely that a large number of banks will be able to get this reduction.
In terms of size, banks will be able to borrow up to 30% of eligible loans net of their borrowing in the 2014 TLTRO (currently EUR 212bn). Eligible loans comprise loans to non-financial corporations and households (excluding mortgages) in the Euro area. According to ECB data, this equals EUR 5.6tn as of January 2016, making the potential size of the new TLTRO programme as large as EUR 1.4tn.
In practice, however, banks are unlikely to take up the maximum amount. Firstly, they will not want an excessive part of their funding to be re- liant on the ECB. Secondly, one of the effects of the ECBs package has been to reduce corporate spreads for banks, thus making central bank funding less attractive at the margin. Additionally, bank deposit growth in the Euro area has been outpacing loan growth, thereby reducing external funding needs for the banks. The actual take up at the time of TLTRO 2 operations will depend on these factors, as well as the banks available collateral.
We estimate the potential take-up and related ECB balance sheet expansion using the following methodology: we take country level data on eligible loans and then apply a different take-up rate for core (25%) and periphery (60%) banks, reflecting the rel- ative attractiveness of the TLTROs for banks in the two parts of the Euro area. Given this, we conclude that total take-up over the course of the operations could amount to EUR 615bn. Banks will be given the opportunity to repay the 2014 TLTRO (EUR 212bn), which does not have the potentially negative rate. Assuming most of that money is repaid, we compute that the new TLTRO should contribute to approximately EUR 405bn of ECB balance sheet expansion over the next year.
Given the potential for a negative rate, the new TLTRO would support profitability of the banks by lowering their funding cost. Based on the estimated take up above and assuming attainment of the current -40bps deposit rate, Euro area banks would gain close to EUR 2.5 billion of profit. For the listed sec- tor, this translates to around 3% of 2017 net income. This is comparable to the earnings headwinds from the 10bps cut announced on Thursday and could potentially lead to a net earnings increase if the take-up is larger or the funding, which is replaced with the TLTRO, more expensive (although some of that might be offset by asset spread contraction).
This stands in sharp contrast to the tiering sys- tem introduced by the BoJ, which avoided banks from paying a negative rate on part of their reserves, but did not shelter them from the net interest mar- gin compression that resulted from the introduction of negative rates. The relative performance of Euro area and Japanese banks after their respective policy announcements reflect, amongst other things, this sharp difference.
- Borio, C., L. Gambacorta, and B. Hofmann (2015): “The influence of monetary policy on bank profitability,” BIS Working Paper, WP n. 514.
- Davies, G. (2016): “Mario Draghi and the ECB learn from experience,” Financial Times [Online; posted 10-March-2016].
1) “Negative Rates Dig a Hole for Bank of Japan,” Wall Street Journal, 10 March 2016
2) The purchase of non financial corporate bonds is likely to be limited by the size of the market to only about E50 billion per annum. But other private asset pools provide much greater possible scale for central bank asset purchases. Nikalaos Panigirtzoglou of J.P. Morgan estimates the following sizes for private asset pools: uncovered bank bonds (E1tr), bank loans (E10tr) and equities (E6tr). The ECB has the legal mandate to purchase significant proportions of these asset pools, though it would only contemplate doing so in extreme macro economic conditions.
3) Matthieu Walterspiler, Fulcrum Asset Management. Marble Arch House, 66 Seymour Street, London, W1H 5BT
4) “ECB announces new series of targeted longer-term refinancing operations (TLTRO II),” European Central Bank, 10 March 2016.