In a world of excessive liquidity, where open market operations by central banks are ineffective, unconventional and unheard-of monetary policies need sometimes to be applied. In this article we will briefly discuss of the overnight interbank lending market in Europe (and changes in how its respective index is calculated) and in particular discuss of “tiering”, a policy adopted by the ECB in September’s meeting, whereby interest rates on the deposit facility would depend on the amount of excess reserves compared to the required reserves that a bank holds.
Interbank lending, EONIA and bank profitability
What is the interbank lending market? Banks constantly exchange money, assets and liabilities for each other. Sometimes a bank may temporarily not have enough money to settle an incoming liability (or satisfy reserve requirements), usually because the assets it holds don’t have an immediate payout. Other times, the reverse may happen, and banks may end up with excessive cash, which would need to be somehow invested or at least deposited. This is the reason for the existence of the interbank lending market: so that temporary deficits/surpluses of cash may be fixed by borrowing/lending to other banks within a 1-day time frame.
In Europe, if a bank can’t find a lender (highly unlikely as we will see) it can borrow directly from the ECB, paying the marginal lending facility rate (0,25%), and the reverse, if a bank cant find a borrower it can deposit cash at the ECB, at the deposit facility rate (before 12/09/2019, -0,4%).
Together those two rates form the so-called interest rate corridor within which we can find all the one-day interest rates of interbank lending. The EONIA (Euro overnight index average) tracks all interbank lending transactions that happen during a day and have a maturity of 1 day and shows the volume-weighted average interest rate for the day.
From the graph above we can notice that, with the exception of periods of particular financial stress, after the financial crisis the EONIA has been de facto set by the deposit facility rate, with the former being much closer to the latter rather than being in the “middle” of the corridor: This is due to the fact that reserves are superabundant making lenders easier to find than borrowers. In short, due to excessive liquidity, one of the main tools for the ECB to manipulate short term interest rates in Europe is setting the deposit facility rate.
As rates on deposits are negative, this implies that when banks are holding reserves, they are making a loss.
Data in Bn€
From the graph above we can see that this is a serious issue, especially from banks from the core Eurozone. Losses due to negative interest rates deposits are estimated to be €7.15bn yearly, which is 0,4% (the reverse sign of the current deposit facility) of the approximately €1788bn of current Eurozone bank reserves, more than 50% of which are owned by German and French banks.
However, the ECB decided to cut the rate by 10bps in September’s meeting, potentially making this problem even worse. How can the ECB avoid this issue? The answer is the ECB’s new policy: tiering.
A brief explanation of tiering and its application across the world
There are four countries in the world right now that make use of this instrument: Japan, Denmark, Switzerland and Sweden.
The system proposed in Europe is essentially the same as the Swiss one, which is based on the concept of minimum reserve multipliers. In the Swiss system, there is a threshold which is equal to 20 times the minimum reserve ratio, currently set at 2.5%. When a Swiss bank owns Reserves/Required reserves>20 it must pay 0,75% (since the current interest rate is -0,75%) on reserves exceeding the threshold and 0% on any reserves below the threshold.
A practical (and simplified) example: Suppose a swiss bank owns 100CHF assets and is funded 50% by deposits. The minimum reserves that the bank would have to own would be 50*0,0125=1.25CHF. Our beloved bank, however, holds 30CHF in reserves, 5CHF in excess to 20*1.25CHF=25CHF. At this point, our bank would receive an interest rate of 0% on the 25CHF and pay 0.75% for the privilege to deposit 5CHF at the Swiss National Bank. Again, in normal conditions those 5CHF would have been lent in the interbank lending market to a bank which is instead in deficit of CHF, in exchange for an interest rate (or at least a rate superior to -0,75%): however, as stated at the start of the article, excess liquidity forces this money to instead be deposited at the central bank and, in fact, the data shows that Swiss banks hold approximately 30 times the required reserve on average.
Tiering need not be based on just minimum reserve multipliers. In Japan, for example, a “temporal” clause exists any reserves accumulated in 2015 pay an interest rate of 0.1%, while any reserves accumulated after following a system similar to the Swiss and Eurozone one, paying 0% and -0.1% interest rates.
Tiering “arbitrage” and its influence on EONIA and €STR
The ECB decided to set the reserve threshold at 6 times the minimum reserve requirement, within which deposits are “excepted” from negative interest rates. A superficial analysis may make us think only in terms of what the cost savings may be for the banks at the current excess reserves that they hold. A deeper analysis reveals that trough arbitrage excess reserves should be redistributed slightly more equally across the Eurozone banks, with everyone benefiting profit-wise.
Here is how it works: The current deposit facility rate is -0.50%, but any reserves before the threshold of 6 times the minimum reserve have an interest rate of 0%. As we saw before, German and French banks may on average hold more than 6 times the minimum reserve, hence the “exception” on interest exists holds for only a part of their reserves. On the other hand, Spanish and Italian banks may hold less than 6 times the minimum reserve. The same is true for larger vs smaller banks, with the former having more excess liquidity than the latter.
From this, it is easy to guess what the next step would be… Banks which hold excess liquidity will lend it to banks which lack excess liquidity at an interest rate higher than -0.50%, while the counterparties will deposit the money right back to the ECB, where their deposits would be excepted and earn an interest of 0%. As larger and “core” banks own more excess liquidity than smaller and more euro-peripherical banks can handle the lending rates should be closer to -0.50% than 0% as “demand” for an exception is smaller than “supply” of excess reserves
This implies that EONIA should track closer to the upper level of the interest rate corridor as banks are lending to each other at a rate higher than the deposit facility rate… However, EONIA, starting the 2nd of October will be calculated differently. The formula will be 8.5 basis points+€STR.
€STR is the new index which will calculate interbank lending rates. The methodology is the following, after aggregating transactions by their interest rate:
- Remove the top and bottom 25% interbank transactions in terms of volume
- Calculate the volume-weighted mean that spans within the two quartiles
As those “arbitrage” operations are likely to have lower volumes than normal interbank lending €STR will likely remove those, resulting in the EONIA tracker being again closer to the deposit facility rate.
Potential problems and TLTRO arbitrage
There is also the issue of a potentially counter-productive (for the ECB monetary goals) arbitrage: the TLTRO arbitrage. As we saw before Spain and Italy are amongst the countries which lack excess reserves on the scale of their Central European cousins. Spain and Italy happen to be the main beneficiaries and users of TLTRO.
Supposedly, Italian/Spanish banks (as well as other eligible ones) may receive money from TLTRO, receive a better discount rate with their operations and then send back the money to the ECB by buying reserves within 6 times the minimum reserve ratio, earning a riskless profit based on the differential between TLTRO loans and 0% ECB deposits.
Combine this with incentives for peripherical banks to accumulate reserves till they reach the threshold and we can easily guess the result: an involuntary monetary tightening as peripherical banks deposit money to the ECB rather than lend to the domestic government/economy.
It’s hard to decipher what the market thinks of tiering itself as besides it Draghi announced the recommencing of QE and lowered interest rates. Both those measures, however, appear to have been largely anticipated by the markets Notice the steepness ahead of a potential QE2
In fact, yields on the 10Y Bund even slightly rose, after an initial large drop, on the day of the ECB meeting, while the spread between ECB deposit rates and bund rates lowered significantly.
According to analysts the 3 banks that benefited the most from the combination of all the ECB’s policies in terms of EPS are Deutsche Bank, Abn Amro and Ubi Banca. However, the market seemed uninterested and reacted negatively to the ECB meeting, despite the -10-bps rate cut being already priced in by the markets. Deutsche’ s stock ended the day at -0,95%, Amro remained flat and the positive gains form UBI Banca, +1.56%, appear more to be a result of the lowering spread of Italian gov’s with the bund. The fact that the latter lowered even further may point out to the direction that the market does not believe that the “accidental monetary tightening” in peripherical countries, which we previously discussed, will occur.
In conclusion, the full effects of tiering are yet to be seen and it is likely that the ECB may change its rules in the future, especially if the tightening scenario ends up true: Only time will tell how effective this tiered system is and how it will evolve or if we can interpret it, together with the lack of enthusiasm showed by the markets for QE, as just another sign of “Japanification”.