TLTRO and reserves: the ECB brings tightening to money markets

From 23 November onward, and until the maturity or early repayment of each TLTRO tranche, the rate at which banks borrow funds will now match the rate at which they are placing them back at the ECB. In other words, the carry opportunity of keeping TLTRO funds will disappear. For the earlier part of TLTRO loans, the previous interest rate calculation mechanism still applies. As our bank analyst colleagues have noted, keeping TLTRO isn’t without cost for banks, so one has to assume significant repayment at the next opportunity. This should not amount to the whole amount borrowed however, as alternative market funding can be more expensive.

As we wrote ahead of the meeting, this is going to cause some challenges to banks’ cash management, but it is the option that risks the least money market disruptions, at least compared to various reserve tiering alternatives. We estimate this will save more than €30bn in interest payment to the ECB, however. In summary, early repayments should be significant. This will result in a commensurate reduction in excess liquidity, and an injection of collateral in the financial system (more on this in the next section).

We would expect a widening of the 3m Euribor-Estr basis from -2bp currently to above 10bp

Excess liquidity is currently elevated and even a drop of, say, €1tr should not result in a jump in money market rates on its own. We do think, however, that it will make them more sensitive to other factors, in particular credit risk. We would expect a widening of the 3m Euribor-Estr basis from -2bp currently to above 10bp. We foresee a less dramatic rise in Estr rates relative to the ECB deposit rate. There isn’t enough history of Estr rates to determine precisely their behaviour when excess liquidity shrinks, but we think the new calculation method compared to Eonia means broader liquidity pools become relevant for its calculation.

One point made by ECB president Christine Lagarde during the press conference is that early TLTRO repayments would also release the collateral placed by banks at the ECB in exchange for funds, thus helping with the current shortage. There is probably some truth to this but note that the shortage concerns mostly German collateral. Whether TLTRO repayments help with this shortage depends on how much German sovereign collateral was posted at the ECB in the first place, and whether the banks that posted it do repay their loans early.

We think the new TLTRO terms should assist the already sizeable tightening of swap spreads

At the margin, we think new TLTRO terms should assist the already sizeable tightening of swap spreads. For one thing, some tiering measures would have actually worsened the collateral shortage, so this is a relief. German collateral released by TLTRO repayments itself might not be much, but overall paying down the TLTROs can mean on aggregate less excess liquidity chasing the same high-quality collateral so dearly in demand. This all adds to additional lending on repo announced by the German treasury to finance the energy support measures, further supply in 2023, and eventually quantitative tightening. More broadly, the age-old problem of collateral shortage is now a monetary transmission problem, suggesting the ECB will be keen to prevent the shortage from worsening.

This being said, some questions remain. The main one is the wall of sovereign funds currently placed at the ECB at the deposit rate which will have to find an alternative home after April 2023.

The last change brought by the ECB relates to the remuneration of minimum reserves banks are required to hold at the ECB from the refinancing rate to the deposit rate. If the corridor (the spread) between the two remains at 50bp, this would only save the ECB €0.8bn per year on the €164bn of bank minimum reserves. Unlike the change to TLTRO terms, this will now become a permanent feature of ECB policy, however. Still, this is somewhat good news for banks, in the sense that the remuneration rate could have been dropped lower, to 0%.

More to the point, we do not think this will make much change to money market rates.