Ever since the first warning of crisis in August 2007, indications of global banking health and liquidity have been turned upside down. Nowhere is that more evident than European banking, especially in the periphery. Observers have been looking for lower interest rates and reduced reliance on the ECB as evidence of “normalcy”.
Starting with the chief interbank rate in Europe, Eonia, it appears as if the ECB has succeeded in that respect. Eonia is the European equivalent of the Fed funds rate, the rate at which banks charge for unsecured overnight funding. We know there is a large quantity of excess liquidity simply from levels in the ECB’s deposit account, so Eonia is a measure of the flow of that excess, hopefully to the peripheral banks that need it most.
From all appearances the near-zero Eonia since July 2012 would indicate, on the surface, a high degree of liquidity even in unsecured terms.
To complete the liquidity picture, however, we have to include the mechanics of ECB monetary management to provide more meaningful context. Monetary policy operates through targets and an established monetary corridor. The deposit rate provides a floor for the corridor – Eonia should never fix below the deposit rate since banks with excess liquidity can park it with the ECB at the deposit rate with zero counterparty risk. The Marginal Lending Facility (MLF) rate functions as the corridor ceiling, akin to the Federal Reserve’s Discount Rate.
Eonia itself should normally trade just above the minimum bid of the ECB’s Main Refinancing Operation (MRO) rate which is set at the midpoint between the MLF and Deposit rate corridor. Since the MRO is collateralized it sets a comparison between unsecured and secured lending in the short term (Eonia is overnight, MRO is one week).
The spread between Eonia and the MRO should be slightly positive since it essentially marks the point at which banks can switch from unsecured terms – collateralized lending rates are typically below unsecured due to the increased security of collateral, meaning a reduction in counterparty risk (all else equal).
While Eonia on occasion would trade slightly below the MRO, beginning in August 2007, and the first indications of eurodollar problems, the Eonia/MRO spread was often negative and, for a time, extremely volatile. This is counterintuitive in that banks should stop unsecured lending at the margins once Eonia terms fall below the MRO, however once banks in that period became distrustful of counterparties and their collateral they opted to park funds with only “strong” banks at whatever rate – the interest rate became a secondary consideration. This was a breakdown in basic monetary mechanics for both “money” stock and flow.
Like LIBOR and GOFO, Eonia is a fixed survey of the largest banks in Europe. Therefore it represents, even on a volume-weighted basis, solely the liquidity of these larger institutions. In that context, the negative Eonia spread makes sense – excess funds pulled away from the periphery and concentrated with “safer” banks, thus depressing Eonia and keeping it below the MRO. The negative spread to the MRO, then, is an indication of a tiered liquidity market.
From August 2007 through the panic in 2008, the ECB increased the level of liquidity (money stock) theoretically available to the European system. But, again, the problem was one of flow and the excess liquidity ended up concentrated in the largest banks. As the ECB ramped up liquidity relief measures, Eonia was depressed all the way to the Deposit rate floor.
The ECB attempted to “normalize” the system in 2011 and it appeared for a short while as if Eonia and interbank lending on an unsecured basis might respond. We know how that turned out as the sovereign collateral problems actually worsened. Eventually, the ECB flooded liquidity through the LTRO’s, increasing excess “reserves” by about €800 billion. But the persistence of the negative Eonia/MRO spread indicates excess liquidity still only trades at the biggest banks. The trade-off for this liquidity in the big banks is either among each other at Eonia or parking it idle at the ECB deposit account (which is why Eonia now trades very close, 6-10 bp, to the Deposit rate).
In other words, the ultra-low Eonia rate is actually indicating reduced effective liquidity throughout Europe. As long as the negative spread to the MRO remains, it shows that peripheral banks are not receiving “their share” of the ECB’s generosity. Instead, these weaker institutions are receiving marginal liquidity almost exclusively through the ECB’s collateralized regimes (either MRO or LTRO). Better liquidity conditions would actually be demonstrated by a rising Eonia rate (as we have gone through the looking glass).
If peripheral banks have been forced into collateralized lending at the ECB, it stands to reason that a reduction in ECB dependence would signal growing interbank health. The most visible case, particularly since the middle of 2012, is the Spanish banking system. While Greece and Ireland were experiencing more dire conditions relative to Spain, it was the sheer size of the Spanish banking system that drove growing dysfunction.
Where Spanish banks had been drawing €106 billion from the ECB (still a large amount) in November 2011, they were soon tapping an amazing €411 billion by August 2012 as depositors pulled out of Spain (and into Switzerland, Denmark and other “safer” alternate markets). This capital flight was shown in the TARGET 2 imbalance at the Bank of Spain.
Since then, however, Spanish bank reliance at the ECB has fallen all the way back to €270 billion, as Spanish banks have reduced both LTRO and MRO lending balances. On its face, this seems to be a very positive development as Spanish deposit balances have increased slightly, taking off some liquidity pressure. But, like the Eonia spread, there is an upside down element to consider here as well.
Looking at the Spanish banking system as whole, we see that Spanish banks are reducing their ECB balances in response to their own deleveraging. Overall, Spanish intermediation is simply shrinking, including the transfer of about €50 billion in “bad debt” to SAREB (Spain’s government-funded bad bank).
Overall liquidity needs, therefore, are largely unchanged in Spain, as they are in Italy. The Italian system’s reliance on the ECB has been constant since the LTRO’s.
Not surprisingly, we see that corresponds to the overall level of banking assets at Italian banks.
The composition of assets inside Italian banks has changed dramatically, as loans are runoff or charged-off and reinvested in Italian sovereign debt. So where Spanish banks are undertaking a significant deleveraging process, Italian banks are not. But in both cases, marginal liquidity has not changed as both systems continue to depend heavily on the ECB proportionately.
Because of the European system’s continued dichotomy, the reduction in borrowing at the ECB is actually indicative of lower lending levels in the Spanish economy (and we would expect to see the same if Italian banks follow into deleveraging). The gripping macro-economic depression in both Italy and Spain is driving this trend. Bad debt levels in Italy have risen from 5.4% of total bank assets at the beginning of 2012 to 6.6% in February 2013 (the latest data). Not only that, the Bank of Italy is also showing a marked reduction in loss recovery rates on delinquent loans – the vast majority of the increase in delinquencies are loans to nonfinancial corporations (from €50 billion in 2010 to €85 billion most recently).
In Spain, the amount of bad loans has dropped in the past two months, but that is entirely due to the transfer of nonperforming assets to SAREB. The latest figures still show an NPL rate of 10.4%, or €162 billion for February.
From all this we can conclude that despite ultra-low lending rates and decreasing absolute volumes at ECB borrowing facilities, liquidity conditions have not much changed. This is likely why recent rumors of an ECB rate cut have begun to circulate more frequently. A 25 bp cut from the ECB would likely have no impact on the European economy, but it might theoretically be useful to reduce the negative Eonia/MRO spread, bringing peripheral bank funding down closer to core banks.
If effective liquidity in Europe was real and growing, there would be no need at all for a rate cut. Instead, a reduction here would produce some potentially serious complications. First, with the deposit rate floor already at zero, the ECB would have to decide if it wanted to narrow the rate corridor. If the deposit rate is kept at zero while the MRO and MLF rates are reduced, the narrowed corridor might actually decrease the likelihood of a pickup in interbank lending, the opposite of what is really intended.
On the other hand, if the corridor size is maintained that would mean a negative deposit rate and floor, something that is largely unpredictable with no real reliable estimates as to what banks in the core would do with their excess balances. The likeliest scenario, in my opinion, for the negative deposit rate would be an increase in negative yields on core or non-euro sovereign debt. Again, this is the opposite of what the ECB would intend. Would we also see Eonia fall below zero?
One final complication is deposit balances. While March figures show little movement for the PIIGS nations after the Cyprus deposit confiscation, it is still not clear, despite assurances otherwise, that Cyprus has not had any effect. In Greece, for example, deposit balances had recovered rather nicely since August. Overall deposits had risen from €206 billion to €235 billion at the end of March, but the growth rate dropped (only a 1% gain in March 2013). The level of household deposits actually fell by €1.2 billion (-0.7%). While this was not the first time household deposits declined slightly, Bank of Greece Governor Giorgos Provopoulos indicated in April that he expected to see a decline in deposit balances for that month.
The situation in Greece was also complicated by the inability of National Bank and Eurobank to raise 10% private capital in an expected consolidation/recapitalization. That set off a wave of heavy selling in the Greek banking sector on April 8, incidentally right around the same timeframe as the crash in gold.
Despite appearances otherwise, we still have good evidence that liquidity in Europe is far from ideal. Instead, if we piece together various snapshots into this opaque and esoteric construction, we see that the repeated assurances that conditions are on the mend are not really matched by the actual liquidity environment.