Bank of America Merrill Lynch analysts have taken objection to everyone interpreting high use of the ECB’s overnight deposit facility as an indicator of ‘bank hoarding’.

It is just not so, they say.

In normal times, for example, there are three main areas where banks can direct their borrowed cash:

1 – the commercial banking system.

2 – turn it into physical cash.

3 – transfer it to governments via auctions.

In more technical terms, liquidity provided by the ECB takes the form of either estimated minimum reserve requirements or net autonomous factors. To unpack what those two terms mean:

Minimum reserves are the ratio of cash banks are required to maintain at the ECB in order to cover their liabilities in the commercial banking system. If they happen to have more borrowed cash on their balance sheet than minimum requirements (let’s say because they are not lending it) the excess usually ends up deposited in the deposit facility so as to earn some interest. Remember, this is borrowed cash, and banks will want to offset the interest they themselves pay to have access to that liquidity.

Net autonomous factors, meanwhile, are made up of demand for physical cash as well as the cash balances accumulated by governments.

In other words ECB liquidity can either be directed to the deposit facility, physical cash in the market or back to governments.

But as the BoAML analysts note, it may also be the result of lower reserve requirements (their emphasis):

Assuming the liquidity stays within the commercial banking system, the additional cash will only disappear from the deposit facility if banks’ reserve requirements increase. Banks reserve requirements increase if the deposit base in the Euro area increases.

The deposit base would have to increase by a factor of roughly 50 (as reserve requirements are 2% of the deposit base), implying loan growth on the back of the tenders, in the scale of €20trn, in order for the excess reserves on the deposit facility to be turned into required reserves and move to the current account of the ECB’s balance sheet. This is clearly unrealistic given the targeted deleveraging of bank balance sheets in Europe of EUR 1.5 tn.

So industry deleveraging — a.k.a an inability to lend — means banks have much lower reserve requirements than they might otherwise have. Less of the cash they borrow from the market, needs to be held back in minimum reserves, and more of it can be directed to the deposit facility.

Or you could say, banks have more than enough cash to cover their official reserve requirements, but feel they need to hold more liquidity to be able to lend on an interbank basis to quash counterparty fears. In that sense they are also pre-hedging the risk of future funding shortfalls.

But does that mean banks are hoarding?

Well, ECB President Mario Draghi stated at Thursday’s press conference that banks that placed funds at the deposit facility are not necessarily the same ones who borrowed in the three-year LTRO suggests that’s not the case. That money did make its way through the system. It did end up in other places.

Here’s an example of how it might have worked:

The set-up: each of Bank A and Bank B has a €250bn loan book to start with, but A is funded fully by deposits while B is funded fully by wholesale funding. Both banks are subjected to reserve requirements of 2% of their deposits and wholesale funding. The reserves are held at the ECB’s current account (C/A). For simplicity, we will ignore the effects of haircut on collateral.

So both banks are in a situation where they lent €250bn to customers, but bank A used customer deposits to make those loans, while bank B used wholesale funding to make those loans.

In the scenario where bank B loses €50bn of wholesale funding (finds itself short €50bn because its wholesale financing opportunities have expired or coming due) it has the opportunity to replace that €50bn with ECB repo, using the loans it made as collateral.

The ECB is effectively ensuring the bank can meet its wholesale obligations. But since it now owes the money to the ECB, rather than the wholesale market, its reserve requirements are lower. It has to borrow slightly less (€1bn — two per cent of €50bn) from the ECB, than it would do from the market to cover the imbalance.

Now consider that Bank A receives €50bn in deposits from its customers who obtained them from B’s redemptions. That bank’s reserve requirement increases by €1bn (two per cent of 50bn).

Netting it all together, Bank A ends up with €44bn in excess cash and places that at the ECB’s deposit facility.

But here’s another scenario. Presume Bank B didn’t utilise the ECB repo facility because it had an imminent wholesale funding shortfall of €50bn, but instead, because it was pre-hedging the risk of coming up against a €50bn wholesale funding shortfall…

… in which case you might have got a situation like this:

* Bank B
Liabilities: ECB repo increases by €50bn (using the loans as collateral)
Assets: places the proceeds from ECB repo at the deposit facility in anticipation of the redemptions

*Bank A: No change to its balance sheet

*ECB
Liabilities: Deposit facility increases by €50bn
Assets: ECB repos increase by €50bn

While that might look like the definition like hoarding liquidity for a rainy day, BoAML add one more scenario to the mix.

Imagine Bank B pre-hedges the disappearance of wholesale funds, and uses the proceeds to buy government bonds in the secondary market instead:

*Bank B
Liabilities ECB repo increases by €50bn (using the loans/bonds as collateral)
Assets Government bond holdings increased by €50bn (using the proceeds from the repo), before selling them to pay for the redemptions

Bank A
Liabilities: receives €50bn in deposits through its customers who sold the government bonds to Bank B; reduces its ECB repo to 0 with these deposits
Assets: its reserve requirements is increased by €1bn; after netting all the transaction, its places the €44bn excess cash at the deposit facility

ECB Liabilities: Deposit facility increases by €44bn, Current account by €1bn Assets: Repos increase by €45bn

And now same thing but in the primary market:

Bank B
Liabilities: ECB repo increases by €50bn (using the loans as collateral)
Assets: Government bond holdings increased by €50bn (using the proceeds from the repo) through purchases in auctions, before selling them to pay for the redemptions

Bank A: No change to its balance sheet

ECB
Liabilities: Government deposits increase by €50bn
Assets: ECB repos increase by €50bn

And that last scenario is arguably the whole point of the exercise. That is to say, it is just as likely that the money washes through the bond markets (helping out yields) on its way to ECB deposit facilities as that it doesn’t.

In all scenarios, deposits at the ECB rise.

And just to throw in one last scenario… in the event that Bank B taps ECB liquidity due to a €50bn funding shortfall, and Bank A receives the deposits from Bank B’s redemptions and uses them to make new loans to the ‘real economy’, those deposits will end up in the banking system once again too, adding to ECB deposits.

The interesting aspect is thus the relative indifference of ECB deposits. Whether liquidity is distributed into the system to cover genuine funding short-falls or to help banks ‘pre-hedge’ funding short-falls; whether funds end up going through bond markets or not, deposits always rise.

So the question really should be, what would it take to see deposits fall?

Related links:
The collateral crunch gets monetary - FT Alphaville
ECB Overnight Deposits Again Hit Record High – WSJ

This entry was posted by Izabella Kaminska on Friday, January 13th, 2012 at 14:48 and is filed under Capital markets. Tagged with ecb, ECB deposits, liquidity, Mario Draghi. Edit this entry.