Tuesday, December 20, 2011

A Risk Assessment on ECB's LTRO


The FT has already reported on how hesitant banks are about buying ever more sovereign debt. In fact they outright dumped  €65bn of bonds in just nine months. Hopes that banks would hold the hand of the sovereigns that back them continue to dim, as the Sarko carry-trade looks increasingly less likely in advance of this Wednesday’s offer of cheap 3-year ECB financing.
The presumption that banks are going to use the 3-year Long Term Refinancing Operation (LTRO) to buy sovereign bonds comes not just from the dreams of certain politicians, but also from the observation that yields at the short end of peripheral curves have come in dramatically.
Spanish bonds provide an example (chart courtesy of SocGen):

From the above, European financials have deteriorated over the last week while the yields on Spain’s government bonds have been coming in. Is this the result of banks buying up the high yielding bonds that they will soon be able to fund exceptionally cheaply?
Not so much, say the analysts at SocGen in their Rates Strategy daily this Monday. There are many factors at play, and true, one of them may be the anticipation of banks putting on carry trades, but the expectations may not transform into reality.
For one thing, banks are going to have to find a way to fund their existing asset holdings — to the extent that they don’t deleverage themselves into nothingness, that is — and a good portion of the current funding for them will roll off in 2012. SocGen points out that for eurozone banks in 2012, €250bn of senior unsecured bank bonds will mature, along with €83bn of government guaranteed debt, plus €19bn of subordinated debt.
Seeing as the unsecured market is somewhere between frozen and inaccessibly expensive, the most relevant candidate for the replacement of that debt is reckoned to be around €185bn of covered bond issuance, a figure which the analysts acknowledge may well be a bit on the high side (though at least it will be supported by another ECB programme to specifically prop up that market).
The rest of the funding needs to come from somewhere. And, well, the ECB is offering…
True, the ECB ties up collateral as equally as covered bonds do, but there is an extra attraction to the LTRO: the banks that take the 3-year funding will in fact have the option to repay any part of it after just a year, hence freeing up the collateral held against the borrowing at the ECB. Nice option… that isn’t too consistent with the whole “carry trade” concept where the maturity of the asset is matched to the term of the funding for it, the rates team at SocGen points out.
Oh, and the collateral posted to the ECB can be relatively low quality. Not like the stuff required for private markets, or for covered bond pools.
One thing that actually joins the LTRO on the supply-side for liquidity, according to SocGen, is the lower reserve requirements that will kick in for the maintenance period starting on January 18th. Falling from two per cent to one per cent will free up some €100bn that was on deposit with the ECB — something that will happen in advance of the second 3-year LTRO at the end of February. However, the SocGen analysts expect that this move will more likely lead to a decrease in weekly main-refinancing operations (MROs), than a decrease in LTRO demand. One to be aware of, anyhow.
But back to how unlikely carry trades are:

There are several obstacles to carry operations, namely the stricter capital requirements, the pressure on banks to deleverage; and the stigma attached to such trades if ever revealed.
Put even more simply, do you think a bank that shows an increase in sovereign bond holdings in their quarterly reports will find it easier or harder to fund itself in private markets?
And what if there is yet another EBA stress test that whacks sovereign holdings and then demands additional capital for potential losses? How clever will a sovereign carry-trading bank look then?
In addition to that, if the bonds were to reverse course and start tanking again, the banks would have to post additional margin on them.
All of that said, could the banks make a dent if they wanted to? Out of some sense of patriotism perhaps? Emphasis ours:
Euro area banks have some 6% of total assets in government bonds (with ratios slightly higher at 7% to 9% in Spain and Italy as per the most recent EBA data). If half of all Spanish and Italian banks (it is unlikely to be the larger names) were to raise the ratio by 1% next year, that would lead them to buy some €8bn-€10bn in each country. Most likely the impact would be far less, and graduated over time. Buying though on such a scale is modest as a percentage of total issuance (some 9% in Spain, 4% in Italy).
That’s a “no” with words. Here’s the same with a graph, courtesy of Deutsche Bank with a couple of FT Alphaville modifications:

In the end, SocGen predicts a demand for €200bn on Wednesday.
RBC is in a similar ballpark, but warns that there’s a risk that the uptake could be lower than expected. There are currently already €350bn in excess reserves parked at the ECB which is much higher than they were prior to previous LTROs:

Furthermore, banks can fund the €432bn of tenders that mature this week with the weekly MRO and 3-month LTRO, so they don’t necessarily need to go to the 3-year tender just to keep things constant.
In addition to this, the RBC team notes that the exact details on the expanded range of eligible collateral hasn’t actually been decided yet, so it may not be clear to banks whether they have newly-eligible assets lying around that they may otherwise be willing to post.
In conclusion (emphasis ours):
A not too small outcome should suggest that banks use the new facility and get longer dated funding on board. This should sooth some anxiety about their funding risks going into 2012. A not too large outcome should also suggest that no unreasonable risks have been taken.
It’s the goldilocks of refinancing operations.