Via Peter Tchir of TF Market Advisors [8],
Now that the hype of LTRO is over (for now) people are starting to focus on the details and some of the potential consequences. This is a first cut based on bits and pieces from various LTRO documents released by the ECB. I haven’t seen anything that resembles a document fully describing the current LTRO’s, but am trying to find it, and will refine this analysis as more details come to light.
Repayable in 1 year
This isn’t a big deal, but helps explain why banks would have taken money to create a safety net for themselves. The loans are 3 years in maturity, but are repayable in a year. That prepayment right is an important factor in why the demand was so high. The LTRO is actually a floating rate loan – it is not fixed. It will move roughly in line with the ECB’s short term rate. Right now that is at 1%, but in theory could be increased. Banks that were borrowing “preemptively” under the facility will value the 1 year prepayment rights. If the loans became expensive as rates rose, and they didn’t need the money, they could just repay the loans.
Floating Rate
The loans have a 3 year maturity, but it seems as though the rate paid will be reset periodically in a way that should track the ECB overnight rate. That isn’t exactly how it is described, but seems to be the jist of it. With the Fed on hold until 2014 and the ECB under Draghi much more accommodative, it seems likely the rate will remain low, but it isn’t guaranteed. That puts a slight damper on the “carry trade” enthusiasts. Italian 2 year bonds yielding 1.78% aren’t that appealing, especially if the funding cost can increase.
Variation Margin
Zerohedge pointed out a spike in additional collateral being posted at the ECB [9]. According to some documents, the ECB is required to impose variation margins on its financing operations. This means that the collateral posted is not a one-time deal. If the collateral a bank has posted declines in value, the banks would have to post additional collateral. This is a big deal. Somehow the world seems to have an image that banks can borrow 3 year money at 1%, pledge an asset against it, and let the carry take effect with no other consequences. That is far from the truth if variation margins are being used.
Having to post variation changes the product a lot. Buying longer dated bonds becomes very risky. They remain volatile and although banks could hold them in non mark to market books to avoid that volatility hitting their P&L, it wouldn’t save them from posting variation margin if the holdings decline in value. That helps explain why the curves are so steep, and really will limit the ability of banks to hold down longer term yields if we get another round of weakness, the death spiral risk is too scary.
Portuguese banks should be of particular concern – again. The 2 year Portuguese bonds have jumped from a price in the low 80’s to the low 90’s. If banks bought these bonds as LTRO the potential for death spirals is on. As the bonds start declining in value, the banks would have to post collateral. Since the Portuguese banks are surviving almost exclusively on central bank money, their only choice would be to pledge some unpledged assets (if they have any), or sell the bonds and try and repay some of the LTRO. Selling bonds would put additional pressure on a then weak market. So the banks will pledge more assets. This does nothing to stop the slide in the underlying bonds, but would subordinate senior unsecured debt holders further. Senior unsecured debtholders will run for the hills again. They will see assets being taken out of the general pool – where they have a claim – and get shifted to the ECB, where ECB has the first rights.
Like anything else, once this becomes a concern in Portugal, the contagion fear is likely to raise it’s head. With €1 trillion of assets part of the LTRO program, even a 2% decline in assets pledged, would require banks to pledge 20 billion of additional collateral. Italian 2 year bonds have jumped 10 points since LTRO. Is that sustainable? Is there no risk they drift down again? Variation margin is leverage at the extreme. It creates risk to the mark to market of the underlying assets, and makes the “carry trade” option far less interesting, or more scary for any institution that has prudent risk management. Ah, yes, that explains why LTRO dependent banks and those most interested in playing the “carry” game are trading weaker than their peers – they are demonstrating that they are not prudent.