Via Peter Tchir, of TF Market Advisors,
So the market has completely latched on to the idea that LTRO is back-door QE.
Does this make any sense and can it even work?
So banks can borrow money for up to 3 years from the ECB. They can buy sovereign bonds with that money. Those bonds would be posted as collateral at the ECB.
The bull case would have banks buying lots of European Sovereign Debt with this program.
The purchases would be focused on Italian and Spanish bonds with maturities less than 3 years.
Buying bonds with a maturity longer than the repo facility is risky. The banks would need to be assured they can roll the debt at the end of the repo period. Some may be convinced, but the bulk of the purchases will be 3 years and in so that they loans can be repaid with the redemption proceeds.
Looking at 2 year yields, Italy and Spain offer some value, with yields of 3.2% and 4.8% respectively. Belgium at 2.4% may attract some interest. France is too tight at 0.9% to see its bonds purchased as part of the LTRO.
Ireland and Portugal yield 7.6% and 13.9% respectively. The yields are attractive, but the risk of default is too high in my opinion for even an aggressive bank to play around with. Ireland has mentioned the possibility of needing to renegotiate their package. Any bank that decides to load up on bonds of Portugal or Ireland would risk backlash from even optimistic investors.
So realistically, the bulls would hope for banks to load up on short dated Spanish and Italian bonds. Will LTRO really encourage so many new purchases? From a funding standpoint, these bonds were already relatively easy to finance. There is a developed and reasonably efficient repo market for these bonds. It would be difficult to get term financing, but was relatively easy to roll the repos. The ability to get the term financing makes it easier for a bank to put the trade on, but it is an incremental improvement, rather than a dramatic change.
So banks buy the bonds and earn the carry and all is good? Not so fast.
New Purchases vs Existing Positions
Banks are struggling to borrow money right now to finance their existing positions. How much of LTRO will be used to finance new bond purchases, rather than to replace existing forms of funding? Any bank that is already running a big sovereign debt position will look to LTRO to replace existing forms of financing. They can eliminate the repo roll risk on bonds they are financing in the repo market, or they could stop attempting to borrow in the interbank market. Those are positives for the banks as they can earn more carry (cheaper financing) and reduce roll risk (3 year term). But that doesn’t create new demand for bonds.
So the LTRO can help the banks with their existing funding problems without a doubt, but it is unclear that encourages new bond purchases.
How much can a bank do?
Let’s say a bank thinks this is the greatest idea ever. Could a bank do this with 100% of their assets. Yes, that is extreme, but it highlights the issues. A bank puts 100% of its assets into LTRO trades. What price would you pay for the equity? If all the bank had on its books was Italian bonds funded to maturity, how much is the equity worth? In MF Global’s case, it was worth ZERO, but there were some concerns about MF’s ability to continue to roll the position. The equity would be a huge leveraged bet on an Italian default. It would be binary. The income is leverage * net spread. The equity would be wiped out if there was a default. Banks will be tying up some amount of equity in the trade (LTRO shouldn’t be providing 100% financing, there is usually a haircut of some sort).
Would private equity gravitate towards banks making that bet, or would it be pulled. What about private creditors? Would you lend to a bank, when the bulk of their book is based on the theory that Italy and Spain won’t default. Here it is important to remember, that most banks are already overexposed to these bonds under the assumption they wouldn’t default. Banks have not accumulated these positions under the assumption that default was a possibility, that fear is relatively new. Will banks that are underweight Spanish and Italian bonds really decide now is the time to buy?
To buy now, you need to believe that the default risk is gone. Since NOTHING about this program addresses solvency, you cannot change your default assumptions. You would be betting that the problem is really liquidity driven and that this program can solve that. But how can you know that? You need to assume every other bank will add significantly to their exposure. No one bank can grow their exposure too large, without losing all access to the public debt markets and seeing their share price drop. So each bank can only add incrementally. Since the solvency problem hasn’t been addressed, you are buying in the hopes that some other bank buys too. If everyone buys and takes on even greater exposure to these weak countries, then the liquidity and debt issuance risk can be addressed. But what if strong banks don’t think it is smart to take on more risk.
Why would a bank that is afraid of default take on more risk? If you have largely avoided the problem, why would you participate now? You have been involved in government led negotiations on Greek debt (or at least you know people who have been). You are afraid that Ireland and Portugal will demand concessions. Will you really take quasi government money to buy government debt? The solvency hasn’t been addressed and you are exposing yourself to political manipulation. Is some carry really worth it? Will you really decide that you can’t find a better investment opportunity (like buying back your own debt or shares?).
Adding to your sovereign debt positions means you really believe there is no risk of default, and that using the program, won’t lead to renewed pressure on your business from politicians. I don’t see that happening. MF Global, although different, is also too fresh in everyone’s mind. If a bet becomes disproportionately large, it can kill you. Not only lenders may flee, but counterparties may also flee. It is possible that firms will choose not to do business with you. There is no shortage of counterparties that clients can deal with. Why do business with a bank that seems to be putting on risk that leads to a binary outcome?
Even without default risk, will there be mark to market margin requirements? Although the bonds are short dated, they could sell off again. If they sell off, you have to come up with margin for the repo agreement. Where will you get that? Probably from the ECB by posting more assets, but that is a risk. The mark to market risk and variation margin doesn’t go away (unless the ECB is providing all this money without variation margin). In a totally circular world, the more banks that participate in size in the program, the less likely that you will have mark to market margin calls, but if only a limited number of people participate, then the risk is high and you will be isolated by the market.
So it is a weird kind of Prisoner’s Dilemma. You want to act last. You want to make sure other banks are participating before committing yourself. By waiting, you would risk earning less carry, as the bonds you purchase will have increased in price, but that impact is minimal. If the leverage is high enough and the cost of funds is low enough, than that opportunity cost has minimal impact on your decision. If you load up on bonds and find that no other bank has participated, then you will stick out like a sore thumb. Investors won’t trust you, and you will likely have a mark to market loss, and will be on the front page of papers questioning why you participated and others didn’t. So it is smart to wait and see what others do. If everyone is waiting to see what someone else will do, we will likely see minimal participation. The weak banks will add to their all-in trade, but the stronger banks will focus on other business lines that can be profitable without so much intervention or risk.
So what will happen?
There will be significant interest in tapping the LTRO for existing positions. Some small amount of incremental purchases may occur at the time, but the banks will use this to finance existing positions. This should help bank credit spreads. It should also show up in measurements like OIS as it would reduce pressure in the interbank funding market. This is positive, but a relatively minor positive, and seems more than priced in.
We will likely see the yield curve in Spain and Italy steepen sharply at the 3 year point. Buying longer dated Italian paper will remain very risky and the banks will be more interested in retaining positions 3 years and in.
I think we have already seen the initial impact. Now we will wait to see rates do well, but will be disappointed. The big banks with risk management departments will decide to decline. The risk/reward just won’t be attractive to them. We will find out that places like DB don’t participate and that small weak banks do. That will actually start another spiral on those weak banks, as people will sell the shares and they won’t find lenders outside of the ECB as no one will trust their discipline. In the end, this won’t do much for the sovereign debt market, but will shine a spotlight on which banks should be shorted and will possibly expedite their default.
I’m still nervous that the IMF or China or someone has a real program, but up 35 points on SPX from yesterday on hope of LTRO seems overdone – again.