Give covered bonds a chance

By Felix Salmon
March 16, 2011
Tim Geithner is throwing his weight behind efforts to build a US market in covered bonds.

" data-share-img="" data-share="twitter,facebook,linkedin,reddit,google" data-share-count="true">

It’s good news that Tim Geithner is throwing his weight behind efforts to build a US market in covered bonds. This made sense back in 2008, when I did my best to explain what exactly covered bonds are, and it makes sense today as well. The more different ways that mortgages can be funded, the less pressure there is on the US government and government-owned agencies like Fannie Mae and Freddie Mac to fund just about everything.

The problem is that the government has lots of different arms, and one such arm — the FDIC — has issues with covered bonds. At the moment, the FDIC can wipe out billions of dollars in unsecured bank debt when it takes over a failing institution, like we saw with WaMu. If banks moved to a covered-bond system, then a lot of bank debt might be secured rather than unsecured, which raises the prospect of bigger losses for the FDIC.

My feeling is that the importance of fixing the broken mortgage-funding system is more urgent, right now, than shoring up the FDIC — especially given that the covered bond market, if and when it emerges, will in the first instance be very small and pose little systemic risk. Covered bonds have an important role to play, here, so let’s at least make them possible. They might well never really take off. But it’s worth a try.


We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see

2 Points:

Anything that Geithner proposes will be a sellout to the big banks. He is a creature of the big banks.

There is not a problem in mortgage funding, there is a problem in the HOUSING MARKET, because housing is still over-valued and still has some downside to go and inventory to be consumes, which means that mortgage
funding is not broken, but in fact working.

I, for one, am not interested in another ill-conceived bank bailout, and if Geithner’s name is on it, then that is what it is.

Posted by Matthew_Saroff | Report as abusive

“Building a market” seems less important to me than getting the related capital regulatory requirements right (though, after you do that, building a market would be of value, too). I don’t know whether there’s any market these days for mortgage securities, but if a bank wanted to give a credit guarantee on the senior tranche of a security of mortgages it had issued, I wouldn’t think that would reduce their appeal.

(In principle, someone else could issue a credit guarantee on any mortgage security; this loses its systemic appeal if the credit risk isn’t being borne, at least in the first pass, by the bank that should in principle have underwritten the mortgage loan in the first place. As of two or three years ago, bank capital rules encouraged banks to get their own mortgages off of their own balance sheet; if that’s still the case, fixing that would be my top priority.)

Posted by dWj | Report as abusive

Thank you for the link to the earlier explanation. It was very helpful. However, something in that article really struck me: “Their main risk is that the bank blows up because the mortgages blow up, and they’ll be left holding a bag of damaged loans – but because two things have to happen rather than just one, that risk is relatively low.”

This statement implicitly assumes independence of the risks: its rater like saying that because loosing off-site power and the backup generators mean two things need to happen, the chance that there will be a total loss of power at the Fukushima reactor is very small. In point of fact, there was a common cause: the earthquake which caused the loss of off-site power also caused the tsunami that knocked out the backup generators.

The same problem is likely to occur here. An economic downturn that might put the bank in trouble, might also put the homeowners whose debt formed the collateral on the bond to be underwater as well.

Common cause failures are a tricky business. I did my Ph.D. work in this area and I don’t have great confidence when building a risk model that I haven’t overlooked some critical common cause failure. I also am not sanguine about the ability of the market to price this kind of risk. After all, this was the problem with the mortgage-backed securities, everybody underestimated the probability of an economic collapse putting most of the mortgages underwater at the same time.

Posted by ralmond | Report as abusive

I think dWj has nailed it. If covered bonds did introduce systemic risk, there’s no benefit to giving them a chance to take off. The riskiness depends more on regulatory treatment than on the instruments themselves. For instance, cross-ownership by banks should be discouraged (and of course, self-ownership via SIVs should be right out.) Therefore they must not be given low risk weights no matter how safe they are. Also, they should not be putable; as a general principle, putable bank debt should attract liquidity penalties.

Posted by Greycap | Report as abusive

Covered bonds are not a simple trade off between the FDIC fund and the future of mortgage finance. The Treasury’s less than thoughtful support speaks to a serious misunderstanding of how to manage systemic risk.

Greycap notes the important, broader point. Simply put, banks should be allowed to issue covered bonds all they want but they should be able to own them.

Especially since we are talking almost exclusively about TBTF banks. Covered bonds are a creature of TBTF.

Why would only TBTF use the covered bond market?

In the U.S., institutionalized, covered bond-like medium-term lending has existed since 1932 in the form of the 12 Federal Home Loan Banks. The FHLB’s function is to lend to banks (now mostly community and regional banks) with collateral in the form of perfected security interests in specific SFR and CRE loans.

In addition, the macroeconomic monetary function of covered bonds, just as in say trust preferred securities, should be to attract new liquidity (or capital) into the banking system not just redistribute monies in the banking system creating a multiplier effect. There already exist a series of time-tested instruments that redistribute liquidity–fed funds, CDS, wholesale loans, etc.

Consequently, if covered bonds are attractive primarily to the largest banks, then allowing them to issue and then allowing any bank to own covered bonds serves only to create additional interconnectedness and additional systemic risk, in effect, re-creating Fannie and Freddie out of the largest TBTF banks.

So sure, give them a chance. But do we really need Fannie and Freddie under another name?

Posted by AABender1 | Report as abusive


Greycap notes the important, broader point. Simply put, banks should be allowed to issue covered bonds all they want but they should NOT be able to own them.

Posted by AABender1 | Report as abusive