Deutsche Bank Is Scaring Bondholders for a Good Cause

A neat trick is for bank capital to be (1) loss-absorbing and (2) risk-averse. Deutsche bank's cocos sort of are.
Not DB's junior bondholders. Photographer: Leigh Vogel/Getty Images for Knowledge Universe

One problem with having banks answer to their shareholders is that bank shareholders are jerks. Their money represents, I don't know, 3 percent of the bank's assets; the rest comes from creditors of various flavors. So the shareholders are very keenly aware that their shares represent an almost at-the-money option on the bank's assets. The value of an option increases with volatility. So bank shareholders like volatility, or they should anyway; some risks are too much even for bank shareholders, so you occasionally see shareholders suggesting that bankers dial it back a bit.

If you are less inclined to enjoy bank volatility, you might work on a theory where banks should answer to the creditors who fund the remaining 97 percent of their assets. 1 This turns out to be hard, though, because bank creditors tend not to be the types of people who want to be answered to. Retail depositors, for instance, are not interested in reading a bank's financial statements before opening a checking account. And lots of sophisticated debt is like this too; repo lenders are no more interested in reading financials than depositors are.

So you're left with pretty much regulators to sort of represent the interests of creditors, which is not a perfect solution either.

A theory you sometimes hear is that banks should be monitored by some tranche of creditors whose risk appetite lies somewhere between the terrifying shareholders and the terrified depositors: that bail-in-able debt, say, will incentivize 2 debt investors to monitor banks and get them to be less risky.

Tracy Alloway has a really interesting post at FT Alphaville about Deutsche Bank's coco offering. We talked a little about these cocos -- contingent capital notes, or Additional Tier 1 notes, or capital notes, or whatever -- the other day; Deutsche actually launched them a while back but put them on hold to announce its much larger equity offering. (It's raising some 8 billion euros in common equity from a Qatari investment and a rights offering; the cocos will raise around $4.5 billion in various currencies.)

What explains the delay? Alloway quotes Marc Holman at TwentyFour Asset Management:

When we held one-to-one talks with DB senior management on 9th May (the last day of the roadshow) we expressed our concerns that capital ratios and leverage were weak compared to other peers in the European banking arena. We were obviously not alone in this train of thought with a number of market commentators expressing similar views, and this must have resonated with senior DB management.

That might be wishful thinking about the influence of coco investors, but then again it might not be. The sequence might literally have been:

  1. Bank tries to raise junior debt. 3
  2. Junior debt investors say, no, you need to fix yourself and raise equity first, so that you're safer and less volatile.

Alloway also quotes from a May 16 Bank of America research report about Deutsche's cocos that pretty much worries itself sick about Deutsche's ability to pay the coupons on the cocos. The issue there is that:

  1. Deutsche Bank (a) doesn't have to pay coupons if it doesn't feel like it 4 and (b) can't pay coupons if it can't meet certain capital requirements;
  2. Deutsche Bank is perilously close to flunking those capital requirements; 5 and
  3. Deutsche Bank's history of being nice to junior bondholders is only so-so.

Bank of America, in its capacity as a credit research provider, is horrified:

DB’s new AT1s would add an additional level of complexity to the issue of whether or not AT1 investors receive their coupons, we think. In some ways, if these trades ever print, we’ve crossed the Rubicon -- and not in a good way. So, we are writing about this transaction because as we were considering the format of the transaction, we became aware that it was raising a number of issues of wider relevance to the AT1 market (and of course to DB as a credit) that have received less focus in prior transactions than perhaps they should.

And if you are contemplating an investment in these cocos, you too should be carefully measuring the water temperature in this particular Rubicon.

But for the rest of us this all seems great, no? The big problem with using junior debt securities to meet capital requirements is that investors have pesky expectations of being paid back: If everyone buys these things for the yield, expecting that they'll always pay coupons and never default, then turning off the coupons can itself trigger a catastrophe. Capital needs to be loss-absorbing; if you're afraid to stop paying coupons on the thing you've labeled "capital," then it doesn't really work to insulate you against loss. It's just debt that you've snuck by regulators by pretending it's equity.

But if there's a real, well-understood risk that you could actually stop paying back your "capital" securities -- and that that might happen well before you run into any trouble, while you're still a (quite levered) going concern -- then they do serve as loss-absorbing equity. More than that, though: They serve as risk-averse equity. Unlike your shareholders, 6 the investors in the coco bonds will demand that you raise more capital, diluting shareholders but making you less risky. That's a pretty good service for bank capital suppliers to provide.

This column does not necessarily reflect the opinion of Bloomberg View's editorial board or Bloomberg LP, its owners and investors.
  1. Really I'm just kidding about the 3 percent/97 percent thing; I understand that it's just a measurement issue. But 3 percent is what Deutsche Bank's balance sheet says.

  2. Sorry Krugman. This is a useful word.

  3. Here I use "junior debt" or "junior bond" to mean stuff like this -- cocos, capital notes, preferred stock, junior sub debt, whatever. It's all done to meet regulatory capital requirements, it's longer-term and junior to real debt, but it's more senior and fixed-income-y than common equity.

  4. I know, that sounds silly, but it's true: "Interest payments will not be made, if the Bank elects to cancel the payment, in whole or in part, at its sole discretion."

  5. As BofA evocatively puts it, they're one "reasonably robust regulatory fine" away from flunking some of them. (That's before this week's capital-raising news though.)

  6. I'm speaking only very loosely of Deutsche's shareholders, who seem supportive of dilution.

To contact the author on this story:
Matthew S Levine at

To contact the editor on this story:
Toby Harshaw at

Before it's here, it's on the Bloomberg Terminal.