European bank capital structurers have noticed a discrepancy in the pricing of bail-in debt that could be putting investors at risk. And with UK ringfencing rules set to exacerbate the differences in bank holding structures, lawyers are calling for asset managers to better understand ownership issues before a downturn.
Market participants told IFLR Practice Insight that senior unsecured debt issued by the holding company (holdco) of a bank is pricing cheaper than tier two notes issued by its operating company (opco), or subsidiary.
The holdco, and as such its creditors, is entirely dependent on dividend upstreams from the subsidiary. But in a bail-in scenario, all of the subsidiary’s creditors – including tier two – are of course ahead in the queue of any equity, including those dividend streams to the parent.
Lawyers are concerned that some asset managers do not understand the difference between a holdco and an opco, and are loading up on what looks like senior unsecured debt without recognising the risks – or that there should be a subordination premium attached. They think a fairly significant repricing event could be ahead.
According to Gerald Podobnik, co-head of FIG origination at Deutsche Bank, there had initially been a ‘healthy’ pricing differential between holdco and opco debt – but that has narrowed now. “It’s been good banks that issue out of holding companies, certainly,” he said, adding that the gap is even wider in the US where, as a result of local bank regulation, far more holdcos exist.
- According to sources, the market is pricing structural subordination;
- Some are concerned asset managers do not understand the difference between structural and contractual subordination;
- They argue that bank holdco debt is by nature riskier than opco debt as the holdco is entirely dependent on upstreams from the opco, which cannot be paid until the opco’s creditors have been paid;
- At its most extreme the impact is the misallocation of risk by portfolio managers, though the consequences will not be known until a bail-in scenario arises;
- Others believe it’s down to current market conditions in which all investors are hunting for yield.
That’s down to a US rule, in place since 2014, requiring foreign banks to establish an intermediate holdco on US soil – effectively a ringfence. The European Commission tabled a similar rule last year, a move many view as retaliation.
“But I would argue that holdco debt is a bit riskier than opco debt, as there’s a higher probability of the holdco running into trouble,” added Podobnik. “The opco has more means available to deal with issues, while the holdco is completely dependent on its subsidiaries’ dividend streams.” Holdcos do very little outside of a resolution scenario, aside from hold the equity of their subsidiaries.But not everyone is convinced. The head of capital structuring at another European investment bank, who spoke on the condition of anonymity, disagrees that holdco debt is fundamentally riskier than that sold by its subsidiaries. “We don’t see any meaningful, discernible difference in the risk,” he said. “The reliance-on-the-opco argument doesn’t really work either. As a creditor you are equally reliant on either legal entity.”
Another hybrid capital structurer at a European bank thinks there is little difference. “On the margin perhaps slightly, but if the bank is in trouble and is going to be bailed in then you’re not exactly going to explain the difference between a holdco and opco to a deposit-holder – it’s a theoretical discussion,” he said. “Failing is failing.”
But he thinks that could change, for UK banks at least, post-ringfencing. For banks with a significant portion of their business within the ringfence, the regulator is keen to recognise losses at the earliest possible point to protect the rest of the bank – so dividend upstreams to the holdco are more limited.
Ignorance vs. appetite
“The market may not have a full understanding of the subordination implicit in ownership structures and exactly how losses are borne,” said Knox McIlwain, counsel at Cleary Gottlieb Steen & Hamilton in London. “Regulators would prefer the market to be buying the senior unsecured notes of the parent on the assumption that they will bear the losses of the whole group, and it's not clear that’s the case yet.”
He thinks that incoming new disclosure requirements under the Financial Stability Board’s total loss-absorbing capacity (TLAC) rule may help investors better understand these issues. But that’s a while off yet; TLAC will not be fully implemented in the EU until 2019.
Instead the onus is on the market to efficiently price the instrument’s risk – which is not happening. Structurers and analysts that spoke to IFLR Practice Insight put this down to the ongoing hunt for yield.
“I think most specialist fund managers do understand the difference between old-style senior opco debt and new holdco debt in this new resolution world we live in,” one London-based senior analyst said. “It’s just the search for yield that makes financials look cheap in relative terms versus investment-grade corporates.”
It could be down to ratings. As Bridget Gandy, co-head of EMEA financial institutions at the agency explained, Fitch rates banks and instruments primarily on the likelihood of repayment rather than where those notes will sit in the pecking order if things go wrong. That approach differs from investors of course, but also from other rating agencies, which tend to draw a bigger distinction between holdco and opco-sold debt.
“F0r it to make any difference which legal entity’s debt you hold, you’d need to be in a resolution scenario,” she said. “Then it matters, as if the opco has obligations to third parties it must honour those before making payments to the holdco. That potentially puts the holdco in some real trouble.”
The potential mislabelling of risk this causes could also be standing in the way of effective portfolio management. Generally asset managers will have much bigger buckets for senior unsecured instruments than for subordinated debt. “But if you’re packaging a higher-risk subordinated credit in a senior unsecured skin, you’re just injecting risk into that smaller bucket that really shouldn’t be there,” said a lawyer based in London who wished to remain anonymous.
Pricing discrepancies are not much of an issue when times are good – many would argue that if that’s what investors are willing to pay, then market price is correct.
“It might only be when we see a holdco structure collapse and structural subordination being tested in practice that there could be some form of significant repricing,” said Deutsche’s Podobnik. The same goes for UK ringfencing, which doesn’t take full effect until January 2019.
And the longer-term implications could be more serious than asset managers incurring losses.
“Permitting this accumulation of risk that can result in surprising losses is essentially what we saw in the financial crisis,” said the lawyer. “It leads not only to misallocation, but also to people holding that much risk that really shouldn’t from a societal standpoint – pension funds for instance.”
He puts the mismatch down to the fundamental shift in the way banks have been restructured since the crisis. When Lehman Brothers failed it broke apart into multiple local components, with the senior unsecured bondholders of Lehman Holdings dependent on the liabilities owed to it by its subsidiaries. But that’s not the case anymore; the theory now is that no one but the equity and debtholders of the parent company will bear losses.
“We don’t even see this when multinational corporates enter insolvency,” said the lawyer. “But if regulators want bondholders and swaps counterparties to base their behaviour on the single point of entry structure, they need to tell them about it.”
Crédit Agricole sold Europe’s first senior non-preferred instrument last December, while Bank of America sold the first TLAC bond into the continent in February this year after the Federal Reserve issued its final rule.
To take a free trial of Practice Insight and read unlimited articles like this, click here.