The need to offer liquidity may have given rise to the so-called “head-to-head” wholesale market, but detractors say it is antiquated and that liquidity would be better served by improving the broker market. Furthermore, with borrowers reluctant to tap existing issues, benchmarks will become more susceptible to being squeezed. William Thornhill explores the liquidity conundrum.
The covered bond sector is the only remaining bond market to maintain what some would say is an anachronistic system of market making – more colloquially known as "head-to-head" or "knock-for-knock" – at a time when all other markets have matured into a so-called "order-driven" system.
"The covered bond sector follows the principles of supranational/agency/sovereigns – book-built, but they still need to get up the curve in terms of secondary market. Head-to-head is like the Dark Ages . . . all other markets stopped some time ago,” summarised a seasoned covered bond banker. “I'm sure every firm would tell you that having a few guys sitting there trying to avoid getting picked off all day is a ridiculously antiquated system.”
In the US there are a few examples of so-called “forced liquidity”, that is, where there is a market making obligation. The European covered bond model therefore ensures a two-way price can be found for any given issue even if it may not innately have liquidity.
"To have liquidity you don't need head-to-head – you need a proper broker market. There is plenty of liquidity in names like BNG, Rentenbank, KfW, EIB, L-Bank, ICO, Eksportfinans without head-to-head, so why does the covered bond sector need it?" added the same banker.
Furthermore, "All you have is small regional brokers playing at it. There should be prices in the brokers as soon as a deal is free to trade, but frequently this is not the case."
Detractors say, head-to-head exaggerates volatility because of its price opaqueness. "Market makers can easily delay by giving a spurious excuse or they could be part of a price cartel, but in the broker and electronic markets, the price you see is the price you get – and it's immediate."
Inevitably, profits that are made by one bank closer to the information loop are likely to be at the expense of another outside the loop, and this provides the key incentive for banks to form price cartels. "Frankly, it’s natural to form cartels – it’s war out there so price information sharing about what's going to happen can really help," summed up one player.
The question of liquidity is indeed a conundrum. Investors, for their part, prioritise not only a bond's performance but also where they can find a bid after launch, be it for one year or 10 years later.
DePfa's Julia Hoggett explained, "DePfa takes the view that responsibility for generating liquidity sits as much with the issuer as it does with lead managers. That’s why we have a minimum commitment to €3bn, rather than saying ‘let's reach €1bn and let other people make liquidity in the deal.’ We look at it the opposite way."
DePfa is the only jumbo covered bond issuer that does not contractually oblige its arrangers to enter into a formal market making commitment. Instead, it realises liquidity in much the same way as the rest of the agency sector – principally by ensuring that its bonds are eligible to trade on MTS. DePfa's view on what constitutes liquidity is therefore somewhat unique.
Traders feel caught between a rock and a hard place. "You have to quote, otherwise MTS [electronic platform] will tell the issuer – but in quoting we must decide whether to stop/loss or go short. Many issuers are reluctant to tap, so squeezes are just going to get more and more frequent."
While the absence of a repurchase facility usually signals the death-knell of a bond, it can become easily squeezed well before then, at which time price transparency becomes even more compelling.
"On the electronic platform, at least you know where the price is going to open up – so in a way it brings discipline and transparency. It is disingenuous to say you are not quoting smaller deals on the electronic platform since you have to make a market head-to-head for those deals anyway. After all, it’s cheaper to take a hit on €10m via EurocreditMTS rather than in €15m via head-to-head," explained SG's Lorenz Altenburg.
Despite negativity surrounding head-to-head, there are still significant advantages that should not be overlooked. DrKW's Kai Walter, managing director responsible for euro covered bond trading, made the case:
"The covered bond market has experienced some difficult times such as in 2002 when spreads were volatile. Head-to-head market making was key to getting business back to normal as quickly as possible. Towards the end of April this year, spreads widened sharply, MTS quotes were pulled and liquidity in the electronic market was non-existent. But you could still find two-way quotes in head-to-head with established houses. I can see how some traders need price guidance and an option to quickly square positions, but we should not use this to force all the liquidity into the broker markets."
The requirement to make two-way prices is all very well if the product trades like a commodity, boasting both depth and liquidity at all times. However, this is not always the case in any market – let alone covered bonds, even for seemingly liquid benchmark sized deals of €3bn.
"Why should we be obliged to quote a two-way price that is tighter than in the government sector in squeezed issues that clients are not even prepared to repo back into the market?" exclaimed one trader. He pointed out that bigger banks with large retail networks are often more capable of sourcing bonds from their client network but others do not have this luxury.
Typically, squeezes develop because the free float reduces over time and the proportion owned by “buy-and-hold” investors tends to increase. At some point a large customer order comes through for an older issue that is mostly tied up with investors who may hold until maturity. In the absence of genuine depth, the bond then becomes squeezed relatively easily.
DePfa's Hoggett illuminated that "a big risk for lead managers is not size but rather a function of where paper has been placed. For example, the EIB 2008s got on to the buy list of retail and never got off, so the actual free-float went down enormously. A market maker that is asked to offer a few hundred thousand of a €3bn-sized deal is likely to feel comfortable until it realises that the actual float is closer to €500m".
The size that needed to ensure liquidity differs over time. This is illustrated by the fact that KfW reduced the size of its euro benchmark from €5bn three years ago to €3bn today in order to maximise performance. MTS also reduced the qualifying size of covered bonds eligible to trade on its platform from €3bn to €2bn –and €1.5bn for the cedulas-only platform.
"There is no easy answer to what is the optimal size. Some issuers will only have €1bn to do and they won't raise more because of external reasons. Investors may charge a bit for that but not much. I don’t think there is a charge from €2bn to €1bn, therefore investors have comfort that a €1bn issue will deliver liquidity, but at the end of the day, the market will define the price for liquidity," Hoggett explained.
“There is a spread to pay to issue in size from €3bn up to €5bn, but not much of a charge from €3bn down to €1bn. Issuers are not being paid for providing liquidity, therefore rationally they are not going to do it – and investors are not demanding spread for issues that are likely to become less liquid."
So a development that may need to get underway is whether market maker obligations should in some way be linked to when the issue was first launched. "A €3bn 10-year which is now a six-year is very different from €3bn five-year that is still a five-year."
Customer reluctance to switch out of older, usually more expensive, issues further compounds the liquidity problem, raising the probability of a squeeze. The absence of switching interest has lately been owed to the dearth of relative value opportunities. This is because the drive for yield has resulted in a sustained credit curve-flattening, so narrowing credit arbitrage possibilities.
The cost of market making
While bankers are anxious to illustrate market-making capability, there is a fear of a growing dichotomy between what is signed up to at an institutional level and what is delivered at the inter-dealer, wholesale level. Earning fees over the near-term clearly needs to be counterbalanced with the longer-term costs of making a market. With trading desks in some of the bigger firms barely breaking even, the trading viability of smaller firms is questioned.
"If a bank feels it is not getting rewarded, then there is a chance that the issuer does not get such a great market making commitment from its banks. Banks are charging internally for the market making commitment and they need recognition for that," said one banker. "The whole debate about liquidity harks back to what Citigroup did in August 2004. That trade made a statement which was ‘let's have a debate about what market maker obligations really mean’. This hasn't happened yet."
Some bankers believe the cost of making a market in any particular issue could amount to several hundred thousand dollars over the life of an issue. Their obligation is therefore essentially a written option.
Typically, US investment houses run very lean operations, quoting several hundred issues in the covered bond and SSA sector. In contrast, the big European houses typically have several traders quoting only covered bonds, with each individual responsible for different maturity buckets.
Clearly, the cost of operating in the latter environment is higher, but then it is argued that clients are better serviced – especially in illiquid lines. Whether this increased service is of notable benefit to the client and produces more enhanced profitability is a moot point.
"The truth is that head-to-head market makers are afraid of losing their jobs," said an experienced market figure. "They have often told me that without head-to-head they would have to cull their trading desks. It also justifies why trading desks lose money – ‘it's not my fault, it's due to market making obligations’, they argue.”
But the heavily manned European firms, which are generally considered the chief proponents of head-to-head, say they make good margins. The inconsistency between different camps on their respective ability to produce a viable business may reflect the essence of how they measure profitability.
If measured on a cost-of-capital basis – where the cost of running an entire operation is considered – the return on capital risks being less compelling than the more lightweight operation. With each desk thought to cost in the region of US$2m per annum, this is understandable. If, however, the same firm sums up the profit each individual market maker earns, then a seemingly viable overall margin is arrived at.
Aside from the profitability and cost of capital employed, news of a bank downsizing its operation probably sends out the wrong signal. "If a bank were to fire its traders, news would quickly get back to issuers – they would lose face and it would give the impression that they aren't capable of making a market," a banker noted.
The market structure
While most trading in spread product is hedged versus swaps, it is by no means the perfect hedge – indeed nothing is perfect if the underlying instrument is illiquid. Again, a contrast with the US is instructive. In the US all the borrowers occupying this space (eg the Treasury, Fannie Mae, Freddie Mac) have large funding needs so they all generate natural liquidity.
However, in Europe there are a very large number of agency and quasi-agency issuers in the Triple A space. Although they theoretically operate as a swap proxy they do not borrow on a consolidated basis – and therefore do not naturally generate liquidity in the same way.
Triple A is an asset class that investors would normally wish to have as liquid, since they are the backstop in the event of a volatile market or can be used if investors wish to take a view on the curve, for example.
Over time it might therefore be assumed that Europe's more fragmented banking system will consolidate, since mortgage markets are unlikely to escape the impact of globalisation.
In this sense the biggest development would be to see, say, a domestic lender borrowing against a non-domestic asset, eg a German borrower whose housing assets might include a basket of regions throughout Europe.
The process by which this might come about could be a function of increasing EC legislative developments, cross-border bank mergers and a more developed third party servicing industry.
The structure of much of the European housing market is fundamentally different to that of the US. The Irish, Spanish and UK sectors have high home ownership and are similar to the US, and their respective governments cannot realistically allow the mortgage market to fail as it is a cornerstone to their economies – but this is less relevant in the rest of Europe.
"Liquidity will be generated by greater consolidation but consolidation probably means increased home ownership. There is only domestic asset origination on the part of mortgage providers, which means they have got to be comfortable lending across regions. That means a greater consolidation of the nature of mortgages across Europe and probably means a change in the culture of mortgage provision," suggested a leading banker.