ETFs: generation swap

More ETF providers are embracing synthetic replication, with the multi-swap counterparty model hailed as the next stage in product evolution. But to what extent does this already exist? Helen Fowler reports

Their supporters are touting multi-swap products as the next generation for ETFs. Others question whether the products live up to their hype, suggesting the title is just a new way of describing what providers already do. 

Over a year has passed since the first such model was created and yet this divided opinion still exists. 

In 2009, ETF Securities (ETFS) set up what the firm claims is Europe’s first multi-counterparty ‘platform’ model, ETF Exchange. The business offers investors synthetic products backed by swaps from a consortium including Bank of America Merrill Lynch, Barclays Capital, Citi and Rabobank. 

The platform aims to improve on the traditional model involving a single institution as both provider and swap counterparty. “You have to believe it’s stronger than the single bank model,” said Mark Weeks, chief executive officer at ETF Exchange. 

Source is another version of the ‘platform’ approach to ETFs that emerged last year. The exchange-traded product (ETP) provider has teamed up with Bank of America Merrill Lynch (a fixture on both Source and ETF Exchange), Goldman Sachs, Morgan Stanley, Nomura, Credit Suisse and JP Morgan, to offer synthetic ETPs. More partners may follow. 

The theory is that the competition between multiple counterparties to write the swap offers better pricing for the investor, as well as more potential for arbitrage. “Market makers route the order to the bank they feel offers the best prices and the most efficient way of dealing with it,” said Weeks. “There is a competitive element in winning the order. The banks have to go out and win the business.” 

Source chief executive Ted Hood said the multi-swap model meant more competition. “When they buy or sell, most investors like to deal with more than one broker because they like to put them in competition,” said Hood. Theoretically, an investor can buy a Source product from one bank then sell it to another, while checking with a third for a better price. Since it went live with its first products in April last year, Source has gathered over $6bn dollars in assets.

Limiting risk

Proponents also claim the multi-swap approach mitigates counterparty risk. ETF Exchange said if one of the swap counterparties on a platform went bust, another one from the consortium could take their place, all without any disruption for end-investors. 

“The beauty of the third generation model is that if one bank were to default, you would close the position with the defaulting bank and re-open it with one of the other banks,” said Weeks at ETF Exchange. “It allows for continuity of the fund.” ETF Exchange has attracted $400m in assets since its launch.

The multi-swap model may appeal to banks that need an ETF offering, but do not want to create it single-handedly. It is also attractive for start-up ETF distributors that do not have in-house capabilities to write swaps. “It’s a new model that just makes more sense for the institutions that do it,” said Christos Costandinides, ETF strategist at Deutsche Bank. 

Under the multi-swap model, banks get to provide the swap underpinning the ETF, typically the most profitable part of the product, but do not have to tackle the marketing and sales effort involved in issuing them. 

There is controversy over how useful the multi-swap model will prove for investors. Costandinides at Deutsche Bank disagrees with claims that if a provider on a multi-swap ETF defaults, their swaps could be transferred to another member provider, without any disruption to the fund. 

“If it’s a rough market, I’m not sure you can make a call and 15 minutes later you have a new counterparty,” he said. “If a bank defaults, it’s not going to be business as usual. Transferring the swap would depend on what else the other banks had on their balance sheets.”

Richard Yarlott, partner at Integral Asset Management, a firm that advises clients on constructing portfolios using ETFs, is also sceptical about the merits of multiple-swap counterparties. “I’m not sure that adding counterparties solves the problem,” he said. “You want your one counterparty to be reliable.” 

Yarlott doubted how well the multi-swap model might hold up in a real crisis, pointing out that it remains untested by a major default. “I’d like to see how any legal dispute involving multi-swap works out,” he said. 

Similar swap models

Among the biggest criticisms of the multi-swap model is that, when examined more carefully, it is not really all that different to certain types of traditional synthetic replication.

European legislation encourages providers of swap-backed ETFs to search across the market for the best swap issuer. In other words, even the traditional ‘single’ swap providers are more flexible than their title might imply. 

The European Union’s Markets in Financial Instruments Directive (Mifid) asks firms to provide best execution for their clients, meaning they must trade under the best conditions possible. But countries have implemented Mifid differently to each other. For example, neither Luxembourg nor Ireland insist managers follow Mifid’s rules on best execution. In contrast, France takes a rigorous approach.

“A French asset management company that wants to do synthetic replication has to go through a competitive bidding process to select their swap counterparty,” said Isabelle Bourcier, global head of ETFs at Lyxor, the provider owned by Société Générale. 

Bourcier said her firm ran contests to find the best swap provider for its ETFs: “We are not a single swap model here at Lyxor. We act in the best interests of unit-holders when selecting a counterparty, doing so under strict criteria.” The firm does not automatically select its Société Générale parent as a swap provider. Bourcier said: “We run beauty contests throughout the year to select the best swap counterparty.”

Indeed multi-swap may not be the revolution its supporters claim, according to one market participant. He said that ‘traditional’ synthetic replication involved rigorous selection of swap counterparties. 

The process involves reviewing the credit ratings and analysis of potential counterparties to determine their strength. Assessing levels on credit default swaps would also help draw up a shortlist. 

Costandinides at Deutsche said: “There is a lot of noise about multi versus single swap. They are being used as marketing ploys and people are throwing accusations at each other.”

Swap competition

As laws aimed at harmonising European investment rules take broader effect, more firms offering synthetic ETFs will be forced to stage a counterparty selection process as a matter of course. The introduction of the fourth and latest Undertakings for Collective Investments in Transferable Securities (Ucits) law next year will require more funds to prove they have tracked down the best provider available in the market. 

Some funds are temporarily exempt from having to prove best execution. But once almost all providers have to prove it, the appeal of the multi-swap model may diminish. 

Some question the multi-swap’s validity as a business model. “If you are a third party ETF distributor what assurances are you going to have to get the best price and execution from a large investment bank?” said Dan Draper, global head of ETFs at Credit Suisse. 

Draper said: “This competitive dynamic is really important. If you lack pricing power as a third party you are a price taker and may not be able to obtain terms as favourable as larger competitors.” 

Some analysts are not convinced by another key selling point of the multi-swap model: that it reduces counterparty risk. “Moving to a multi-swap model is in many ways not a positive development at all,” said Costandinides at Deutsche. “Multiple counterparties don’t necessarily mean a reduction in risk.”

The multi-swap model could potentially lead to higher levels of counterparty exposure, said Costandinides. Under Ucits rules, the more counterparties you have, the more credit exposure you are allowed, up to a limit of 40%. 

Critics also accuse the multi-swap model of lacking transparency. “What determines the duration of your exposure is the swap agreement,” said Costandinides. “These are customised, tailored documents, each one with a different duration. Also, you never know what exposure you have to each counterparty.” 

Furthermore, it can be difficult to see what collateral is posted against the swaps that exchange the Libor-based yield on funds from the ETF for the total return of the underlying index. 

Multi-swap platforms say they make all the necessary information on counterparty exposure available to investors weekly. 

Under the single swap provider model, the issuer controls all aspects of the replication process, from the swap pricing to the collateral provided and the management fees. Providers dealing with a consortium of swap providers from other firms may not enjoy the same degree of control. 

But perhaps one of the biggest criticisms of the multi-swap model is that, when all is said and done, it is no more than a side issue. Users should not waste too much time worrying about structural details, warn experts. 

Hood at Source conceded that structural risk should not obscure more important considerations such as the investment decision. “You don’t want the tail to wag the dog,” he said. “Structural risk is the tail and the dog is what the ETF is going to do.”

“The biggest risk to understand is the investment risk,” said Costandinides. “You have to make sure you are educated about what you are doing, as opposed to how many counterparties you have. For me, that’s secondary. It doesn’t really matter.”