Since the credit crisis shook up the global financial system in 2007-2008, investors have become increasingly concerned over products using derivatives. Swap-based ETFs are no exception. The two most common criticisms of synthetic ETFs are their complexity and lack of transparency. In this article, we will try to demystify these funds and shed some light on the risks involved in investing in them.
Unlike a cash-based ETF, a synthetic ETF doesn’t hold the underlying index constituents. Instead, it holds a basket of securities which may be completely unrelated to the index it is tracking and to which investors have recourse in case of issuer’s failure. The fund typically will enter into a swap arrangement in which it gives away the performance of the collateral basket in return for the performance of the fund’s reference index. The swap counterparty is usually an investment bank.
This structure has been widely embraced in Europe because it reduces tracking error, it is lower in cost and it enhances market access. In fact, today there are more ETFs that use the synthetic replication method than those that use physical replication, although cash-based ETFs have the greater share of assets in Europe. Despite having numerous undeniable advantages, swap-based ETFs bear counterparty and collateral risks that shouldn’t be overlooked.
Collateral to Offset Potential Damage
Within the context of a synthetic replication ETF, counterparty risk is the risk that the swap writer fails to fulfill its obligations. Under UCITS III, this risk is limited to 10%, which means that the ETF issuer or the swap counterparty must provide collateral amounting to at least 90% of the ETF’s net asset value (NAV). The basket of securities held (or pledged) as collateral is marked-to-market on a daily basis to ensure that its value doesn’t fall under that regulatory limit. In practice, all ETF issuers maintain a margin of safety, which varies from one provider to another. They all maintain different levels of collateral (measured as a percentage of the funds’ NAV) and reset their swaps at different trigger points.
To learn more about these various collateral management policies, we have spoken to the main synthetic ETF issuers in Europe. As you can see in the table below, there is quite a diverse range of practices amongst them. Collateral levels can oscillate between 90% and 120% of each fund’s NAV, resulting in under-collateralisation (<100%) or over-collateralisation (>100%) at any given point in time.
All providers have set average “minimum collateral” thresholds for their ETFs. Each time the marked-to-market value of the collateral falls under that minimum, the fund provider resets the swap and asks the counterparty to deliver additional collateral (cash or securities). Resetting these swaps brings counterparty exposure to zero as the collateral posted to the fund is typically reset back to 100% of the fund’s NAV. Some providers such as db x-trackers, ETF Securities and Comstage bring their funds’ collateral back to over 100% (counterparty exposures become negative) while Amundi maintains a minimum counterparty exposure of 2% to 3%, resulting in a permanent under-collateralisation.
It’s worth noting that all the limits in the accompanying table can vary from one asset class to another, depending on their relative levels of volatility. For example, fixed income funds will typically have lower thresholds than equities because they are less volatile. It’s also understood that ETF providers maintain some flexibility around these limits so long as the 90% UCITS threshold (10% maximum exposure) is not breached. In terms of best practices, we think it is safe to say that the higher the level of collateralisation, the more protection is provided to investors in the event of a counterparty defaulting.
Collateral Quality and Frequency of Swap Resets
The level of collateralisation is not the only factor that should be taken into consideration when assessing investor protections in swap-based ETFs. Various additional factors come into play, in particular the quality of collateral as well as the frequency of swap resets.
The collateral will only play a role if the swap provider fails and no replacement is found. In that hypothetical situation, the ETF provider would have to quickly liquidate the substitute basket of assets, which is likely to be uncorrelated to the underlying index. This is the reason why collateral baskets are usually composed of securities that are liquid (blue chip equities, government obligations, etc.) and preferably traded in or near the same time zone as the market where the ETF is traded.
Swap-based ETF providers have defined different sets of criteria for what they will accept into their funds’ collateral baskets, with some being more conservative than others. For example, Amundi and Lyxor’s equity ETFs hold only European stocks as collateral and rule out investing in Japanese securities, whereas others like Source and db x-trackers may accept Japanese equities (some db-x trackers have held up to 35% of their collateral baskets in Japanese shares). The issue with holding Asian stocks as collateral is that they might not be able to be sold in a timely manner in the event of a default on the part of the swap counterparty due to the lack of overlap between normal European and Asian trading hours.
All the ETF providers we talked to, except ETF Securities, disclose snapshots of fund collateral on an annual and/or semi-annual basis and upon request. ETF Securities would only reveal the collateral schedule detailing the restrictions on the collateral that the funds accept. As a result, investors don’t know how many counterparties (between 2 and 4) are involved in each of the firm’s ETFs. We think the company’s policy should be more transparent: ETF Securities should disclose the actual composition of its collateral baskets and the names of the counterparties.
Another factor worth paying attention to is how frequently the swap is reset. Swaps are usually reset when (i) the exposure to a counterparty reaches the trigger point set by the ETF provider (see the table), (ii) whenever there is a subscription or redemption at the fund level, and/or (iii) on a regular basis. Resetting a swap to zero eliminates (temporarily) counterparty exposure. So the more frequent the reset, the better, in terms of investors’ protection--although it does result in additional costs for the fund.
Room for More Transparency
In summary: all of the leading ETF providers in Europe have synthetic replication ETFs. Even iShares, which is renowned for its physical replication funds, has 19 swap-based funds on its Munich platform. Better tracking and access to new asset classes are undeniable advantages. However, investors need to fully understand the counterparty risk embedded in the swaps and determine the level of risk they feel comfortable with. For that, they need to check the relevant ETF prospectus to see what policies the manager follows. However, in many cases we think the information provided on prospectuses and websites doesn’t fully enable investors to make informed investment decisions. While most ETF issuers have been quite candid with us about the operational details of their collateral policies, they don’t disclose them on their websites or in their filings. As far as collateral is concerned, we believe that fund providers should commit to full transparency --as it’s supposed to be one of the key advantages of ETF ownership. Investors shouldn’t have to call or send an e-mail to request snapshots of their ETF’s collateral basket. Information like the identity of swap counterparties, the composition of the collateral basket and how often swaps are reset is information investors have the right to know. It should be made readily available to them.
