A recent Financial Stability Board (FSB) report and a few subsequent blog posts and articles have created a bit of a stir among ETF investors, and for what sounds like good reason. With headlines like “ETFs are Becoming a Shadow Banking System” and “ETFs: The Next Financial Time Bomb?” I myself would be alarmed about the safety of investing in ETFs – if I didn’t understand that there’s more to the story than what’s being reported.
The FSB’s April report, entitled “Potential financial stability issues arising from recent trends in Exchange Traded Funds (ETFs)” focused on two potential risks – first, the risks associated with the structure of ‘synthetic ETFs’ (ETFs that use derivatives such as swaps); and second, the use of a practice called securities lending in ETFs. For the sake of pithiness, I’m going to tackle the former issue here and the latter in another post, which you’ll see published on our blog shortly.
The first thing to understand is that a majority of the synthetic ETFs exist outside of the United States, especially in terms of AUM. In the US, the vast majority of ETFs are what are referred to by the report as ‘plain vanilla’ or what we typically call ‘physical ETFs’ (i.e. ETFs that hold the physical securities). Physical ETFs are regulated under the Investment Company Act of 1940, which limits the use of derivatives and prohibits affiliated transactions (for example, a bank can’t serve as both sponsor and swap counterparty). Even most leveraged and inverse ETFs are regulated under the Act and, therefore, their use of derivatives is also limited.
In addition, the SEC announced last year that they are further evaluating the use of derivatives in ETFs, which means that there likely won’t be any synthetic ETFs launching in the US until the SEC has better defined their use in structures like ETFs and Mutual Funds.
The FSB report states that “although most of the ETF market remains plain vanilla, there has been an increase in product variety and, in some cases, complexity, albeit with some differences across regions and markets.” One of the “different markets” they’re referring to is Europe, where the Undertaking for Collective Investment in Transferable Securities Directive (UCITS) is much more permissive about allowing ETFs to gain exposure using derivatives. So in the context of the FSB report, the discussion is mainly focused on ‘synthetic’ ETF products listed in Europe that gain their entire exposure via swaps.
There are several ways to structure a swap based product. The FSB report highlighted the potential risks – in terms of quality and liquidity of posted collateral – associated with an “unfunded” swap ETF structure* using an affiliated or parent bank as the swap counterparty. This is in contrast to using multiple counterparties, setting standards for collateral, keeping the sponsor independent from the swap counterparty, and using a fully funded swap with the collateral managed outside the fund by a third party. In the regions where iShares would consider using a synthetic model, this is our preferred method.
There are times when using a synthetic model makes a lot of sense. For example, it can be a good way of gaining exposure to markets that can’t be accessed through physical-based funds (such as commodities and some hard to access emerging markets countries). Furthermore, ETFs that gain their full exposure via a swap have virtually no tracking error, since the performance doesn’t hinge on the provider’s ability to track the index. But you can’t (and shouldn’t) hide the risk from investors. The tradeoff of using a swap to deliver the return of the index is the added risk of a default by the swap counterparty(s).
iShares will consider using synthetic structures in various regions, but we’ll only do it when we feel it’s the best approach. For example, we don’t believe the unfunded model highlighted in the FSB report is the best way to serve our clients. We in the ETF industry need to do a better job highlighting the risk and rewards of these types of structures so that investors can make better informed decisions on what they are willing to put in their portfolio.
So is there cause for concern, as the articles above suggest? Not in my estimation. The synthetic ETFs being referenced in the report and the media still only make up about 13% of global ETF assets, and within this category there are several different structures available, some of which are simply better than others. To imply that all ETFs are broken because of a few that carry less familiar risks is a stretch, to say the least.
Stay tuned for my next blog post, where I’ll be covering the FSB report’s concerns about securities lending in ETFs.
* There are several synthetic ETF models available, but the general idea of most is that they derive their returns not from actually holding the underlying securities of the index they’re seeking to track, but rather from entering into a swap arrangement wherein a financial intermediary promises to deliver the return of the index the ETF is seeking to track. In an ‘unfunded’ swap ETF structure, the intermediary (considered a counterparty in this transaction) delivers a basket of collateral assets to the ETF sponsor in exchange for cash, and the total return of the collateral is then delivered back to the counterparty.