ETFs have grown by leaps and bounds over the past decade. From plain-vanilla offerings that turned the most liquid stocks in various national stock venues into investable bundles capable of being bought and sold at the price of a single stock, they have morphed, over the past decade into many other things.
Instead of tracking simply the top-volume stocks, such as the TSX 60, the S&P 500, the FTSE 100 and so on, they have added total market exposure. Not that they buy the total market – in the lower reaches of trading volume, representation means a workaround for illiquidity. Nevertheless, in the right hands, tracking errors have been quite low – total market ETFs have tracked and sometimes exceeded their index benchmarks.
But other assets – think commodities – are not so easy to invest in for somewhat similar reasons: they involve sampling for the less liquid tails of the target market. There’s a huge illiquidity tail here. But even where the underlying investments are investable, it may be cheaper to do something else, namely a total return swap, to skip the less tradeable parts of the market.
Thus, we get inverse ETFs, leveraged ETFs, commodity ETFs. With an inverse ETF, the underlying investments aren’t physically shorted – it would be too costly to do that. Instead, the futures are shorted – or better, the ETF provider enters into a total return swap on a target index. Therein lies difficulties, according to a pair of reports, one from the IMF and the other from the World Bank.
The IMF report (scroll to page 68) starts out with the notation that: “ETFs emulate the returns on an index by physically replicating the underlying index constituents, by synthetically replicating the index returns using swaps and other derivatives, or by using some combination of the two. U.S.-based ETFs typically use the physical replication technique due to regulatory constraints.”
The concern here is that “[i]n synthetic replication, ETF managers hold a basket of assets, different from the benchmark index’s constituents, and swap the returns of this basket for the actual returns on the reference index through total return swaps (TRS).”
Tracking error therefore falls on the counterparty. But not all counterparties are robust enough to survive a financial crisis – as was seen during the Lehman Brothers collapse. There, the financial crisis affected investors in two ways. First, retail principal-protected notes based on Lehman’s own credit capacity failed, leaving investors considerably out of pocket – even if they bought the notes from some one other than Lehman. So that’s counterparty risk one: your guarantor fails. Here’s counterparty risk two. Since Lehman Brothers was a major prime broker, it rehypothecated assets – it lent the assets already on its loan book to other customers – leaving the named owners, and the first-level borrowers –hedge funds among them – stranded in the whole receivership process of who had claim to what.
Could that happen with synthetic ETFs? The IMF report notes: “While synthetic replication eliminates tracking error, it comes at the cost of higher counterparty credit risk. Because the counterparties’ creditworthiness guarantees the return on these funds, ETFs, and subsequently investors, are exposed to the risk of one or more counterparties defaulting.”
How could that happen? Primarily through securities lending, which is an essential channel for market liquidity. Risks are supposed to be mitigated by the borrower putting up 102% or 105% collateral.
Here, the IMF reports: “Securities lending poses yet another counterparty risk, in which a default of the securities borrower could potentially leave the ETF provider scrambling to replace the securities it lent out. Tracking errors can be partially offset by lending securities to hedge funds and other institutions for short-selling and receiving a fee in return.
“Regulation currently requires ETF providers to be able to recall securities lent at a short notice and to adequately collateralize such lending. However, participants claim this process currently lacks transparency and that the cash reinvestment guidelines have not been clearly laid out by regulators. In addition, the ETF provider is exposed to the mark-to-market losses on the securities it holds in the swap basket.”
Apart from that, there is the impact of a run on the market – a flash crash, like May 6, 2010.
“While most ETFs are supported by one or more market makers, there is no guarantee of active trading under illiquid conditions. Analysts point to the so-called flash crash in May 2010 as an example of the risks ETFs are susceptible to, when market makers were overwhelmed by a surge in computer-driven selling. Market makers stopped offering bid-ask quotes, fueling volatility further and the eventual meltdown in equity prices on the Dow Jones index triggered heavy losses for some ETFs. In addition to risks posed by market makers, some illiquid emerging market assets also present challenges to ETF liquidity, as the issuing and redeeming of creation units become increasingly difficult under stressed conditions. Some market makers use derivatives to side-step the illiquidity issue, but given that such instruments are either absent or too expensive in most emerging markets, turnovers in such ETFs are typically low.”
Even outside of emerging markets, the derivatives may be no more safe nor liquid. Says the IMF report: “In a variation of the swap-based ETF, the provider sometimes transfers all the cash from investors to the TRS counterparty, which in turn pledges collateral to the ETF’s account at the fund’s custodian bank.
“In such a scenario, if the swap counterparty were to default, it could potentially lead the bankruptcy administrator to freeze all ETF assets, preventing the ETF from liquidating its assets if the need arises.”
Next Post: The World Bank report.