Since the onset of the financial crisis in 2007 the asset management business has barely grown. One glimmer of light in this enduring industry gloom has been exchange-traded funds. Assets under management for ETFs have grown from $766 billion to $2.09 trillion. The compound annual growth rate since 2000 has been 30%.
It is now 20 years since State Street established the first ETF. The popularity of the product is not surprising. By offering cheap index exposure across a range of asset classes, ETFs have revolutionized investment. The man in the street can now deploy his savings in markets in the manner of George Soros. Global macro has been democratized.
ETFs are a great product, and for the sake of full disclosure it should be declared that your correspondent has invested in them for many years. However, success might be breeding complacency. ETFs have proliferated well beyond their beginnings as vanilla beta products tracking liquid equity markets. You can now buy inverse (short) ETFs and invest in asset classes ranging from emerging market debt to equity volatility.
Therein lies the problem. ETFs hold out the prospect of instant liquidity. However, many of the assets ETFs track are inherently illiquid. That was brought home in the recent dislocation in bond markets. Emerging market debt ETFs fell to substantial discounts to their underlying net asset values. It was also reported that State Street stopped accepting cash redemptions for municipal bond ETFs.
This situation is not likely to improve any time soon. Investment banks used to warehouse and transfer risk in markets through their market-making businesses. Ill-judged regulation has meant that banks no longer have the appetite or ability to hold inventory. In a market crisis they have to price stocks, bonds and other assets to go. That exacerbates volatility and the likelihood of prices gapping.
If you add ETFs into that mix it becomes even more combustible. Imagine that you are a trader at an investment bank and you take a call on a large sell order from your mate who happens to work at an authorized participant in the ETF market. You know he or she is a forced seller. The need to destroy the units of an ETF implies large redemptions of the underlying fund.
Typically, each unit represents 50,000 shares. Any trader will know that if there is one stressed seller in the market there will likely be others. Underlying liquidity conditions in those physical securities are also likely to be poor as other investors seek to sell. A forced seller in a volatile and illiquid market will send alarm bells ringing in the dealer community and they will price those assets accordingly. This, in turn, increases the likelihood of market dislocations.
However, not all ETFs are backed by cash securities. Many, in Europe in particular, are effectively total-return swaps. In the jargon, they are synthetically replicating the index. This eliminates tracking-error risk, as providers like to point out. But it introduces others. The first is that the ETF buyer is exposed to the risk that the swap counterparty will go bust. Collateralizing the swap mitigates this. However, collateralization also comes with problems.
First, ETFs are sold as transparent, easily understood products. Probably, few buyers realize that they are often buying a derivative. If the swap is collateralized (funded), that collateral does not necessarily match the underlying assets. There is almost no point using the swap if the assets backing it track the index you might as well structure a physically backed ETF.
This means there is a mismatch risk between the swap and the assets. It also gives the ETF supplier the opportunity to use inventory or securities it would not be able to fund in other collateral markets to back the swap. Using illiquid securities to back a supposedly liquid ETF smacks of term transformation. Any ETF can theoretically be fully redeemed tomorrow.
It is possible to think of a situation in a market crisis where a large swap counterparty is subject to a run. As credit lines disappear, rumours would spread. Delivering cash against collateral assets would be difficult, as funding via the repo and stock-lending markets would soon dry up. The prospect of a seizure in collateral markets might rapidly undermine the ability of ETF managers to provide liquidity.
Of course, the likelihood of a fund being rapidly redeemed is small. But the dislocation in bond markets in June and the knock-on effects in ETFs is a timely reminder that financial engineering cannot make illiquid assets liquid. That would be alchemy, not dissimilar to turning sub-prime slime into triple-A rated CDOs. Regulators would also do well to consider the extent to which banks with ETF businesses using synthetic replication are engaged in term transformation.
ETFs have been a roaring success. In the past 20 years they have proliferated and been transformed. They are no longer simple beta products and they are big enough to have an impact on broader markets. After the flash crash in 2010 and the problems experienced by some ETFs in Junes market dislocation an ailing canary should be noted in the coal mine.
Andrew Capon has won multiple awards for commentary and journalism on markets, investment and asset management. The views expressed are the authors own.