Collateral: when it hits the fan

by , 27th November 2017

This article is part of our ETF education series. It explains everything investors need to know about collateral in ETFs. It looks at why ETF issuers use collateral, what rights do investors have to collateral in the event of defaults, and what types of securities get used as collateral. 

In the beginning, there was no collateral in ETFs. Investors had faith in things like swaps, stock lending and investment banks. ETFs never used collateral and no-one ever asked for it. But then the 2008 financial crisis happened. It hit the fan. And investors wanted a more security. Now collateral is everywhere.

Despite its ubiquity, collateral remains one of the opaquer wings of the business. Investors rarely know how it works, what collateral their ETF holds, or what claim they have to it in the event of a default.

But how does collateralisation work? What gets used as collateral? And what are the costs?

Collateral in swap-backed ETFs

Swaps are often thought to be of interest only to corporate finance pointy heads. But in the past, the majority of European ETFs used swaps to track their indexes. This number peaked at 60% almost ten years ago, but is now down to around 22%, according to Morningstar data.

Source: Morningstar, A Guided Tour of the European ETF Marketplace

They do this because swaps are, for big European investment-banks-cum-ETF-issuers, often easier and cheaper than buying every single share in an index. (In some European countries, especially France and Germany, the ETF issuer and the investment bank will often be part of the same company.) They also make it easier to track less liquid investment products – like emerging market mid caps or junk bonds – because few investors want to hold those products long term. They also open up investment positions – such as 4x the US dollar to euro exchange rate, and other highly leveraged positions – that are not possible to physically buy. Swaps also mean there is very little tracking error as the performance of the index is guaranteed by the swap provider.

But swaps have drawbacks. Crucially, this includes counterparty risk. Because the swap is being provided by an investment bank – or sometimes two investment banks – whatever risks these investment banks face, the owner of the swap-backed ETF also faces. This on top of the risk that the asset they invest in fails to perform. In the absolute worst-case scenario swap providers can default – meaning the ETF and the swap it rode in on can collapse. This is where collateral comes in.

Different kinds of swaps, different kinds of collateral

ETFs use two different kinds of swaps when tracking an index: funded and unfunded. Unfunded are easier to understand and more common, so we’ll start with that.

Unfunded swaps and collateral

When ETF issuers knock on investment banks’ doors and ask for swaps to back up their ETFs, investment banks require them to buy something called a “collateral basket” – a mix of assets that are tax efficient, liquid and safe. In practice, a collateral basket is usually full of government bonds, cash and sometimes blue-chip shares. In Europe, collateral baskets must comply with UCITS regulations on diversification and liquidity. 

With collateral basket in hand, the ETF issuer then swaps the performance of this basket for the performance of whatever index it wants to track – be it emerging market construction companies, junk bonds or whatever else. This is why it gets called a “swap”: the ETF issuer swaps something, the basket, with the investment bank, the performance of an index.

If the investment bank goes broke, the collateral basket boomerangs back to the ETF issuer.

The beauty of an unfunded swap is that market forces guarantee that the collateral will cover the value of their ETF. Why? Because the return on the collateral basket is what gives the investment bank their profits. If they’re paying out more on the performance of the index (i.e. their end of the swap) than they’re taking in on the performance of the collateral basket, they’ll lose money. They, therefore, have every incentive to ensure that there is enough collateral to cover the swap.

Source: Morningstar, ibid.

Funded swaps and collateral

Funded swaps are different, and in many ways funded swaps aren’t actually swaps at all, but more like an exchange traded note.

Funded swaps receive the performance of an index from an investment bank, as with unfunded swaps. But instead of giving the investment bank a collateral basket, they give them cash. The investment bank then goes and turns this cash into a collateral basket themselves.

What happens next is crucial and something any investor considering a funded-swap-backed-ETF should know.

In some cases, the investment bank will give the collateral to a segregated account with a transfer of title in place. But other times they will put it in a pledge account. This difference is extremely important because it determines what rights investors have to collateral 

When a default occurs in ETFs using a segregated account, the transfer of title means that the collateral is transferred from the segregated account to the fund’s account and the ETF issuer has direct access to the collateral. The pledge account, however, is not so simple. When defaults arise, a pledge has to be enforced. In this scenario, it is possible that bankruptcy administrators – whoever they may be – freeze the assets in the pledge account.

If your ETF uses swaps, it’s very important to know what kind: unfunded or funded? And if funded, what kind of account holds the collateral?

Source: Morningstar, ibid.

Physical ETFs

ETFs that buy and hold the underlying stocks in their index are called physical ETFs. The vast majority of ETFs around the world are physical.

It is sometimes thought that physical ETFs do not use collateral. But this is a mistake. Physical ETFs use collateral too and involve counterparty risk much like swap-backed ETFs. This happens through securities lending – described below.

See this article on securities lending

Securities lending

Physical ETFs often lend out stocks held in their fund. How much they lend out depends on the ETF and the issuer. Some cannot lend stocks, others can lend out most of them. ETF issuers do this so they can make more money as stock borrowers pay lending fees.

Whether physical ETFs are safer than swap-based ETFs has been subject to debate for several years. The majority opinion at the time of writing is that physical ETFs are safer because, in the event of a counterparty default, investors still have a legal claim to the stocks that make up whatever index the ETF is tracking. But stock lending complicates this, as we’ll now discuss.

Collateral in physical ETFs

ETF issuers lend out equities

Collateral comes into physical ETFs because when a fund lends out stocks – usually to short sellers – there is always a chance the stocks won’t come back. To cover this, exchanges and industry practice require that borrowers post collateral. The collateral they put up must be high quality and liquid – again, usually treasuries or cash. The collateral must exceed the value of the stock that was borrowed – usually, in the United States at least, that value is between 102 and 105%. This means that if the borrower defaults, the ETF manager can sell the collateral and use the proceeds to buy back the stock that they lost.

But posting collateral is not a silver bullet for counterparty risk.

Let’s take short sellers as an example (those borrowing stocks from ETF issuers are almost always short sellers).  And let’s say our imaginary short seller borrows Apple stock and sells it, hoping the price drops. But instead of dropping, Apple announces that it is launching a new phone, causing its stock to rally strongly. What can happen here is that the value of Apple’s stock can come to exceed the value of the collateral posted. This can mean that the borrower cannot return the Apple stock he borrowed – because he sold it. And it also means that the ETF issuer cannot use the collateral, which, again, can be as little as 102% of the value of the stock borrowed, to repurchase the Apple stock either. The result is a potential default.

ETF issuers also lend out collateral

But things can get worse. As it turns out, ETF issuers don’t just lend out the stocks in their funds: they also lend out the collateral that short-sellers pledge. This stage is crucial because, as Vanguard notes: “The most significant risk in securities lending lies not in the [stock] lending itself, but in the reinvestment of the cash collateral.”

If issuers receive cash collateral, as they very often do, more often than not they throw it over to the money market to earn interest on it. (In practice, the lender and the borrower often share the gains of money lent, or the gains are used to offset the rental fee on the stock borrowed).

The ETF issuer has discretion over how the collateral gets reinvested. On a low-risk approach, it would invest in short-term government bonds. In a higher risk approach, in lower rated and longer maturity securities. The latter can be risky, because if money market products they invest in can go bust – as happened with Lehman Brothers in 2008 – they can be left with nothing.