In the aftermath of the market’s sharp sell-off in June, some concerns were raised about how exchange-traded funds (ETFs) fared during the selling spree. An article in the Financial Times on June 21 noted that as a result of the heavy selling pressure, a couple of U.S.-listed ETFs traded below the value of their underlying assets. Several other news stories have since been published that highlight the perceived disconnection between an ETF’s price and its net asset value (NAV) during periods of heightened volatility.
There are important lessons to be gleaned from the flags raised by the sell-off and subsequent media coverage. Sometimes an event like the market volatility spurred by the U.S. Federal Reserve statement on the afternoon of June 19 provides a great catalyst to encourage further education.
ETFs have been heralded for their ability to consistently trade at or very near to their underlying net asset value. So when a reputable news source claims that the mechanism broke down, the natural reaction of many is to start ringing the alarms. Before delving into the issue, let’s begin by reviewing the mechanism in question.
The information required to calculate the NAV for an ETF that tracks, say, the S&P/TSX 60 index is simply the current prices for all 60 index constituents aggregated in proportion to their weight in the index. The intraday NAV allows market participants to see whether the ETF is trading at a discount or premium to its underlying portfolio. This can be done easily throughout the trading day, because the ETF and its underlying securities trade together during the same market hours. At any given time during the day, market participants can get an accurate quote for the ETF’s NAV based on how the underlying stocks are trading.
With this in mind, let’s consider the ETFs in question, which included an emerging markets fund and a municipal bond ETF.
As you may have already surmised, the primary issue with the emerging markets ETF was that the underlying securities trade in different time zones on various exchanges around the globe. In fact, about 75% of emerging markets are closed while Canadian and U.S. markets are open for trading. Therefore, unlike the S&P/TSX 60 ETF example where the ETF’s market and that of the underlying stocks overlap perfectly, the emerging markets ETF trades while the markets where the majority of its underlying stocks trade are closed.
In this case there is actually very little value in the ETF’s intraday NAV, which is the last close price for the stocks. Imagine you have an ETF that is trading actively, but the NAV is basically frozen in time because it is based on the last trade prices of markets that have been closed for several hours. That means there’s a very good likelihood that a perceived “premium or discount” will appear when some market-moving news or event occurs. In reality, the NAV is a stale value, while the ETF market price represents the most current value that active market participants assign to the securities.
Which number is a more accurate reflection of true market value? In the case of the market’s recent sharp sell-off, the ETFs were actually acting as the price discovery mechanism. As the only open and functioning market, it was the ETF that was providing the market’s consensus view by discounting the new information that was hitting the wires at the time.
A municipal bond ETF was the other example cited in the financial press, and presents a similar case. Again, the issue at hand relates to operational differences between the ETF and the underlying market that it tracks. In this case we’re dealing with relatively illiquid bonds that still trade over the telephone and an ETF that trades electronically over a centralized exchange.
Unlike stocks, which are standardized and trade relatively frequently, bonds lack homogeneity and in many instances can go days or weeks without trading. That makes an index based on the last traded prices pretty useless for an ETF that trades frequently throughout the day.
Most bond indexes therefore rely on matrix pricing models to come up with more accurate and up-to-date index values. Matrix pricing models provide a major improvement over stale bond prices, but have their limitations. Particularly during periods where the markets react swiftly and fiercely to newly released information (as we saw on June 19 when the U.S. Federal Reserve released its statement in the afternoon).
So, we have another scenario where the ETF is the only real active market, where investors can bid to buy or sell in real time. That means that there’s a good chance that the ETF market price will drift away from the stale index price, as the new information is discounted by the market.
If a “premium” or “discount” temporarily develops, does that mean the ETF structure is broken, or that ETFs are a ticking time bomb for financial markets? Despite what some misinformed media reports would lead you to believe, that’s hardly the case. In fact, ETFs shine as the only real and functioning market during such a fast-moving period of market stress.
I use quotation marks and label the “premiums” or “discounts” as phantom because they are not real and are extremely temporary in nature. Most often what we will see in these cases is that the premiums or discounts will collapse within a day or two. In the case of the emerging markets ETF, the premium or discount disappears when the underlying market opens again for trading and the stocks immediately converge to what the ETF was already foreshadowing. In the case of the municipal bond ETF, the premium or discount collapses as soon as a few municipal bond transactions are registered and the matrix pricing models that drive the indexes have a chance to catch up.
Without a firm understanding of how ETFs function, it’s easy to see how one might come to the wrong conclusion regarding phantom premiums or discounts for ETFs. The flurry of misinformed articles from the financial media is a case in point. Hopefully, investors will take the time to dig a little deeper rather than running for the hills.
John Gabriel is an ETF strategist with Morningstar Canada, responsible for Canadian ETF research.