On Wednesday, the US senate will hold a subcommittee hearing on exchange traded funds and their impact on market structure.
Three key questions the senators will be asking in the opinion of Nicholas Colas, chief market strategist at ConvergeEx, will be:
• Do ETFs affect market volatility?
• Do leveraged and inverse ETFs affect market volatility differently than other ETFs?
• Do ETFs present systemic risk to the financial system?
Commenting on the matter will be a range of market experts, including Noel Archard, a managing director at the world’s-biggest ETF firm iShares, as well as Harold Bradley, the chief investment officer at the non-profit Kauffman Foundation, who penned one of the first comprehensive critiques of exchange traded funds last year.
As Colas notes, there are likely to be as many impressions of the answers to these questions as there are people offering opinions — a reflection of the subject’s divisive impact on the market.
Whatever the views, as Colas points out, there’s no escaping some basic market facts:
In one corner, we have global equity markets that have gone from “Zero to Hero” in a few short weeks. From the lows on the S&P Futures just two weeks ago at 1072, we are up 11% (even after yesterday’s pullback). The CBOE VIX Index has gone, ping-pong-like from 32 to 45 to 28 to 33. In 20-odd trading days. The long end of the yield curve – the ultimate safe haven trade in the current tape – has gone from yielding 3.35% to 2.75% to 3.14% over the same period. So we should see evidence of all this volatility in ETF assets.
Yet for the leveraged/inverse fund critics out there, Colas points out, these funds have hardly seen any inflows in that time frame.
In fact, he notes:
The bottom line is that during the last month of extreme market volatility these funds have received just over $1.2 billion in fresh money. This is the equivalent of $60 million/day. Taken in total, these funds comprise 3.6% of total ETF assets.
Critics of the leveraged and inversed products state these ETFs can both exacerbate intraday moves to the downside as well as artificially lift stocks into an already-strong close. There is no publicly available data to support that assumption, at least as far as I have seen. Anecdotes abound; proof seems in short supply. What is provable, as the numbers shown above indicate, is that this segment of the ETF marketplace is very small relative to almost any other category. Anything is possible, of course, but the assertion that $36 billion of ETF products can consistently determine the intraday course of U.S. stock markets needs more substantiation than is currently on offer.
Something which doesn’t actually surprise us here over at FT Alphaville. The leveraged/inverse ETF debate may be a red herring.
Whilst we too have little in the way of proof, our inclination is to agree with the idea that end of day volatility has very little to do with leveraged or inverse ETFs per se.
Rather, we would say, it’s due to the mechanics embedded in the way all ETFs trade.
Gary Gastineau (the unofficial father of ETFs) recently noted, for example, that the ETF market suffers from a major over-dependence on market-on-close (MOC) orders which can create some warped pricing incentives at the end of the day:
Market makers in even the most thinly traded ETFs understand that the midpoint of their daily 4 p.m. quote will be preserved in prospectuses and on ETF Web sites for years to come. These market makers have a stake in attracting traders to the ETFs they trade. Consequently, they monitor their real-time bid/offer NAV calculations closely as 4 p.m. approaches. Even if they have to widen or otherwise change their spread for a few seconds, they will work to get the midpoint of their bid and offer as close to the expected 4 p.m. NAV as possible. (6) Their 4 p.m. quote is the most widely scrutinized and least useful bid/offer of the day.
Publication of this premium and discount information based on 4 p.m. ETF share quotes and NAV calculations has led to overuse of MOC orders, especially for ETFs that are thinly traded. Most investors do not realize that MOC transactions in ETFs are not reflected in most ETF reported premiums or discounts in any way. Nonetheless, MOC orders often are used by individuals and defined contribution retirement plan investors who are accustomed to buying and selling no-load mutual fund shares at NAV. Publication of this premium and discount information accounts for the fact that MOC trades account for a disproportional share of ETF trading volume.
The idea that volatility is being caused by a last minute rush to ensure NAV/price deviations are eliminated with the use (or abuse) of MOC orders is definitely one factor that should be investigated more closely.
The other factor to investigate, we would argue, is the ongoing tendency for ETFs to experience counter intuitive flows with respect to market dynamics.
As Colas himself notes:
The last month has actually been a bit sleepy in terms of money flows into equity products. Overall, equity linked funds are up some $1.4 billion in new capital. However, one fund (EFA, the iShares MSCI EAFE product) represents all of that increase, and then some. In the past month EFA added $3.2 billion in new money, so without that addition the equity ETF asset class would have lost $1.8 billion.
Of course, if ETFs are being used more prominently for shorting by hedge funds, principal trading firms and market makers rather than long-only asset managers then the idea of counter-intuitive flows suddenly makes more sense.
FT Alphaville recently received some interesting data from Ancerno, a provider of trade cost analytics to institutional investors, which supports the above notion. In August, for example, Ancerno found that while five of the top six traded securities globally were ETFs, only one of the top six securities traded by institutional clients was an ETF — the SPY . This they say is only because the SPY is used as a default investment by fund managers who have a restriction on the amount of uninvested cash they can hold.
It’s clear, therefore, that the market likes to trade ETFs much more than institution investors (and often for shorting specifically).
What’s more, many of these shorts are created on what one might call “a phantom basis” — using a mechanism known as “create-to-lend“. In the US market, this is known as the easiest and most popular way to short ETFs.
While it’s relatively complicated to understand, it can best be described as a mechanism which allows for the creation of fresh units (for shorting purposes) without a simultaneous increase in assets under management.
Because the short units are sold into a market which has no real demand for them, the selling pressure creates a deviation which encourages authorised participants to immediately redeem the excess units. The short interest remains high, but the overall shares outstanding and AuM is thus unchanged.
In the event of a major price slide in the market, however, the incentive to close out those shorts kicks in. If the short covering is large enough, it can create significant new demand for units leading to a deviation large enough to encourage the creation of fresh underlying stock by market makers. This time, however, the new units do add to assets under management, at least until a new deviation which favours redemption comes about.
So, the act of shorting-covering by hedge funds and principal traders potentially encourages the growth of assets under management when stock prices are falling, while the act of shorting not only neutralises the creation of shares for shorting, but possibly creates deviations that favour redemptions by market makers who have inventory they are waiting to offload.
Either way both ETF volatility and counter-intuitive flows deserve further investigation, with particular attention given to the role of market-on-close orders and the “create to lend” mechanism.
Related links:
Kauffman: ETFs are the problem, not HFT – FT Alphaville
Leveraged ETFs: not for retail investors – FT Alphaville
Is there such a thing as the “FT effect”? - FT Alphaville
The end of diversification? – FT Alphaville