‘Shadow trading’ ETFs

Well well . . . This, from our mainFT colleague Steve Johnson, looks intriguing:

Insider traders have used exchange traded funds to conceal billions of dollars’ worth of trades, according to a team of academics who say their finding may be just the “tip of the iceberg”.

Their analysis suggests at least $2.75bn worth of anomalous trades occurred in US-listed ETFs before merger and acquisition announcements between 2009 and 2021.

“Our findings suggest insider trading is more pervasive than just the ‘direct’ forms that have been the focus of research and enforcement to date,” the academics from institutions in Sweden and Australia said in the paper, Using ETFs to Conceal Insider Trading.

Here is the full paper, written by Elza Eglite and Dans Staermans from the Stockholm School of Economics in Riga, and Vinay Patel and Talis Putnins of the University of Technology Sydney. They say they find “significant levels of shadow trading” in some sector ETFs ahead of M&A announcements, which adds up to $212mn of trading a year.

But we have some doubts.

Intuitively, the paper makes sense. The SEC has become increasingly sophisticated and adept at spotting traditional insider trading, using things like the new Consolidated Audit Trail. But as Matt Levine wrote recently, a company’s material nonpublic information affects lots of securities, not just the obvious candidates (ie the company’s stock, or short-dated out-of-the-money call options on its stock).

Suspicious trading in those less-obvious securities may be harder for regulators to detect. And even if regulators do find sketchy trades, it could be harder to prosecute than “classic” insider trading.

As Steve points out, the first test case is still wending its way though the courts. Here’s the paper’s argument for the usefulness of using ETFs to “shadow trade”.

ETFs provide an attractive instrument for insiders to trade their private information for several reasons. First, the stock that is the subject of the information may be a constituent of the ETF, so that one can get a direct exposure to the company’s share price via the ETF, but in a vehicle that is more subtle than trading the company shares directly, helping reduce scrutiny from law enforcement. Second, ETFs are cost-effective and often more liquid than the underlying company shares, potentially reducing the price impact of insider trades. Both theoretical and empirical evidence shows that insiders trade in highly liquid assets so that they can hide their information and maximise their trading profits. Third, shadow trading in ETFs prior to price-sensitive news allows insiders to benefit from increases in the price of both the source firm and related firms.

It’s a fun paper, to be sure. The uptick in sector ETF trading volumes ahead of market-moving corporate announcements is eye-catching, and worth thinking about. In practice there are some problems, however.

First, in most sectors, insider knowledge won’t help you predict what an ETF will do, unless the information is about one of the biggest companies in the fund. And, well, those companies just don’t tend to be acquired very often. In most other cases, it won’t have much of an impact.

For example, the 2017 news that Amazon had agreed to buy Whole Foods for more than $13bn sent the supermarket chain’s stock up by almost 30 per cent (and Amazon up 2.4 per cent). But the consumer staples sector ETF fell by about 1.6 per cent.

In theory it’s possible to use very niche ETFs — the paper cites the iShares Expanded Tech-Software Sector ETF, and Vanguard’s industrials and healthcare ETFs — but a trader would still need to find one where the company they want to “shadow trade” accounts for a large weighting.

In the three ETFs cited above, no one company makes up more than 8.9 per cent of the portfolio. Even the spivvier ETFs usually have pretty strict weighting ceilings. In other words, if you are a nefarious CEO who wants to “shadow trade” your company’s stock, it might be difficult to find an ETF that will react strongly to the news. You could use options for a more leveraged bet, but only very large ETFs have robust derivatives markets referencing them, so it would still be hard to get a lot of pop out of the trade.

Say our ne’er-do-well CEO does find an appropriate ETF. It’s not always clear how the market will interpret the broader impact of, say, an M&A deal, or bumper earnings. The news could be good or bad for other stocks in that sector — for example, good earnings might mean industry-wide tailwinds, or simply that a company is benefiting from the weakness of rivals — making buying a sector-specific ETF a bit of a pot luck.

Let’s revisit the example of Amazon buying Whole Foods: how would someone profitably bet on the consumer-staples ETF with advance knowledge of that deal? A trader would need to not only correctly predict that investors’ worries about competition from Amazon would outweigh the mega-rally in Whole Foods, but also structure a short position with enough conviction to make it worth the cost and legal risk. That’s a tall order for a price move of less than 2 per cent.

Lastly, while the academics’ headline number of $2.75bn looks large, it averages out to $212mn of “surplus” trading per year in sector ETFs. That sum is low compared to actual trading volumes; QQQ alone trades an average $15bn a day. The energy sector-focused XLF trades ca $2bn a day. Even the most listless of the smaller ETFs in the paper trades almost $16mn a day, on average. So over the course of an entire year, across all sector ETFs, $212mn starts to look like a statistical oddity.

There could well be some “shadow trading” in ETFs happening — in fact it would be surprising if no one tried it. But it’s unlikely to be as pervasive as the paper’s authors suggest, and probably not as profitable as the insiders would like, either.

Click here to visit the ETF Hub