How to make ETFs less risky

By Felix Salmon
October 25, 2011
Congressional testimony last week.

" data-share-img="" data-share="twitter,facebook,linkedin,reddit,google" data-share-count="true">

Harold Bradley and Bob Litan made some very good points about ETFs in their Congressional testimony last week — testimony which Paul Amery today greets as “a mixed bag”. But it’s hard to argue with this:

We have enough history with financial innovations to at least raise questions when we see an innovation growing at very rapid rates. ETFs are no exception. We believe that these instruments may now be undermining the fundamental role of equities markets in pricing securities to ensure that capital is efficiently allocated to growing businesses. When individual common stocks increasingly behave as if they are derivatives of frequently traded and interlinked ETF baskets, then it is trading in the ETFs that is driving the prices of the underlying stocks rather than the other way around.

There’s a tragedy of the commons going on here: for any given individual investor, ETFs make a huge amount of sense. But if individual investors — and a lot of institutions, too — all pile into ETFs en masse, then stocks lose their price-discovery role, and large deleterious effects can start emerging.

For instance, one big reason to buy a market index rather than a handful of individual stocks is that the index provides diversification. But that diversification is disappearing as people pile into ETFs:

High co-movement of securities is not new, often occurring when markets reflect crowd panic or euphoria. What is new, however, is how ETFs decrease diversification benefits, with stocks and sectors worldwide moving together, even when there is no panic. Stocks move together today more than at any time in modern market history with recent data indicating that individual common stock prices that make up the S&P 500 index now move with the index 86% of the time.

Here’s one particularly striking chart:

correlation.tiff

This isn’t the S&P 500, which has even higher correlations. It’s just large-cap stocks. Which are all moving together in lockstep — much more than they were even in 1987, and more even than they were in the Great Depression.

Amery comments:

It’s common sense that the more investors buy and sell equities via index trackers like ETFs and futures, the higher you’d expect internal index correlations to be. And since ETFs have been growing in importance, the greater you’d expect their impact on index stocks’ co-movements to be.

Researchers at JP Morgan came to similar conclusions last year, arguing that high-frequency trading is also playing a role in boosting internal index correlations.

There’s nothing forcing investors, by the way, to purchase those index trackers where such herd effects are most prominent. You don’t have to buy the S&P 500—you could easily choose an index put together in different ways and featuring different stocks. Arguably, avoiding “overowned” indices should be near the top of a list of any investor’s priorities.

This I find a bit less convincing. As the above chart shows, there are very high co-movements even outside the S&P 500, it’d be hard to find any kind of broad stock index without a high correlation. And if you only get a small benefit in terms of extra diversification, then the costs, in terms of higher ETF fees for unusual instruments, are likely to be higher than the benefits from diversification.

That said, there are lots of people who make money from the huge amounts of cash sloshing into ETFs, and especially from the way in which the S&P 500 ETFs have to rebalance every time the index changes. And the more popular S&P 500 ETFs become, the more we’re all just funneling cash — some 30bp per year, it’s estimated — to those rebalancing arbs.

What is to be done? At the margin, Bradley and Litan’s idea that ETFs should have tighter circuit breakers than individual stocks is a good one. Amery reckons that “it’s important not to leave loopholes where trading in the ETF is stopped, but that in most of the underlying stocks can carry on” — but I don’t see that. If you want to keep on trading individual stocks, that’s fine — just stop the ETF driving their movement.

But fiddling around with circuit breakers won’t make any difference on a day-to-day basis. And cracking down on ETFs more generally risks throwing the baby out with the bathwater. My feeling is that the best thing to do here is simply adopt the kind of financial-transactions tax that the Europeans are talking about — a 0.1% Tobin tax on all ETF trades. That would drive away the day-traders and rebalancing-arbs from the ETF market, leaving it to the buy-and-hold investors that ETFs are very good for. It can’t do much harm, and it could well do quite a lot of good, from a systemic-risk perspective.

Comments
14 comments so far

“At the margin, Bradley and Litan’s idea that ETFs should have tighter circuit breakers than individual stocks is a good one.”

If the purpose of an ETF (as it is with an index) is to reduce volatility risk, then this should have been in place already.

I really don’t see what “baby” is with the dirty bathwater here, Felix. ETFs are and have been constructed to fool B&H investors, not attract them.

If you want the benefits of indexing, you buy an index (QQQ, MSCI, SPX, DJI, VIX, RUT, to name a few).

If you want more expenses–with the possibility that you’ll beat one of those–you buy iShares (IWM, IWN) or Vanguard (VTWx, VRTGx) or some other “tracking stock.”

If you want something that claims to be related to the above but gives you no good information about its balancing act, you buy an ETF. And then you look up in wonderment when you find out they overweighted NFLX…

Posted by klhoughton | Report as abusive

Felix, you’ve brought up the concept of “rising correlations” more than once, and I *still* don’t see how it jives with my experience. Very much wish we could sit down over a beer and hash out our disparate perspectives, because we are clearly coming from very different directions.

Correlations may be high, but they aren’t nearly perfect. If you have invested in Oracle over the past year-plus, your experience is very different from somebody who has invested in HP over that span of time. Over the past four years, those investing in WFC have lost 30%. Those investing in BAC have lost 90%. Over the past year, PEP and KO have opened a 10% gap, then closed it, then opened a 15% gap. All reasonably similar companies with VERY different performance. Isn’t that what price discovery is about?

I wonder if the observed rise in correlations has to do with the fact that the market is being driven by larger macro-concerns than in past years? European default is legitimately bad for nearly every company (to varying degrees). QE2 supports asset values equally broadly. This isn’t necessarily a failure in price discovery, but a recognition that values depend on much more than individual company execution these days.

Posted by TFF | Report as abusive

Seems like there should be a trading strategy in here: identify the least liquid, most correlated components of large ETFs. These stocks should be most susceptible to overshooting due to churning in the ETF. Buy when the ETF declines and sell when the ETF rises (I suppose you can short the ETF as a source of funds). What do you think?

Posted by TGDC | Report as abusive

“Correlations are very high!” says Felix. Well sure, I think, the principal uncertainty in the world’s economy is the question of when aggregate demand will recover. This obviously affects practically all businesses, so one would expect that the arrival of information bearing on this question would induce strong correlations.

But wait! Felix says it’s all due to the evils of ETF’s, and produces a graph of historical realized correlations, claiming that correlations today are higher “even than they were in the Great Depression.” A glance at this graph, though, shows the opposite: although the highest spike is recent, they area under the curve between 1931-1934 is clearly greater than between 2008-2011. And in fact, the JP Morgan paper linked through to by Felix shows that today’s realized 2Y (i.e. integrated) correlations are not much greater than those of 1987, never mind the depression.

What’s more the primary cause identified by the JP Morgan paper is the “macroeconomic environment” by which they mean exactly my initial guess.

How can you write so confidently about something you can’t read, Felix? You present with remarkable symptoms of alexia without agraphia.

Posted by Greycap | Report as abusive

well said, Greycap… I would have just said “correlation is not causation” – without the pun intended on the correlation of high correlations…

Posted by KidDynamite | Report as abusive

Mr. Salmon – you wrote that your feeling was that a nominal Tobin tax will drive away day-traders and rebalancing arbs. I concur with the intent – but won’t this stymie liquidity in the ETFs, causing bid-ask spreads to widen? Then the buy-and-hold investors “that ETFs are good for” will end up buying at a price other than net asset value – sort of like an load fund where the load is extensible, variable, and not terribly transparent … Why do those buy-and-hold investors care about short-term correlations / volatility anyway, if they are truly buy-and-hold? Perhaps human nature dictates they are not so buy-and-hold after all …

Posted by Dr_Stonewafer | Report as abusive

ps – Felix: “rebalancing arbs” don’t trade the ETFs – they trade the underlyings. so your punitive tax wouldn’t impact them.

Posted by KidDynamite | Report as abusive

@Dr_Stonewater – talking about rebalancing arbs, I believe Felix was referring to those who make money by trading the stocks that the ETFs will have to buy and sell in their quarterly/annual rebalancings.

you’re talking about NAV arbs – and that’s another good point – you WANT arbs to keep prices in line.

(you want “Rebalancing arbs” also, in fact, to take the price risk for you – they buy the stocks and sell them back to the ETF, lessening the ETF trading impact)

Posted by KidDynamite | Report as abusive

“First do no harm.” Let’s make sure that there is a real problem before we try to fix it.

To restate earlier comments, to blame ETFs is to blame a purely technical factor when there are obvious huge fundamental uncertainties in play which affect the whole market. And even on the technical side, there are also enormous futures markets, which probably have more effect on the market as a whole than ETFs. It takes much more than finger-pointing at a couple of graphs to make a convincing case here.

There may be smaller sectors where ETFs do have an effect (Gold?), but even then that does not make it in any way wrong for correlations to rise when there are market-wide or sector-wide effects.

Finally, just because things are different now from the past does not mean that the past was right and the present is wrong. Maybe the markets were more inefficient in the past. Is the time axis on that graph meant to imply that the rigged market of the 1920s is something we should aspire to?

Posted by Hayes | Report as abusive

If ETFs are causing securities to become mispriced, then delta-neutral arbs must exist aplenty. These are lovely when you can find them in liquid markets. Lever-up, put on the trade, profit. What’s not to love?

Actually, it’s quite a bit more likely that no such arbs exist. Any time someone claims the market is mispricing an asset, you have to ask how they came up with their valuation. Most of the time, it’s the valuation model, not the market that’s wrong. I think these two guys from the Kauffman Foundation are just cranky that the foundation’s endowment is getting its face ripped off in this market. It can’t be their models, so they have to find something to blame.

Real life is heteroskedastic. Models work right up until the point they don’t. Correlations go through cycles, too. There’s nothing you can do, except adjust. I mean, anyone who trades knows this instinctively.

Posted by haggers | Report as abusive

“Models work right up until the point they don’t.”

Models work until they come into broad use. Then they stop working. The consensus is thus always wrong.

Posted by TFF | Report as abusive

The thing that always seems to get lost here is that ETFs actually represent a tiny amount of trading compared to the volume of the underlying. S&P 500 ETFs currently hold 1% of the market cap of the S&P 500. Index funds in general however hold over 10%. That’s a trend that started back in the ’70s and has never waned.

Correlations are up, yes. But they’re also up between all asset classes, not just within large cap stocks. Are you suggesting ETFs are to blame for the increased correlations of Gold and Bonds too?

Finally, don’t forget that far more notional value is getting traded in the futures market than in ETFs OR the underlying, day in and day out, and that really drives the prices of the underlying. It was futures trading, not ETFs, for instance, that caused the infamous “flash crash.”

Posted by DaveNadig | Report as abusive

Golden Networking has created an instructional DVD for executives and professionals in high frequency trading, entitled, “The Speed Traders Workshop 2012”. This 4-disc DVD set walks professionals and non-professionals through the main issues, challenges and opportunities practitioners face to consistently capture alpha using fast computers. The DVD uses non-mathematical terminology and provides insights to successfully implementing speed trading while avoiding minefields. It is the ideal reference for professionals in the world of alternative investments. Visit http://bit.ly/znIl3E for more information

Posted by kaylabi | Report as abusive

Golden Networking has created an instructional DVD for executives and professionals in high frequency trading, entitled, “The Speed Traders Workshop 2012”. This 4-disc DVD set walks professionals and non-professionals through the main issues, challenges and opportunities practitioners face to consistently capture alpha using fast computers. The DVD uses non-mathematical terminology and provides insights to successfully implementing speed trading while avoiding minefields. It is the ideal reference for professionals in the world of alternative investments. Visit http://bit.ly/znIl3E for more information

Posted by kaylabi | Report as abusive
Post Your Comment

We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see http://blogs.reuters.com/fulldisclosure/2010/09/27/toward-a-more-thoughtful-conversation-on-stories/