ETF Flows Can Reveal Market Sentiment

If you want to ferret out signals from ETF flows, you have to dig deep into the data. Here's how to do it.

Dave Nadig's recent blog on the August ETF fund flows—Investors Fleeing To This Treasury ETF—provides a great overview of where ETF investors moved in aggregate in August.

The teaser for Nadig's article painted a dark picture: "If you're reading tea leaves in ETF flows, this one's ominous."

As Dave pointed out, three of the top 10 ETFs for inflows in August were the kind investors only buy when they are scared: the iShares 7-10 Year Treasury ETF (IEF | A-58), iShares 1-3 Year Treasury ETF (SHY | A-97) and iShares 3-7 Year Treasury Bond ETF (IEI | A-72).

Is the market panicking?

It's the kind of signal I see a lot of investors trying to draw out of ETF fund flows. Unfortunately, to truly separate the signal from the noise, you have to go a level deeper.

Investors Or Investor?

Let's take the most popular ETF last month: the aforementioned IEF, which pulled in $1.48 billion in new money in August. If investors as a group were flocking to this ETF, it would be cause for concern.

But one of the great things about the free ETF fund flows tool on ETF.com is that you can pull up a chart of any ETF over any time period and look at the flows both in aggregate and on a day-by-day basis.

To do that, just enter the ETF's ticker into the search bar, set your dates and click "submit." Then click the blue arrow next to each fund that says "Details" and you pop up a chart that shows each day's flows.

Here's the chart for IEF in August:

It's easy to see what sticks out: essentially all of the flows occurred on a single day—Aug. 18. To be specific, $1.24 billion out of $1.48 billion in inflows came on that day.

You see a similar pattern for the other two Treasury ETFs, with 95 percent of their inflows occurring on Aug. 1.

Here are the flows for SHY:

And here they are for IEI:

The fact that all those flows occurred on one day suggests that a single large investor made these decisions, as opposed to "investors" in general. In fact, we can probably even guess which investor made these bets: Good Harbor Financial.

Good Harbor is a large, well-run ETF strategist firm with $10 billion invested in quant-driven ETF strategies. It rebalances its portfolios twice a month—once on the 15^th and once at the end of the month—which coincides perfectly with the timing of these flows (due to settlement delays, fund flows don't always appear the day an investor buys a fund).

What we're seeing in the flows for IEF, SHY and IEI is Good Harbor getting nervous, not the market in general. That's a worthwhile signal—Good Harbor has a great track record—but it's just one opinion. And in fact, a quick glance at this morning's preliminary flows numbers suggest big outflows in both SHY and IEI, two of the Good Harbor targets in question.

Looking For Real Market Sentiment?

So how do you look for real sentiment swings in fund-flow data?

You need to dig a level deeper.

For starters, to minimize the Good Harbor problem, I look for ETFs that are regularly getting inflows. The tenth-most-popular ETF in August, for instance, was the iShares iBoxx $ High Yield Corporate Bond ETF (HYG | B-74). It captured net inflows on 12 of the 22 trading days last month. While that could be one investor dripping into the fund day by day, it's more likely a result of herd activity.

The next step is to look at competing ETFs to see if they attracted flows as well. If "the market" wants to buy HYG, it should want to buy all plain-vanilla high-yield bond ETFs. If only one high-yield bond ETF got inflows, it was more likely caused by a single investor.

The ETF most commonly compared with HYG is the SPDR Barclays High Yield Bond ETF (JNK | B-77). Using our fund flows tool again, I can see that JNK had $248 million in inflows in August, with positive flows on 10 different trading days.

Clearly, the market as a whole is embracing high yield.

Data In Segments

This concept of the market swarming around a particular theme is why we always report flows on the "segment" level. "Segments" represent the way we group together ETFs that compete to provide exposure to the same area of the market: all the Large Cap U.S. Equity ETFs, all the Corporate High Yield Bond ETFs, and so on. We have close to 400 categories in all.

Grouping flows by segment can reveal some very interesting data. For instance, while no single utilities ETF ranked in the top 10 for outflows, Dave's blog showed that U.S. utilities had the second-largest aggregate outflows of any segment last month, losing $1.9 billion. If you then drill down to look at the individual ETFs, you'll see it's a bloodbath, with all 8 ETFs experiencing outflows.

Now that's sentiment!

Sure, some of those ETFs show the same one-day outflow pattern as the Treasury ETFs mentioned earlier. But not all of them do. And when they're selling utilities from iShares, State Street Global Advisors, First Trust, Vanguard, Guggenheim, Fidelity and PowerShares, you know you've got something.

All this takes extra work and institutional knowledge. But no one said it would be easy. In a market where large actors can obscure the masses, you have to do your homework to ferret out what "the market" is really telling you.

Note: If you're interested in seeing the kind of data I look at to analyze flows, email me at [email protected] and I'll send you my spreadsheet.

At the time this article was written, the author held no positions in the ETFs mentioned. Contact Matt Hougan at [email protected].