Christine Benz: Hi, I'm Christine Benz for Morningstar.com.
The ability to trade exchange-traded funds throughout the day has led some market watchers to conjecture that ETFs could lead to bad investor outcomes.
Joining me to share some recent research on this topic is Fran Kinniry. He is a principal with Vanguard's Investment Strategy Group.
Fran, thank you for being here.
Fran Kinniry: Thank you, Christine.
Benz: You recently looked at the dollar-weighted returns in exchange-traded funds versus traditional mutual funds, and you compared the results.
First, let's talk about what dollar-weighted returns are, and how they're different from the familiar total returns that investor see when they look at their funds.
Kinniry: Sure. Just to level-set everyone, funds, if you go look on the web or you look on Morningstar, you get published time-weighted returns, which is known as TWR, the abbreviation. That's just the return of the fund or the investment itself. It does not take into account any cash flow in or out.
Benz: So that's just this basket of securities, what it returned?
Kinniry: Basket of securities, what it returned. And the easiest example, if I put a dollar in on Jan. 1, no [other] cash flows in or out [during the year], my time-weighted return would then equal my dollar-weighted return, because I had no cash coming or going.
The second form, of which Morningstar has done a great job and other providers are starting to look at is, well, how did the investor do? And the investor returns are known as IRR or internal rate of return. What that capture is, well, how did the actual dollars go in? In my prior example, I said if I put a dollar in on Jan. 1, and held it, those two would be equal. Now, suppose that I put a dollar in on Jan. 1, but then I put another dollar in on June 1. And if my returns in the first half of the year were different from my second half of the year, I would have a different result in my dollar-weighted return IRR to TWR.
Benz: … So you looked at flows into ETFs and traditional mutual funds. I think one of the assertions about ETFs is that they would encourage this wild trading, that investors might be inclined to buy the hot ETF and sell it at an inopportune time.
What did you find when you examined the dollar-weighted returns versus the time-weighted returns for ETFs and traditional mutual funds?Read Full Transcript
Kinniry: What we found is, like a lot of the findings in the investment business, there is a lot of intuition. But when you actually go do the research, the intuition may not match the experience. What I mean by that is, there has been a lot written about ETFs being bad for your health, or because of the intraday trading and the ability to get in and out, that investors would misuse them.
So, our research tried to look at where the IRRs and TWRs of ETFs were really doing exactly that. What we found is, no, that's not the case. It's very time-period dependent, meaning that if you look at it in one period of three- and five-year returns, and then just move it to a different three- and five-year returns, the results flip 180 degrees.
Benz: Why would that happen?
Kinniry: We've learned a lot about IRR to TWR, regardless of ETFs [or mutual funds]--because Morningstar and others have tracked IRR/TWR on traditional mutual funds for a long period of time. What we've seen is that the IRR/TWR is very time-period dependent on the type of performance in the market. Meaning that as long as the performance continues to do well, my earlier example of a dollar in [in January] and a dollar in [in June]--if the first half of the year, the market was up 5%, and you put another dollar in on June 1, and the market continued to go up another 5% or 10%, the IRR could actually be higher.
Really when the IRR/TWR starts to go lower is when you have a cyclical or mean reversion. … Now think about more money coming in and then you get the reversion. And so IRR/TWR has always been time-period dependent. In momentum markets, where the market is continuing its current cycle, we'll see IRR actually be above TWR. But then when it does revert, you have peak assets, and then you have declining returns. That's when you really start to see IRR go much lower than TWR.
Benz: You also found that, in addition to it being time period dependent, it's also dependent on the asset level within a fund and the magnitude of flows within a fund. Let's talk about that piece.
Kinniry: Yes. So, not only is it time-period dependent, if my earlier example of the dollar, what really matters is cash flow in, relative to base assets. So, there were a lot being critical of Vanguard … looking just at Vanguard traditional index versus Vanguard ETF. And think about our traditional index funds--they've been around for 20 to 30 years. So they have large base assets and still strong cash flow, but cash flow relative to base asset is quite small.
Now, flip that to something that's new: ETFs. Even if we had the same cash flow, positive cash flow in ETFs, as we did in our same mutual funds, let's say the S&P 500 ETF and the S&P 500 mutual fund, let's say they both took in $5 billion in a year. Well relative to base assets, that's a huge difference on a fund that's been around for 30 years than a newer fund that may [have been] around only five or 10. So, the time period, the performance of its momentum or mean reverting, and cash flow to base assets are the primary drivers in IRR to TWR.
Benz: Did you see any differential in fund types? For example, did the more volatile fund types tend to have worse internal rates of return than less volatile fund types?
Kinniry: Absolutely. We confirmed a lot of the research that Morningstar has done, and some of our own work confirms that the more narrow the investment space, so let's say, sector funds or concentrated funds. So, the more narrow and the more volatile the return streams tends to be where investors really hurt themselves. When we say hurt themselves is when we see IRRs to TWR is being very wide. And they can be positive for a while. So, even our own sector funds at Vanguard, we have seen our energy and health-care funds where the IRR to TWR is significantly higher, but then when you get that mean reversion, you have peak assets, and you get the full blunt of the negative returns.
So, what we saw is that the more narrow the investment landscape, sector funds, concentrated funds, and the more dispersed the returns, the wider the IRR/TWR is.
The conclusion to that is the more investor behavior matters, the more investors may be harmed by concentrating in sector funds or more narrow segments. When we looked at holistic funds, something like the Total Stock Market or the Total Bond Market or even single fund solutions balanced mandates...
Kinniry: Target-date funds, we see that the IRR to TWR is extremely tight, which really means that it's very, very good at minimizing, if not eliminating, investor behavior.
Benz: Well, Fran, thank you so much for sharing your insights on this important topic.
Kinniry: Thank you, Christine. I appreciate it.