The observer effect on muni ETFs

By Felix Salmon
January 14, 2011

How much are municipal bonds worth? There are lots of indices purporting to follow the market, but these days all the attention is on exchange-traded funds, which are plunging alarmingly. The volatility in municipal ETFs has already caused Vanguard to pull its plans to issue three more such funds, and the total amount of money in them seems to be falling fast:

Individual investors, who represent two-thirds of the market, withdrew more than $13bn from muni bond mutual and exchange traded funds in the last two months of 2010, an outflow that exceeded the sums they cashed out at the height of the financial crisis in 2008, Thomson Reuters data show.

When individual investors only owned specific bonds or muni mutual funds, outflows in the muni market were much less visible than they are today. With ETFs, outflows cause immediate and significant price drops, and that kind of volatility can in turn prompt even more selling from muni investors—who are, after all, risk-averse by their nature. Aaron Pressman writes:

Amid the turmoil, ETF investors have complained that some muni ETFs are not properly tracking their underlying market indexes. During the fourth quarter, for example, the share price of the $2 billion iShares S&P National AMT-Free Municipal Bond Fund, closed as much as 2.4 percent below the value of its assets, according to the iShares web site. The fund closed on Wednesday at a 0.5 percent discount.

Tracking problems arise because muni ETFs own only a portion of the thousands of bonds included in their underlying indexes. Fund managers try to create a representative sample of bonds but the technique sometimes goes astray when the market moves sharply.

But Dan Seymour has a different take, in a fascinating article about the increasing disconnect between municipal bond indices and the ETFs which try to track them, and comes to the conclusion that the ETFs actually do a much better job of price discovery than indices do—given that the indices, by their nature, comprise thousands of bonds which simply aren’t trading at all.

In other words, it’s not the fact that ETFs don’t own all bonds in the index which is the problem here, so much as the fact that the price of the index is a largely arbitrary measure, determined by implausible pricing services with the impossible job of evaluating the value of bonds that aren’t trading.

Financial markets often operate on the basis of a deliberate refusal to mark to market. In a crisis, all banks are insolvent on a mark-to-market basis, since there’s no real bid for that kind of quantity of loans. But they keep on servicing the loans, mark down the really bad ones, and often manage to come out the other side.

Similarly, in the municipal-bond market, in times of turmoil issuers simply stop issuing, and investors just keep hold of their bonds. Eventually, the market reopens, and things get back to normal; the official indices need not have moved very much at all in the interim. Many individual investors need be none the wiser; and as far as institutional investors are concerned, their best course of action is to simply sit tight, rather than try to sell into a bidless market.

The rise of the muni ETF changes all that. ETFs trade every day, and everybody knows their price. If bids dry up, as they’re doing right now, the price falls, in a very visible manner, to whatever the market-clearing level might be. In turn, that fall in price only serves to exacerbate the problem and the panic.

The observer effect, in physics, refers to the way in which a phenomenon is changed by the act of observing it. ETFs would love to be neutral reflections of the state of the muni market, but they’re not: they change the dynamics of the market dramatically, and not necessarily in a good way. Munis aren’t as safe, now, as they used to be, and the existence of ETFs is one reason why.


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

Muni ETFs have more visible volatility, but they are also much more liquid. In the old days, during a market crisis there was no way to sell your bonds at a reasonable price, right when you may very well need cash. In that sense munis are much safer now than before.

Posted by Stevensaysyes | Report as abusive

“Munis aren’t as safe, now, as they used to be…”

Maybe in the trader’s growth-orientation context, but I think most buyers (OK, individual buyers) of bonds (v. ETFs) are looking at tax-free income with a very strong prospect of getting their money back when the bond matures or is called.

Meredith Whitney notwithstanding, most munis (state & local) are very “safe” (ie–return of principal) investments. Still, I wouldn’t buy CA, NV, FL, MI, and a few other state/local bonds.

Posted by Lilguy | Report as abusive

Just hang in there and hold on to your Muni Bonds & ETF’s. The big guys are buying all they can and will make a killing in a few months once the fear is over.

Posted by ZipQuote | Report as abusive


Aaron Pressman here. Great point, which was also made in the longer, updated version of my story which bizarrely our own web site doesn’t offer up instead of the one you linked to. It’s here 145220110113

And includes this bit: The intraday discounts arose because trading in the ETF was actually more active and more visible than private trading of the underlying bonds among broker/dealers, according to Matt Tucker, head of fixed income strategy for iShares.

“The ETF has allowed a view into the municipal market on a real-time basis for the first time,” Tucker said. Typically, the bond prices catch up to the ETF share price within days, he said.

Posted by ampressman | Report as abusive

Real time, but not real investors. Mom and pop own bonds or muni funds, not ETFs.

Posted by maynardGkeynes | Report as abusive

It’s marginal price setting via a tiny fraction of beneficial owners.

The price accuracy of another $1.X trillion in bonds might arguably “suffer” from a lack of instantaneous MTM, but then again, it’s arguably that a huge swath of their owners aren’t interested in selling, either. The muni market’s unusual heterogeneity means that massive selling of “benchmark” bonds used in ETF proxies may not have nearly as much a relationship to the actual bonds and their holders of much small bond deals from completely different municipalities in other regions.

We reject such thinking when looking at Treasuries or other very-like corporate issues because they are sufficiently comparable that, notwithstanding a major credit difference, they “should” trade the same based on interest-rate discounting. The same doesn’t always apply with municipals.

Put another way, relatively small volume + marginal price setting may not deliver any more “accurate” a price than matrix/comparative + evaluative price setting when it all comes out in the wash.

Posted by fixedincome | Report as abusive