Bubble In Municipal Bond Closed-End Funds Warns Of An Overheating Market

| About: iShares National (MUB)

Summary

Municipal bond closed-end funds are trading significantly above average discounts to NAV.

This trading behavior occurs less than 2.25% of the time based on z score.

Such a rare occurrence speaks volumes about the frothiness of the market at large.

It is common to see individual closed-end funds (CEFs) trading at a premium or discount to NAV. It is also fairly common to see individual CEFs trade away from their long-term average discount/premium. But when the average discount of all 176 municipal bond CEFs trades at 2 standard deviations above average, we know for a fact it is unusual. It is also slightly concerning, not only for investors in CEFs, but also for investors as a whole. Could this infrequent event warn us of over-exuberance in the broader market?

The "Munis"

A municipal bond is issued by a state, municipality or county. They come in all shapes and sizes, but their main appeal is their ability to provide tax-free income. A coupon of 5% will yield more after tax compared to a 5% corporate bond.

Closed-end funds allow an investor to diversify and spread the risk over a range of munis. They also focus on specific types such as high yield or munis from certain states. When choosing a CEF, there are many considerations, and an investor must look beyond only yield and discount. Use of leverage, size of the fund, and distribution ratio are all important, but none of these will help you when muni closed-end funds are trading in a mini bubble.

Munis CEFs Are Overvalued

There is strong demand for municipal bonds and this has kept yields low and prices high. This is normal market behavior, and I have no way of saying if current bond valuations are fair.

Closed-end funds are somewhat easier to value. They track the NAV of the municipal bonds they hold no matter if that NAV is high or low. There are always discrepancies; no fund will track its NAV 1:1, but there will be an average discount or premium that is followed for the majority of the time.

The z-statistic compares a fund's current discount/premium to the average discount/premium. The calculation is as follows:

Z = (current discount - average discount)/standard deviation of the discount

A negative z score means the current discount is lower than average, and vice versa. More importantly, it shows us how significant this discount is. A fund is expected to trade between a z score of -2 and +2 for 95.5% of the time. So when we see outliers, we can be fairly certain there is a temporary mispricing.

Individual funds can trade with outlying z scores quite often for all kinds of reasons. What is more meaningful and unusual is the current z score average for all funds. Below you can see figures relating to all 88 national focused funds. These tend to be more liquid and reliable as an indicator.

Source: Author's database

The average z score is 2.08, and we know that this happens less than 2.25% of the time.

Whatever is driving this trading behavior can only be guessed at. What we know for a fact is that the current situation is unlikely to continue and the discounts should return to average levels.

A Basket Of Funds

Another way to detect mispricing is to construct a basket of funds and compare it to the benchmark Muni Bond ETF, MUB. The MUB has a very high correlation to the NAV of all the CEF holdings, which makes it a reliable comparison.

We constructed a portfolio of 10 CEFs and compared its profit to that of MUB. Each portfolio is constructed in a way that the profits of each portfolio move with a slope of 1.

Source: Author's software

Throughout 2015, the muni CEFs (portfolio 1) consistently underperformed MUB (portfolio 2). Any deviations from this behavior were small and temporary, moving in a 1 standard deviation range. But during 2016, profits generated by MUB have stalled while our muni CEFs have continued higher. The extent of this is best illustrated by the chart of profit/loss of portfolio 1 long vs. profit/loss of portfolio 2 short. This is trading at over 2 standard deviations from average.

So what has changed in 2016? A higher valuation of portfolio 1 may be justified if for some reason its NAV were outperforming MUBs. But the last two charts show this not to be the case. The NAV of portfolio 1 is actually decreasing in comparison to MUB as its valuation increases. The last part of the CEF rally follows no logic.

How I Trade It

The standard deviation proposed is not large enough for a nice hedge trade. I am shorting some individual muni closed-end funds, but I have to be at my desk all the time due to the risks (see below). The idea of the article is not to make specific trade recommendations, but rather to make the reader aware of illogical buying in this part of the market. I would not buy municipal bonds or any closed-end funds associated with them, and I am cautious on the market as a whole.

Risks

The limited potential and liquidity in such low volatile instruments is a big risk. Shorting some of the low volume funds can turn out to be a real disaster, so you must constantly monitor the hard to borrow rate. A good example would be the behavior of BHV.

A 60% chance of success with a risk reward ratio of 2:1 is not a realistic expectation for a short. In reality, the risk is much larger and hard to measure in advance. You may end up being right in general, but still make pennies from the trade.

Conclusion

A reasonable buyer will always try to shop around for a good deal. In municipal closed-end funds, this is not currently possible, but that does not stop the buying. Closed-end funds are easier to evaluate than most of the instruments traded on the exchanges and they are giving us a warning on the market at large.

After receiving some really nice arguments in the comments section, I decided to show the following chart:

Here you can see how the average premium/discount of 9 of the mentioned funds in the article is distributed for the last 998 trading days. I did not make any further calculations to determine if the distribution is normal, because it is quite obvious. Even this does not add any doubt in my trade, because I have risk management and reaction, but hope it might be helpful for readers. The average discount is -1.6% and the mode is in the range 8-11%. It is hard for me to interpret this 998 days data and it is up to the reader to make his/her conclusions.

Disclosure: I/we have no positions in any stocks mentioned, but may initiate a long position in MUB over the next 72 hours.

I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.

Additional disclosure: I am short BTA, IIM, VKI, VGM