December 09, 2015 Articles

Municipal Bond Pricing, Part I: Fair Pricing

The first in a two-part series examines the current standards for establishing fair prices, including identifying a market price for a security trade

By Pamela W. Peterson

Securities regulators have focused on unfair pricing of municipal bonds and other types of municipal securities within the past few years. See Press Release, Financial Industry Regulatory Authority (FINRA), FINRA Fines Oppenheimer (Dec. 9, 2013). FINRA found in the Oppenheimer case that in a one-year period, the firm (through its employee) priced 89 customer transactions from 5.01 percent to 15.57 percent above the firm’s contemporaneous cost. In 54 of those transactions, the markups exceeded 9.4 percent. Experience tells us that a regulatory focus on any particular area of the securities laws is often followed by a spate of private client claims based on those same rule violations.

In part one of this two-part series, we examine the current standards for establishing fair prices, including identifying a market price for a security trade. We also address some of the issues and limitations associated with establishing fair prices for municipal securities trading in practice, including the application of a markup or markdown by the dealer.

For municipal bond pricing, it is unfortunately easy to misconstrue both what regulators have said and what information is available to the public about municipal bond trading. The best source for such information is a website maintained by the Municipal Securities Rulemaking Board (MSRB) for Electronic Municipal Market Access (better known by its acronym, EMMA). The MSRB sets the rules (with approval by the Securities and Exchange Commission (SEC)) and issues interpretations; FINRA enforces them. The standard for fair pricing of municipal securities is that bond prices must bear a reasonable relationship to the market price, including markup, markdown, or commissions. MSRB Rule G-30 requires that dealers trade with customers at prices that are fair and reasonable, taking into consideration all relevant factors. The MSRB’s August 2, 2012, notice concerning Rule G-17 (the fair dealing rule) provides a relevant interpretation.

This “reasonable relationship” to market price is not—at least not strictly—a numerically based standard. All regulators have resolutely refused to define a simple percentage threshold for fair markups and markdowns or to suggest exactly how much of a variation from market price makes a customer price “unfair.” There is no bright-line test. The existing standard is referred to as objective because it references that legal myth, “reasonableness,” but reasonable people all too often disagree about what “reasonable” means, and when any asset class suffers a decline in value, the disagreements tend to get sharper.

So, as Americans anticipate that the Federal Reserve Board will eventually allow interest rates to rise (and this will, inevitably, cause a decline in the market value of many interest-rate sensitive securities), now seems an appropriate time to talk about the pricing of municipal bonds at the time of purchase or sale. How can a customer know if he or she is getting a fair price? If later challenges arise, how can that price be defended or attacked?

Market Price
First, let’s consider what “market price” means. Since 2009, when the MSRB established EMMA, the general public has been able to “see” the trading in municipal bonds. However, while EMMA’s trading data appear to be enticingly transparent at first glance, a closer inspection leaves the user with many questions. Because EMMA discloses several kinds of trades (new issue, interdealer, and customer—both institutional and retail), and because most municipal bond CUSIPs are very thinly traded, a great deal of what’s visible on EMMA can be misunderstood or misinterpreted. One aim of this article is to clear up as many of those misunderstandings as possible.

Why does it matter if a security is “thinly traded”? The simplest way to find a market price at a particular point in time, as far as regulators are concerned, is to look for a “contemporaneous” trade between two dealers. The regulators believe that sophisticated entities—trading between themselves with full experience and electronic access to relevant data including bid/ask data, credit quality information, and up-to-the-minute news—are capable of arriving at a sound market price. In fact, this is the simplest starting point for examining EMMA data: Look for a recent interdealer trade in the specific security that has no special condition indicator on it. (However, this won’t work for new offerings as there are no prior trades to review.)

But far more often than not, there is no reasonably contemporaneous trade of any nature. Why not? In the registered equities markets, it’s common to find millions of outstanding shares of any particular stock. To pick one example, in its 2014 annual report, Microsoft reported 8,299 million outstanding shares. And because equities are traded on exchanges, with readily available price information, it is very easy to identify a market price that incorporates millions of trades that reflect the investment beliefs of a diverse set of market participants. Internet access is really all that is required to find and verify a market price for a public company.

Contrast this with a fairly typical municipal security. There are about 1.3 million outstanding municipal CUSIPs, three orders of magnitude more than outstanding registered equities, so “typical” is understandably hard to define. However, one of the largest issuers (by outstanding par amount of bonds and also by frequency of issuance) is the state of California, so let’s look at some of its recent general obligation debt (a large issue by a large issuer).

On March 3, 2015, California issued a set of general obligation and general obligation refunding bonds with an aggregate par value of $1,944,865,000. This example is deliberately chosen to emphasize that it is not only small issuers and issues that are thinly traded. There were 17 separate maturities ranging from 2016 to 2045 in this issuance. Consider the bonds that mature in 2027 (CUSIP 13063CQU7). The total par amount of this single maturity was over $63 million, which is quite large for a single maturity. Smaller issuers rarely have a single maturity of this size. Looking at the trading history on EMMA, we see that since the issuance date, there had been a total of 50 trades in this CUSIP as of August 31, 2015. If we filter out all trades over a par amount of $100,000 (such larger trades are considered to be institutional customer trades), we find a total of 12 retail-customer-size trades. See the trade activity here. You must expand the 5/1/2015 trade information to see that it’s two $50,000 par-amount trades. On average, that’s two retail-customer-size trades per month. Not to mention the fact that there are slightly over 12,000 outstanding bonds bearing this CUSIP, not millions or billions of security units, to be traded.

So, unlike finding the market price for equities, which is established by second-by-second trading, finding the market price for municipal securities—even large par-amount issues by well-known issuers—will likely require other methods. Most municipal securities are thinly traded. Relying on trading data—which are thin to begin with and which may be days, weeks, or months old—is usually insufficient.

The supplementary material for MSRB Rule G-30 refers to the use of “all relevant factors” to determine the fairness and reasonableness of prices. What are those other relevant factors? Section .02 of the Rule G-30 Supplementary Materials states, “The most important factor in determining whether the aggregate price to the customer is fair and reasonable is that the yield should be comparable to the yield on other securities of comparable quality, maturity, coupon rate, and block size then available in the market.” (Emphasis added.) If a dealer can show, for example, that it established a market price for a retail customer who wants to make a secondary market purchase of Iowa school district bonds by looking at several issues of other similar Iowa school districts and found some relatively recent trades where the credit rating was the same, the maturity was nearly the same, the coupon (nominal interest) rates were close, and the par amounts of the trade were at retail-trade sizes, it has likely established a case for having found a market price.

Also, among other things, section .03 of the Rule G-30 Supplementary Materials allows a fair and reasonable price to be based on “the best judgment of the dealer concerning the fair market value of the securities when the transaction occurs and, where applicable, of any securities exchanged or traded in connection with the transaction”—a shaky support on which to rest a fair-pricing defense. Experienced traders often do have an excellent sense of market value (like baseball scouts or racehorse stock agents or antiques appraisers), but it’s very hard to demonstrate how they arrived at good, but intuitive, decisions.

Also, often only some of these relevant factors can be found for otherwise highly comparable bonds. A dealer might then widen the search. If Iowa school district bonds aren’t trading, maybe Iowa water and sewer district bonds with similar ratings, maturity, coupons, and block sizes are. Maybe school district bonds issued by the neighboring states of Minnesota and Missouri have traded recently. Maybe the dealer can find Iowa school district bonds that have similar credit ratings but fairly different maturities, and apply a simple mathematical yield-curve analysis to get a good estimate of what the yield (and, therefore, the price) on these bonds should be.

Real life, however, is sometimes stubbornly noncompliant with regulatory theories. As indications of comparableness become fainter and fainter—the bonds are less similar, the trades farther away in time, the coupons less similar because the reference bonds were issued under far different interest-rate conditions—the dealer can be faced with unpalatable choices. It may decide to take the legal risk of basing its market price on somewhat tenuous data. It may entirely refuse to do the trade, informing the customer that it is unable to determine a market price and therefore cannot assume the risk of trading at a price that may be retroactively determined to be unfair. In some cases, dealers have advised the customer that there is no way to determine a market price, but, if the customer acknowledges that a diligent search cannot produce evidence on which to base any conclusion about market price and still desires to seek bids or look for offers, the dealer will facilitate the trade on an as-agent basis with a fully disclosed commission. The legal risk to a dealer who acts as a principal when it has no hard data to defend its price is considerable. By acting as agent and informing the client of the pricing problem, the dealer at least allows a client who must liquidate a position a way to do so.

This brings us to the second consideration in finding a fair aggregate price with a reasonable relationship to the market price for a retail customer trade: How much is the dealer making on the trade? The amount of markup, markdown or commissions charged must not lower the yield so much that it is no longer reasonably related to market yields. How much money should a brokerage firm be allowed to make when a customer is buying or selling a municipal security? What’s a fair markup or markdown?

It’s here that the data are most often used incorrectly. The typical misunderstandings that arise from examining EMMA data come from several sources. First, there’s the failure to understand that EMMA shows at least three kinds of data: interdealer trades, retail customer trades, and institutional customer trades. Further confusing the picture, EMMA also reflects trades in which the customer is not charged any transactional fee because he or she has an advisory relationship with the dealer and pays either a fixed or a percentage-of-assets annual fee. (This will change on May 23, 2016, when a special condition indicator for such trades will be added—no help for those examining historical data, however.) The pricing is different for interdealer, retail customer, advisory customer, and institutional customer trades for reasons discussed below. Second, because trades are not instantaneously reported but may be reported within a 15-minute window, it is entirely possible that the earlier of two close-in-time trades shown on EMMA may have not been “visible” to the parties who made the second trade. If the price for each of those trades was independently arrived at with different assumptions, the second trade might appear to be “unfair” even though the mechanism used by the second dealer was actually reasonable based on the information available to it at the time it set a price.

How Much Can a Dealer Charge?
Now let’s think about markups, markdowns, and commissions. Much has been made of the fact that the vast majority of dealer-customer trades are done on a principal basis, and the MSRB rules do not (currently) require that a dealer disclose markup and markdown on a principal trade, only the aggregate price (and yield). The suspicion is that unfair markups and markdowns are concealed by this practice (the MSRB has a current proposal to disclose markups and markdowns on certain close-in-time trades). However, principal trades often benefit customers by allowing faster sales or purchases when speed of execution is paramount or by providing liquidity for relatively unknown bonds. In contrast, if the dealer acts as an agent, facilitating but not taking one side of the trade, it charges a commission and must display it on the trade confirmation. This commission disclosure must be valued against the serious impairment to liquidity that customers would face if all trades were required to be done on an as-agent basis.

At least one large municipal dealer began voluntarily disclosing these numbers on confirmations in 2014, and nothing prevents a customer from asking to have the markup or markdown disclosed. The resistance exhibited by some dealers is not against providing this information to the customer but against publicizing a number that may trigger “price war” competition.

In part two of this series, we will continue our examination of the issue of fair pricing, with a specific focus on the use of EMMA trade data in establishing fair pricing and the problems that arise when trying to interpret those data. We will also examine common practices in determining markups and markdowns on municipal transactions, and the underlying economic rationale behind those practices.

Keywords: litigation, securities, fair pricing, municipal bonds, fair and reasonable, FINRA