With bid-ask spreads for retail-sized trades averaging over 200 basis points and nearlythree to five times the size of spreads on institutional-sized trades, retail investorsin the municipal market lose about a half-year's worth of return on the average triple-Abond just to transact, according to a forthcoming study in the Journal of Fixed Income.
Selling a retail-sized block of muni bonds - $100,000 or less - without taking a seriousdiscount, , takes three times longer than an institutional-sized piece of $500,000 ormore, the study authors contend.
"Punishing" as this environment may be for retail, these findings are no surprise, saidlead author Arthur Warga, who is dean of C.T. Bauer College of Business at theUniversity of Houston and a former Lehman Brothers analyst. "It tells us there aredifferent markets for the two."
Retail investors drive most of the market's trading activity and account for 40% to 50%of the dollar volume in the muni market. As of March 31, individuals held more than athird of the $1.9 trillion total municipal bonds outstanding, according to The BondMarket Association.
"Municipal bonds, unlike the more standard government securities, are backed bydifferent cash flow protections and issued in quantities that are often very low andsmall in value. And that creates problems for a marketplace that is really structured totrade in institutional-sized transactions," Warga said.
Using 33,925 individual bid-ask spread pairs representing 78% of the total dollar volumeof trades in May 2000, Warga and co-author Gwangheon Hong, assistant professor at theCollege of Business at Saginaw Valley State University in Michigan, found that forretail-only trades, sized at $100,000 or less, the mean percentage spread was 246 basispoints. For institutional-only trades, sized at $500,000 or more, the mean spread was 79basis points.
What then is an acceptable spread?
While the Municipal Securities Rulemaking Board's G-30 rule says brokers and dealersshould charge prices that are "fair and reasonable," with "1.5 million differentmunicipal bond issues outstanding, it would make little sense for the MSRB to adopt asingle numerical guideline for such a diverse market," MSRB executive directorChristopher Taylor said. He added that the SEC has taken action against brokerages forspreads in the vicinity of 3%.
The National Association of Securities Dealers fined eight brokerages $610,000 in Junefor abusive pricing. The firms were fined for marking up securities by as much as 122%,ranging down to trades with spreads of 6.85%.
Hong and Warga's study, which uses MSRB trade data from May 2000, is biased toward themost liquid municipal issues, according to Warga, who also directs the Fixed IncomeResearch Program that maintains databases on bond pricing.
"It's likely that even wider spreads could result from the municipal issues we ignoredas a result of our experimental design," Warga added, though his is the second studythis year arriving at a similar conclusion on retail spreads.
A Securities and Exchange Commission study released in February by Lawrence Harris andMichael Piwowar also found that bid-ask spreads for retail-sized transactions averagewell over two percent of par value. And in a related and soon-to-be-published study,Richard Green, Burton Hollifield, and Norman Schurhoff at Carnegie Mellon Universityfind that dealers "operate as oligopolists" and earn higher average markups on small andmedium transactions, while behaving very competitively in the institutional market,Green said.
According to the TBMA, the "advent of real-time transparency in January [2005] willsignificantly enhance investors' ability to make wise decisions." Then, broker-dealerswill report prices and other data from most of their trades to a clearinghouse within 15minutes after a trade occurs.
But during testimony to the Senate Banking Committee in June, Warga said that"transparency has had virtually no effect whatsoever on the costly retail structure,cost structure for retail trades and municipal bond market," citing a 2003 studyVanguard Group's Christopher Ryon conducted. Ryon's study used one day's worth oftransactions, about 30,000 trades, and showed that spreads haven't shrunk since moreprice reporting has become available. The MSRB began reporting some transactioninformation in 1994, and has increased the amount of information available since.
TBMA said it plans to make the real-time reporting data available free-of-charge on itswebsite, www.investinginbonds.com.
"We're all waiting to see what the impact of immediate reporting is," Warga said.
"Just because you have price reporting doesn't mean there's a relevant trade for you toobserve," Warga said, as most bonds are bought and held, and have "no reason to trade."
"But where bonds do trade, [there will be] a benchmark, and retail investors deservethat," Warga said.
Even if a particular bond doesn't trade often, "getting information about comparablebonds is extremely valuable, and in the case of muni bonds, there are lots and lots ofcomparables," Carnegie Mellon's Green said.
Excluding situations where, for example, bondholders have to dispose of large quantitiesof bonds for reasons having nothing to do with the bonds themselves, ultimately, thebenefits of price reporting outweigh the disadvantages, Warga said.
"It will employ and provide a much richer database for mutual funds to base their netasset value calculations on. It will also give banks the ability to monitor whethersomething is going wrong or not," Warga said.
