The fees that bankers are paid to underwrite municipal bonds have been declining since financial markets started their recovery in 2009, prompting some dealers to focus on lower quality or more complex deals to boost profitability.

Through May 23, state and local governments have paid underwriters on average $5.16 per $1,000 of bonds sold this year, according to Thomson Reuters data.

That number has come down almost 17% since it peaked at $6.21 in 2009, when deals were riskier to underwrite and fewer banks were competing for business following the financial downturn.

The spread has steadily declined in the years that followed, dropping to $5.94 in 2010, $5.62 in 2011, and $5.52 last year.

Underwriters and other market participants attribute the squeeze in spreads mostly to low issuance of new bonds, which has heightened competition.

“We saw a noticeable difference in takedowns beginning in the middle of 2011,” said Todd Frazier a managing director and head of the pricing group at Public Financial Management Inc., referring to underwriting fees. “It’s no coincidence that in 2011 the market was down about 40%. I think underwriters felt the need to bring in business and to maintain market share.”

David Manges, a managing director and municipal trading manager for BNY Mellon Capital Markets, said dealers will continue to drive for market share as long as they’re making money.

“They’re going to drive to find deals on the competitive side, and on the negotiated side, bankers will push to get market share by reducing spread,” he said.

While underwriters compete on rate as well as spread, he said, it’s often easier to cut the spread before they cut the rate, since the rate has to be attractive enough to sell to investors.

“As underwriters compete for the issuers’ attention, one of the things they’ll talk about is their willingness to reduce spread to produce the lowest rate,” Manges said. “The lower the spread, the better the deal for the issuer.”

Matt Fabian, a managing director at Municipal Market Advisors, said the declining spreads may also reflect a more aggressive stance by financial advisors and issuers in general.

“This is a seller’s market, after all,” he said.

On the underwriting side, Fabian said that dealers are feeling competition because there is less issuance, with most of that issuance coming from larger issuers doing refunding deals.

Out of the $377 billion of municipal bonds sold last year, only $145 billion was new money.

In 2011, new money also totaled close to $145 billion, out of the total $286 billion of issuance. Before that, new money issuance surpassed $200 billion every year since 2002, according to Thomson Reuters.

Refunding bonds tend to be relatively short-termed, with maturities around 10 to 15 years.

This means the underwriting spread on these deals tends to be lower.

“As you shorten the maturity of these bond issues and the duration of total cash flows, the takedowns will also be compressed for that reason,” said Herman Charbonneau, executive vice president and manager of public finance at Roosevelt & Cross.

The “takedown” is one component of the underwriter’s discount, and is the compensation to the underwriter for distributing the bonds.

Charbonneau said that on deals today, he rarely sees a takedown over $5.00.

Other components of the underwriting spread include the structuring or management fee, compensation for risk, and any related expenses.

“So you’re looking at a relatively complex situation, where some of the decline is a decline in takedowns, generated by competitive considerations, and some of it is generated by the relatively short maturity of many refunding issues,” Charbonneau said.

In addition to the decline in the takedown component, the risk component of the underwriter’s fee has also contributed to the overall decline in underwriting spreads.

Underwriters take on some amount of risk related to interest rate fluctuation during the short period it takes them to underwrite the securities.

“With yields as low as they are and having dropped consistently now for the past three or four years, that risk component has dropped to a pretty slight margin,” Manges said.

In order to make up for lower spreads on some deals, underwriters may try to focus on other deals consisting of lower quality bonds where they can get slightly higher fees.

On the higher quality bond issues, underwriters will be more aggressive and tend to squeeze down on their takedown fee as they seek business and make proposals or bids to issuers, Charbonneau said.

“Elsewhere in the market where you’re trying to deal with the issuance of low quality bonds that are, say, at the bottom of the investment grade category, the work that you do as an investment banker has higher value,” he said. “There’s more of a skilled component that goes into the structuring of those issues and into the marketing process so you can charge a little bit more there.”

Focusing on issues where the skill component of the work makes a greater contribution to its profitability can help keep up their takedown and structuring fees, Charbonneau said.

“The lower the credit rating, the more speculative the credit, the higher the takedowns are, and that hasn’t changed,” said Frazier.

“I don’t think you can create those deals,” he said, “but certainly as they come up there is more competition among underwriters for those deals, but no more so than on any other deal.” MMA’s Fabian calls it a balance.

“A dealer needs enough lower risk bonds that, even with less compensation, will trade easier and pose less risk to their balance sheet,” he said. “But dabbling in riskier stuff isn’t a bad idea if done in moderation.”

Although underwriters have been receiving less compensation for their work, it doesn’t mean their work is any less valuable.

“At PFM, we maintain a database of takedowns and make sure the quality of the underwriting isn’t affected by lower takedowns,” Frazier said. “We haven’t seen that yet.”

He added that for the industry, the concern is that the broker dealer community will be pared down because underwriting becomes a less profitable business.

“A smaller universe of broker dealers over time may not be healthy,” Frazier said. “Long term, we want to maintain healthy competition.”

Though spreads have been declining, they have yet to come to close to the all-time low of $4.89 in 2008, according to available data going back to 1986.

In fact, prior to the sudden drop in spreads before the financial downturn, followed by a spike after, spreads had been steadily decreasing since 1986, when the average spread was $12.92.

Spreads came down between $7.00 to $10.00 averages in the 1990s, $4.00 to $6.00 averages after 2000, to the $5.20 average today.

“It’s low,” Charbonneau said. “It’s not a small decline — and every decline has some impact on profitability, but it’s not catastrophic.”

Breaking down the average spread among deal types, competitive issues offer greater fees than negotiated issues.

In 2012, the average underwriting spread among competitive deals was $6.17 and among negotiated deals it was $5.40.

Through May 23, the average competitive spread has come down to $5.29, but it’s still higher than the average negotiated spread of $5.12.

“This is because there’s more risk involved on a competitive deal than a negotiated deal,” said Peter Stare, a senior vice president at First Southwest.

In a negotiated deal, Stare said, the underwriter is not committing to underwrite the bonds until they have gone through the order period and have a balance of, for example, around 10%.

For a competitive deal, however, apart from any bonds sold in a presale, the underwriter is basically committing to underwrite the entire amount, taking on more risk.

He added that this recent decline in spreads is somewhat cyclical.

“I think we see this happen when spreads get tight and we experience a difficult market,” Stare said.

“It’s been pretty easy to be tight on spreads in the past couple years because we’ve basically been in a bull market,” Stare said. “but if the market starts to back up dramatically and people are still bidding the spreads that they’re currently bidding and start taking losses, then I think we’ll start to see spreads widen out.”

If interest rates start to go back up substantially, Stare said there’s a possibility that some underwriters may reevaluate whether it’s worth it to continue underwriting deals at such low spreads.

“It’s strictly a numbers game. If we get a protracted bear market in municipals, then I think you’ll have some people rethink whether they want to be in the market or you’ll see spreads widen out.”

However, he doesn’t think that rates will go up any time soon.

In addition to a rise in interest rates, Roosevelt’s Charbonneau said other reasons why spreads would stay down is if the average quality of bonds being sold in the marketplace rises, or if lower quality bond issuance is limited.

Manges said that spreads will stay low or go lower until one of two things ¬happens: either interest rates start to rise or there is a credit problem with some issuer or sector of issuers that makes it more difficult to place their bonds.

“Spreads can stay here or they can go lower. It’s a matter of competition,” Manges said. “Underwriters will continue to push to buy deals by reducing their spreads and they will do that until it proves to be unprofitable.”