Rule G-23:
There were a lot of great papers presented at the ninth annual Brookings Municipal Finance Conference that concluded yesterday. Two in particular stood out for me due to their implications for current regulatory issues.
MSRB Rule G-23 has been controversial since the MSRB rewrote the rule in 2011. Before that time, Rule G-23 allowed dealer Municipal Advisors to serve their clients as FAs, resign as FA when the deal was ready for pricing, and then, as long as the issuer agreed, either underwrite the bonds in a negotiated offering or bid on the competitive auction. Resigning as FA in order to underwrite a negotiated deal carries obvious conflicts of interest. But those conflicts are much less obvious in sealed bid competitive auctions.
The two papers from the Brookings event inform the discussion over whether G-23 should permit dealer FAs to resign and bid the deal, but their conclusions are conflicting. The first is by the MSRB’s sharp Chief Economist Simon Wu. Simon looked at trends in the number of bidders in competitive new issue sales and the effect that more bidders have on issuer deal execution. Simon “found that the average number of competitive bids received gradually increased over the past 10 years, from an average of 4.4 competitive bids per issuance in 2009 to an average of 5.7 competitive bids per issuance in the first half of 2019,” across location and deal size. That’s consistent with what we’ve heard from member firms about increasingly broader coverage of competitive issuers.
Simon also “found that the winning bidder’s primary offering spread was negatively correlated with the number of competitive bids received after controlling for characteristics of each offering, such as offering size, time to maturity and yield, etc. Therefore, all things being equal, soliciting more competitive bids does indeed improve an issuer’s selling price and reduce the yield cost for the issuer.” Of course that makes perfect intuitive sense, and it’s great to see the numbers back it up.
We’ve been trying to make this point for years in the context of Rule G-23. The more bidders at an auction, the better the execution for the issuer. This is especially true for issuers with little coverage. The benefits to an issuer of attracting a third, fourth or fifth bid can be significant, as Simon shows. That’s why restricting the number of bidders is not in the interest of issuers, even with a well-intentioned rule like G-23. Let the dealer FA bid the deal. There’s only upside for the issuer.
The second paper, by Dan Garrett at the Wharton School, tells a different story. Dan looked at the question of whether the 2011 changes to Rule G-23 affected issuer execution by comparing deals before and after the rule change and found they did, to the benefit of issuers. He concluded that when he dealer FA resigns and bids the deal, it discourages other bidders from participating. As a result, the bids are weaker.
Clearly Dan’s paper doesn’t support the point we’ve been trying to make that revisiting the G-23 prohibition on dealer FA bidding would benefit issuers. The question I ask myself, however, is whether the rise in the number of bidders generally that Simon found might offset some of the effect that Dan found during the 2008-2015 period.
And maybe this is stretching Dan’s conclusion a bit—he’d certainly say it is—but I wonder if his findings have implications for the conversation about whether MAs should be permitted to solicit investors in private placements. Dan found that issuer execution is best when the roles of MA and dealer are separate. When the SEC’s Temporary Conditional Exemption expires in December, it will provide an opportunity for people like Simon, Dan and others to examine the effects. I can’t wait. I love this stuff.
