Last Friday evening we posted a comprehensive report on municipal bond mark-ups. This week we've had several posts covering topics within our report. We're wrapping up the week with an example which illustrates some of our observations.
In February 2005 the City of Carlsbad issued $33,085,000 in tax exempt bonds underwritten by a Stone & Youngberg. This small San Francisco-based brokerage firm specialized in municipal finance and was recently bought by Stifel Nicholas. The Offering Circular is available to view online.
The issue included $10,840,000 in 5.2% bonds due September 2, 2035 issued at $99.694 to yield 5.22%. The trading in this series is available on the EMMA website. The bonds were sold to investors between February 2, 2005 and February 8, 2005 on a when issued basis to settle on February 17, 2005.
We capture a screen image of the MSRB EMMA trades below.
42 small trades were done over 5 or 6 days at exactly $102.319. The average inter-dealer price during this period was $100.00 and so the markup charged against this benchmark was 2.32%. Excluding these 42 customer purchases done at exactly the same price, the average customer purchase price was $99.69 and so the 42 investors who were sold these bonds paid $2.63 more than investors buying the same bonds at the same time.
Can a markup of 2.3% to 2.6% be justified on these trades? Almost certainly, No.
This difference cannot be explained by the relatively small sizes in the 42 suspect trades since there were also trades of $10,000, $20,000, $25,000 and $50,000 all executed at $99.692.
The 42 trades must have all been done by the same broker-dealer because of the uniformity of the $102.319 price. This spread is over $25,600 and reflects either additional compensation beyond the disclosed spread to the underwriter, Stone & Youngberg, or a riskless profit earned by a different broker-dealer who purchased $1,000,000 at $100 and was selling the bonds instantaneously at $102.319.
That the 42 trades were done over 5 or 6 days and at the same price to the tenth of a penny makes clear that the dealer determined the price customers would be charged and sprinkled the bonds throughout customer accounts.
These trades highlight potential abuses which are difficult to spot with just the public facing EMMA data we've been reporting on this week. For example, imagine this trading by the same dealer in 42 other similar bonds. If the dealer purchased $975,000 in one bond in each of the 42 accounts a 2.6% markup would be clearly excessive. Even breaking the 42 $975,000 positions into 126 $325,000 trades sprinkled 3 trades into each of 42 accounts would not be disperse enough to make the 2.6% markup appear not excessive. The trading in small lots we observe in this bond may have been purely to disguise the excessiveness of the markups being charged.
We see other examples where almost the entire series is sold at a substantial markup to the IPO price by the underwriter. In many cases, the markups are earned by the underwriter swamp the disclosed underwriter spread. Issuers need to be as concerned about these markups as investors. Issuers should check the early trading in their newly issued bonds to make sure the underwriters and advisors haven't caused them to significantly underprice their bond offerings thereby wasting taxpayer money.
