With secondary market trading data from EMMA, hospitals have another tool to check yields on their bonds and see if they left money on the table. After reviewing 275,000 trades from bond issues going back to 2015, we share our findings and some tips on how to ensure best pricing.
Management guru Peter Drucker is quoted saying that you can't manage what you can't measure.
But in the public bond markets, where yields set by underwriters on day of pricing can cost big bucks over the life of the debt, measuring is easier said than done.
On pre-pricing calls, underwriters will show how similar issues priced, comparing proposed credit spreads over the "AAA" MMD index. But even within a single sector like hospitals, credit spreads can vary significantly from one bond issue to another depending on factors including ratings, maturities, security, call features, even state of issuance. There may be less than a handful of true comparables, making the comparison unreliable.
On the day of pricing, the hospital may be told the offering is a great success and is oversubscribed by 3 or 4 times. Not only is this hard to verify without seeing the order books, but investors routinely inflate their indications of interest when they anticipate getting only a portion of what they ask for, so oversubscription doesn;t mean much. If demand actually exceeds supply, how does the hospital know if the underwriter took advantage of the imbalance and adjusted yields accordingly?
Hospitals shouldn't give up just yet, because there is another way to measure underwriter performance: reviewing prices paid by investors buying and selling the bonds in the secondary market.
If investors view a bond as underpriced (the bonds pay a higher yield than what investors would have been content with on pricing day), they will bid it up in secondary trading. The opposite is true and if the bond is overpriced: investors will pay a lower price in secondary trading. The higher the markup, the more the hospital should be concerned about yields being above market.
"The higher bonds trade relative to their initial offering price, the more likely the hospital left money on the table."
We started our analysis by fetching market trades on the EMMA website. Most publicly-sold tax-exempt bond issue offering statements are posted there along with secondary trades whose yields are reported based on the "yield to worse" convention.
Above: Sample trade activity from EMMA website (emma.msrb.org).
We went back to January 2015 and retrieved 275,000 "Customer Bought" trades that took place in the first 30 days following initial offering for 368 hospital bond issues.
Before we get into details, here are some important things to keep in mind:
- Bonds must be traded. Issues that tend to go to "buy and hold" investors will show minimal trading, some no trading at all. We excluded issues where Customer Bought trade volume represented less than 5% of issue par in the first 30 days of trading.
- Prices must be adjusted for changes in rates. Before being compared to the initial offering price, EMMA trade prices must be recalculated for changes in rates. Bond prices move inversely with rates, so a rise in rates will make the underwriter look better as investors pay a lower price in secondary trading (bond is worth less). Conversely, a drop in rates will make the underwriter look worse as investors pay a higher price (bond is worth more). We recalculated trade prices based on a yield equal to the original credit spread plus MMD as of the trade date, using yield to worse.
- Trade prices include transaction costs. This can include dealer commissions and profits investors make when selling smaller blocks, which don't have much to do with underwriter performance and more to do with how the primary offering is distributed among investors. In late 2016, the SEC approved new rules requiring dealers to disclose markups on retail trades. The rules go into effect in May 2018 but won't be much help as disclosure won't be required if a bond is sold out of inventory acquired by a dealer prior to the transaction date. We did not adjust for transaction costs.
- Borrowers can change. Trade prices can be impacted by changes in the borrower's credit quality or shifts in credit spreads since initial pricing. We did not adjust for these changes since they usually take several weeks to impact prices.
"Before markups can be calculated, EMMA trades must be adjusted for changes in rates since day of pricing."
Our first step was to calculate the markup percentage for each issue as the sum of all markup dollars divided by traded par.
We then ranked issues based on how they deviated from the median:
- Top Perfomers. 184 issues (half the total number of issues) traded at a markup of less than the median of 0.97%, a reasonable markup considering dealer commissions and other transaction costs.
- Average Perfomers. 116 issues (one-third of total) traded at a markup of the median plus one standard deviation: between 0.97% and 1.81%. Not great, but still acceptable.
- Bottom Performers. 68 issues (about one-fifth of total) traded at a markup greater than the median plus one standard deviation: 1.81%. Money left on the table.
Above: Distribution of secondary trading markups as percentage of traded par.
The 68 issues in the bottom category saw a median markup of 2.3% of traded par, or 1.9 percentage point more than the top tier median of 0.4%. On a $50 million bond issue, this additional markup is equivalent to adding $1 million per year to the hospital's cost of funds.
"The median markup during the first 30 days of trading was 0.97% of traded par."
Here are the bottom performers in secondary trading:
Above: Hospital bonds issued since 2015 with a secondary trading markup greater than 1.81% of traded par.
Markups did not vary much based on par amount sold.
So how can hospitals use this information to minimize their cost of funds, particularly since trading patterns do not emerge until after pricing?
Armed with secondary trading analysis, there several steps hospital management can take to ensure optimal pricing:
- Look back. During underwriter selection, the analysis can be used to review the underwriter's previous offerings to identify less than stellar track records; past performance may not be a predictor of future returns, but a poor pricing history can be part of the scorecard.
- Be clear. Before pricing, management should make it clear to the underwriter that secondary trading will be scrutinized and results will be used in future underwriter selection.
- Stay actively involved. Management should be involved in all pre-pricing discussions as well as pricing calls, to demonstrate how focused the hospital is on pricing as part of the financing process.
There is no question that underwriters who price bond offerings day in and day out have an advantage over hospitals that do not access the markets on a frequent basis.
To level the playing field, hospitals can bring in a financial advisor or other third-party with the knowledge and systems already in place to analyze trading activity.
"Make it clear that secondary trading will be analyzed and used for future underwriting RFP's."
With an understanding of its inner workings and limitations, hospitals can apply secondary trade analysis to measure underwriter performances and avoid leaving too much money on the table on day of pricing.
Note: This article expands topics discussed in Hospital Bond Trading Markups (July 2016).