Leading the discussion are the PRI’s Carmen Nuzzo, Head of Fixed Income, and Katherine Ng, Head of Research. They’re joined by Eric Friedland, Director of Municipal Research, Lord Abbett, and John Hund, Assistant Professor of Finance, University of Georgia.
Interested in the resources discussed in this episode?
Find the episode transcript here
The Principles for Responsible Investment podcast transcript
THE US OPIOID CRISIS AND MUNICIPAL FUNDING WITH CARMEN NUZZO, HEAD OF FIXED INCOME, PRI, KATHERINE NG, HEAD OF RESEARCH, PRI, ERIC FRIEDLAND, DIRECTOR OF MUNICIPAL RESEARCH, LORD ABBETT, AND JOHN HUND, ASSISTANT PROFESSOR OF FINANCE, UNIVERSITY OF GEORGIA
Note: The Principles for Responsible Investment podcast is designed to be heard. If you are unable to do this, this transcript offers an insight into the episode.
Transcripts are generated using a combination of speech recognition software and human transcribers, and may contain the occasional error. Please check the corresponding audio before quoting in print.
Subscribe to the channel via Apple podcasts, Spotify, or wherever you listen to your podcasts.
Hello and welcome to this new PRI podcast. My name is Carmen Nuzzo and I’m the head of fixed income at the PRI. Opioid abuse has taken the lives of over half a million people in the US over the last twenty years and while the crisis is a national one, an aspect that doesn’t grab the headline as much is how it plays out at the local level and how the shape of the epidemic changes drastically from one community to another. These local differences have repercussions, not just on lives lost and healthcare costs but also on US municipal bonds and the cost of local authorities’ funding. 2 I’m co-hosting today’s podcast with my colleague Katherine Ng who’s head of research at the PRI.
Our joining of forces is testament to how this topic is piquing the interest not just of investors but also of the academic world. Our guests today are John Hunt, assistant professor of finance university of Georgia. John co-authored a paper precisely on the effects of the opioid crisis on municipal finance, and Eric Friedland, director of municipal research at Lord Abbott, who will bring the investor perspective. Right, I’ll jump in with the first question, John can you tell us about this crisis and its impact in the us?
Thank you very much Catherine and Carmen and Eric for this opportunity to talk about the opioid crisis in in America. Over the last 20 years opioid poisoning and deaths from it have risen to become the number one cause of injury death in the United States, rising about 8,000 deaths per year in 2000 to over eighty thousand in the most recent statistics and we focus on deaths in our study as a measurable indicator of the of the toll of the crisis. But yeah, obviously the toll is much higher. In terms of destruction of families, social fabric and economic costs. We focus on the economic costs and in particular a slice of those economic costs. The capital market consequences for local communities overall estimates of the economic costs to communities run somewhere close to over $1 trn a year, that’s from a couple studies in the society of actuaries and that factors in a host of things including direct costs to communities and diminished labour productivity and other sorts of sort of economic costs. So, it’s a real thing it’s a very big public health epidemic and it has immense economic costs.
John, how have the effects played out over the last twenty years?
So Katherine, there’s been really about three phases of the opioid epidemic in the United States. The first was heavily influenced by marketing by Purdue Pharma and it was a prescription opioid crisis that was primarily centred in rural and suburban areas. After about twenty ten there were crackdowns in the United States on prescription opioids and that drove addicts towards illicit opioids primarily heroin for a few years until the current phase which is very heavily driven by synthetic 3 fentanyl, which is much stronger and easily sort of overdosed opioid. It has caused deaths to spike in recent years and really moved the crisis from many community from the rural suburban sort of origin communities to sort of all counties and cities being affected in this way. So, what we’ve done in our in our paper is we’ve looked at these costs and in particular we’ve looked at the costs associated with the crisis that are born by local and non-state municipalities the cities towns and counties across the United States and what we really do is we try to quantify the effect of the crisis on local budgets by matching each year areas with high opioid abuse with areas that are somewhat identical in degree of urbanization and very similar with regard to income property values, population similar trends and GDP growth, employment growth unemployment rates, those sorts of things and what we show is at least at the very local granular level that there are significant differences between the high abuse and very economically similar low abuse counties particularly in, in terms of lower received properties and sales taxes and at higher expenditures on things like law enforcement and corrections prison prison expenditures this lowered revenue and and higher expenditures creates a lot of budgetary pressures in places with higher degrees of opioid abuse relative to areas where there isn’t abuse that have similar economic fundamentals.
Wow. When you examined the effect of the crisis on muni bonds, what did you find?
Thanks Katherine. First of all, I’d like to thank my co-authors Kim Cornnegie and Zangan of Penn State and Zhan Ye, they were obviously huge contributors and a big reason that we got this sort of very large research project off the ground. To do this justice, we realised we had to get the best possible data. So what we did is we went out and collected effectively super granular data that covers the universe of Geo Bond issuance, every death in the United States at the death certificate level. And every opioid pill shipped over the period via a database as well as a host of other socioeconomic and economic fundamentals. Um, we put this data together and show that in communities with high opioid abuse, there are municipalities that have higher yields on their primary issuance and that they have slightly higher ratings. But more importantly, they often have much lower issuance amounts relative to economically similar areas that don’t have as high of an opioid abuse um ah high as high levels. Of opioid abuse, one of the things that we’re worried about is that we are going to essentially just pick up correlation and not causation and so what we try to do is exploit the supply chain of very addictive opioids and effectively finding that. Places that had very high levels of easily divertible supply of opioids as well as sort of were subjected to very strong initial 4 marketing of Oxycontin by Purdue Pharma in the pre sample years from 96 to 2002 so those two things heavily predict deaths in much later periods. Ten, Fifteen years later and abuse as well so what we do is we effectively show that that municipalities with really similar economic conditions but that had very different supplies of addictive opioids have very different yields and issuance amounts of municipal bonds.
Thank you John and I encourage you to take a look either at the article itself that John co-authored or the blog that is available on the PRI website. It really made me think about the opioid crisis through a different lens especially because of the link to the yields and the issuance by different municipalities and also highlights how different the issuers in the muni market are not just by type but also depending on the state they are located in and their and geographical and economic exposure. Let me turn to you Eric as a market predicter and petitioner and an expert on the munibond market. What are your reflections on this study?
Yes, thank you Carmen and it’s a pleasure to be here speaking with John and Katherine as well. I thought the study was very interesting and it calls attention to an affliction that has both serious social and financial repercussions as an analyst. My focus is on how this type of issue can impact credit quality. Demographics have a significant effect in counties where there is opioid overuse in many cases. There are higher incidence of population loss crime and emergency medical activity. This can result in lower tax revenue higher police and health care costs and declines in economic development. All of which can manifest itself in structural budget deficits, higher debt burdens and deferred capex, these weaker credit factors will result in lower rating agency ratings which will require the counties to pay higher rates of interest to borrow money. One finding of the study that I had not considered is that counties with opioid overuse are not able to access the capital markets as easily as others so their ability to raise funds to correct these issues is limited. This can create a death spiral that is hard to reverse, some would say that the higher yields on these bonds create an opportunity for investors to earn higher returns. However, the risk of further credit deterioration downgrade in many cases would outweigh the potential high return that could be realised by purchasing the bonds, the study addresses a critical issue and for the most part I believe it is well done. However, one of the things that concerns me is that the data set only includes a small part of the municipal market. Many investors assume the municipal market includes only state and local general obligation bonds which is only what is 5 included in this study. However, that only encompasses about 30% of the muni market, within the state and local sector there are lease in special tax bonds and then they’re over a dozen revenue sectors ranging from water sewer, electric, healthcare, education, etc. I know it would be challenging to include all of these bonds but excluding them may limit the conclusions that can be made also as and it’s important to note that rating agency analysts and the research teams at management firms such as Lord Abbott consider a very broad range of both quantitative and qualitative factors when formulating credit ratings and opinions. The study does focus on several key financial metrics but it does not incorporate many of the others that would be routinely be considered. Having said that one of the biggest challenges for municipal analyst is a lack of disclosure from issuers. It is very difficult to get data regarding opioid overuse, getting data on opioid prescriptions, dispensed opioid treatment levels substance abuse services and narcan programs would be helpful. The final thing I would mention about the study is that understanding cause and effect is difficult, opioid overuse may be one of several factors that results in population loss. Tax-based decline and financial distress may be the case that these conditions can actually lead to an environment that is riper for opioid overuse. Also, there are many other factors other than high opioid use that cause these conditions such as education levels, poverty and employment opportunities. So, although opioid use is an important factor, it’s hard to know how significant of a contributor it is to credit distress.
Eric, you know nobody disputes the fact that investment decisions are taken based on a variety of factors and not just one. But if we keep the attention for the time being on social factors. You’ve mentioned already a couple but what other material factors you would cite as relevant for the for muni bond valuations?
Yeah let me start by saying that for all of our holdings, we incorporate a proprietary ESG analysis and scoring process, when we assess social issues we are looking to see how labour and income access to education and housing demographic change in health and safety are likely to affect government credit worthiness. For example, we may assess the fiscal and social implications of poverty on economic competitiveness growth and ultimately per capita income levels and is not just the metrics we are looking at impact and how proactive management is addressing any negative social factors. 6 Also trying to understand how bond finance projects may provide important services to underserved populations for certain sectors, social factors will have a more significant impact on our overall ESG assessment. This is the case for the local government hospital and education sectors where for public power and water sewer, social issues are important. But environmental more. So, for every sector there are unique social factors and metrics that we assess. To provide a few examples, I’ll start with education. We assess quality by looking at academic performance, its measured by school ratings and test scores, student retention, teacher retention and teacher credentials, assess poverty reduction in economic growth by looking at high school graduation rates post-graduation outcomes and a breadth of offerings such as technical education programmes and for reduced inequalities. We look to see that there are programmes such as bilingual education and English proficiency and that low income students are served by free or reduced breakfast and lunch. Also evaluate the response to Covid-19 where school’s ability to shift to remote learning without leaving the most vulnerable students behind has been critical, another sector where social issues are critical is hospitals. Quality is assessed by looking at star ratings and readmission rates, serving the underserved is assessed by looking at the pair mix of patients in particular, percentage of them on medicaid or self-pay and education and training are assessed by the quality of clinical residency programmes and employee engagement. As a final example, there’s a local government sector here. We focus on income inequality by looking at poverty rates, homeless rates and gene rates, the need for social services is assessed by looking at age of population and mortality rate. Public safety is assessed by looking at residents per police and the crime index and the desirability of area is assessed by looking at vacancy rates, a number of hospitals’ social factors can have both a positive and negative impact on credit quality. Some examples of positive are follows. There’s a private university in the southeast that provides strong financial aid, a robust diversity inclusion programme, strong career placement and good retention of faculty, this has resulted in strong enrolment growth and financial performance. There’s a hospital in Oregon with high quality ratings and low cost of care and engagement with local communities to address racism in public health has helped lead to higher operating margins and higher utilisation and then looking at negative impact. There are two examples I can provide. There’s a school in California where poor test scores, enrolment declines and labour conflict resulted in lower state funding and poor finances. Then there’s a charter school in Arizona where conflicts of interest between board members and management resulted in the overpayment of several projects poor accounting and poor student outcomes.
Eric, you’ve mentioned quite a bit of different examples and data there I’ll turn to you first then, then John, as this research discussed the importance of comprehensive local data. What are the data challenges when it comes to assessing social factors in municipal bonds?
Perhaps I can start off, all of the ESG factors, social factors are by far the most difficult to assess, they’re not easily defined. They can be ambiguous, less tangible and data is more difficult to find. There’s a lack of global consensus on how they should be viewed. Factors can be politically charged, amplified by the news media and sensitive to address. The other thing is that the connection to credit is indirect, only becomes apparent over the long term as I mentioned earlier Femuni’s disclosure is a challenge not just for ESG, but in general. There are no uniform disclosure requirements and compliance is not monitored from data. There are no bulk data providers as you would have for corporates, muni ESG data is often disaggregated, incomplete and they’re overlapping data sources. We’re starting to see some ESG-related disclosure and offering statements. But as far and few between, there is one agency that is working to change this. The group is called the Municipal Securities Rulemaking Board, also known as the and MSRB ,for those of you who aren’t familiar with the MSRB, they were authorised 19875 by Congress to regulate the activities of broker dealers and banks that buy, sell and underwrite municipal securities. And although it’s a self-regulatory agency, it’s overseen by congress and the SEC one of the ways the MSRB protects investors and municipal entities in the municipal market is by increasing the transparency and availability of market information and they provide free access to municipal Bond Trade Data offering documents and continuing disclosures through their online platform. The NSRB is now seeking input from investors in municipal securities regarding ESG-related disclosures and ESG-labeled bonds. They’re asking market participants 12 questions such as how important is ESG-related information? Is there adequate access to information? What type of information would be important? How could issuers better provide information and how related information is material to investment decisions. Our firm has responded and hopefully this survey is going to result in the MSRB encouraging issuers to provide better disclosure through their platform.
So I mean I can jump in I guess, disclosure at the local level for ESG as Eric pointed out is very limited. We had you know specifically with regard to the opioid crisis. It’s very, very difficult to get information on deaths at the county level, at a very granular data level in order to do sort of meaningful work and that only occurs with a significant lag so we were able to access. For instance, very detailed granular budgetary data, very detailed revenue data very detailed employment data for local communities but that only comes through the census at a lag which is not useful to investors. So, there’s a huge need again for both disclosure of basic municipal finance data at the budgetary level which still as Eric pointed out is still not the case and above and beyond that disclosure of data at on ESG factors, which is now you know, at best scattered across a multitude of different sources. We really do feel like it is a limitation. That we don’t look at the revenue bonds that he speaks of so I’d have very different sort of cash flow expenditure sort of sources. You know it’s difficult to compare so we wanted to focus on the most homogeneous source of bonds those that had sort of recourse to local tax revenue as a benchmark to most cleanly identify our effects. The other thing that that we do spend a lot of time working on is trying to differentiate the key the key struggle that Eric talked about which is disentangling sort of correlation and economic general economic decline from the pure effect of opioids. That’s why I spoke about the fact that we compare areas with high supply of opioids to places that have high low supply of opioids that have. Almost identical economic fundamentals. In particular we um control for lots and lots and lots of economic fundamentals. But more importantly, what we’re really doing is essentially finding the channel that is defined by the shift in supply of addictive opioids. And that shift in supply sort of allows us to identify causality rather than just correlation and we also sort of do the sort of obvious things where we control for credit. You know in our yields sort of results. We’re sort of thinking about the yield in a in a county that’s similar economically. Similarly rated of a bond that has similar duration, convexity issue size, use of proceeds, and state tax regimes. So, what we’re really doing is trying to compare like for like, is just in areas where there were lots and lots and lots of opioids. Either drop by Purdue Pharma marketing or by through an easily divertible supply chain. So in some ways some ways, we feel like we’ve tried sort of from at a first order to disentangle this correlation and causation. By using shifts in the supply of opioids to really sort of push back and think not just correlation but a causal sort of fundamental risk factor channel.
Thank you Eric and John for a very insightful discussion. I’m really glad that you found the interaction useful and this is definitely something that we want to encourage bridging links with between the academic work and the practitioners. Unfortunately, it brings us to the end of our conversation. Before I leave, I would like to remind you listening that the PRI launched a dedicated work stream on ESG consideration in sub-sovereign debt in 2020, you can find our first report focusing specifically on us municipal bond, as well as other resources that we’ve produced so far on http://wwww.umpi.org/subsovereigndebt. Please also take a look at unpri.org/fixedincome for other resources related to fixed income more broadly and workstreams dedicated to different debt instruments and of course get in touched if you want to be more involved in future work.
And if you’re interested in more from the research programme where we seek to bridge the gap between investors and academics and also investigate thought leadership areas you can join us at the academic network conference within Pri in Person on the 20-21 September in Barcelona and of course online. You can access and engage with experts, research insights and data via our seminar series, blogs podcasts, and resource library at http://unpri.org/academicresearch. You can find links to all of these resources in the episode description. Thanks for listening and tune in for our next podcast episode on corporate purpose with PRI senior analyst of corporate governance, Betina Vaz Boni.