Case Studies
RavenPack | February 03, 2020
In a recent study of the U.S. muni-bond market, Professor Stephanie F. Cheng at Tulane University takes an interesting and novel approach to applying RavenPack news sentiment and event data.
The problem with the U.S. municipal bond market is its “opacity,'' she says, but one way around this is to use corporate earnings aggregated at a state level, which can be gauged via earnings news sentiment using the RavenPack analytics platform.
Corporate earnings news shed light on the municipal bond market by providing an estimate for the economic wellbeing of the municipality in question, and, based on that, aid in the valuation of its bonds.
We decided to ask Professor Cheng some more questions about her research and her views of alternative data in an interview, and her answers are set out below:
Municipal bonds continue to play a vital role in financing local public services throughout the U.S. In fact, recent years have seen substantial growth in local government debt issuance—from $182.9 billion in 1996 to $445.8 billion in 2016. As of 2017, the secondary market comprises $3.8 trillion in outstanding bonds.
Recently, state bonds have attracted much attention from the public and media. For instance, Illinois was downgraded to a near-junk credit rating by Standard & Poor’s and Moody’s in June of 2017.
However, investors in this market face high information opacity relative to the corporate sector, primarily because the Tower Amendment limits the Securities and Exchange Commission and the Municipal Securities Rulemaking Board’s authority over the issuer. For example, between 2014 and 2016, the SEC found that 96 percent of the municipal underwriting market had failed to ensure that the issuers provide continuing disclosure in the secondary market.
This information opacity motivates me to investigate whether alternative information signals can inform investors in such a market.
I find that monthly earnings signals that are aggregated at the state level predict states’ future economic development and are positively associated with contemporaneous state-bond returns.
Further, the effect is stronger when corporate managers disseminate earnings news more extensively and when news coverage on public firms’ earnings signals is greater.
These findings suggest that public firms’ earnings announcements provide bondholders with real-time signals about regional economic performance, and the business media plays an important role in this information dissemination process.
My study requires data aggregation for a comprehensive set of public firms. The key is to obtain a relatively complete set of public firms. Thus, I need to rely on various datasets to construct my aggregate measures. For news coverage, I rely on RavenPack.
RavenPack has comprehensive coverage on public firms’ earnings and non-financial signals. I also use RavenPack’s sentiment to proxy for the content of the broader news coverage on public firms.
Yes. I use the data for two purposes. First, I use the data to construct a cross-sectional variable that proxies for the media coverage on public firms’ earnings information. Second, I use sentiment scores to proxy for the information contained in the actual new articles.
Yes. Investors are always actively searching for forward-looking information to facilitate their investment decisions. In the current era of algorithm trading and robotic journalism, exploring new alternative information could provide investors with competitive information advantage over others.
Media has a great influence over capital markets. News reporters have incentives to uncover and report negative events. Their incentives are actually quite assigned with bondholders’ non-linear payoff function. Thus, news articles naturally would contain information relevant to the bond market. Thus, I think this is a promising area.
I am studying the disciplinary effect of media on local government reporting quality.
You can find her profile on Tulane University website and her latest work on her Google Scholar link .
Further research showing how to use Sentiment and Event Data for a variety of Fixed Income instruments is available here .
Please use your business email. If you don't have one, please email us at info@ravenpack.com.
We will process your personal data with the purpose of managing your personal account on RavenPack and offering our services. You can exercise your rights of access, rectification, erasure, restriction of processing, data portability and objection by emailing us at privacy@ravenpack.com. For more information, you can check out our Privacy Policy.
Your request has been recorded and a team member will be in touch soon.
High inflation has returned in developed markets after decades of lying low. In our latest paper, we show how to build an inflation-based asset allocation strategy using sentiment data and we illustrate that sentiment-based strategies outperform models that depend merely on past observed inflation values.
This year's RavenPack Research Symposium brought two intense days of knowledge sharing in London and New York, from 25 top experts in natural language processing, quantitative investing and machine learning. Together, we explored how firms can leverage new language models to generate alpha, better manage risk and respond to calls for more sustainable investment practices.
Human capital is at the heart of value creation. Our latest research demonstrates how unprecedented workforce insights, sourced from over 200 million job postings, can generate more alpha.