Municipal bond market innovation continues to lag most other major financial markets – the sector is arguably light years behind the equity market in transparency, timely reporting, electronic trading and analytics. The sector’s history of low default rates against corporate bonds is likely one of the reasons why investors are somewhat blasé about lagged financial reporting, thin disclosures, and the reluctance to include alternative data in the investment process.

It is safe to say that corporations with publically traded debt and/or equity provide more transparency to their investors than municipal bond issuers, albeit not necessarily by choice. Indeed, there are several reliable vendors that aggregate historical corporate financial data, versus relatively few for municipal bonds.

The municipal bond market should consider leveraging some of the same innovation that equity markets have already adapted, including technology that rapidly identifies sound investments, analyzes credit and monitors positions in a cost effective manner. As the hedge fund sector demonstrated in the past, some investors will turn a blind eye to higher management fees, if a money manager produces above average returns over peers or pertinent index benchmarks.

Many of those outperformers in the equity market have successfully deployed a “quantamental” approach – which takes the sector expertise of an analyst and improves investment decisions through a combination of machine learning and alternative data which identifies “diamonds in the rough” and avoids “landmines”. That same approach can be applied to high yield and unrated municipal bonds to potentially enhance a portfolio’s performance and accurately price risk.

In the case of more plain vanilla strategies and certain SMAs, investors will bargain shop based on fees, which has been driving down fees and profitability across the wealth management industry. This drive to more of a low cost asset accumulation model will require AI based tools and not the hiring of more analysts to rapidly analyze new issue and secondary credits, create accurate and comprehensive marketing material for pitching bonds to their clients, and automated surveillance tools to identify local or regional economic/financial distress using financial statement and public/alternative data sources.

The holy grail of municipal bond analytics will likely mimic that of an industrial supply chain, where every source of revenue and expenses will be identified or estimated through a non-traditional data proxy. These metrics can then be compared to changes in liabilities and the tax paying population (citizens and corporations).

An investor would begin with an aggregate view of every potential bond offered by the dealer community – coupled with MSRB trade price history and government bond yields – and supplemented with accurate evaluated bond prices/yields to fill in the days where a round lot did not trade.

The next layer will use natural language processing (NLP)-driven news-to-CUSIP mapping applications, and alternative datasets – such as US port ship traffic and US Customs bill of lading data – to proxy revenue through the flow of goods in and out of a state, while mining through publically available bespoke data from data.gov to enhance standard economic data releases.

The biggest leap will be made when the performance of the largest private employers for the issuer is added to the credit picture, enabling the identification of a growing or shrinking tax base. Lastly, all of the aforementioned elements will be combined with financial statement data to model which factors drive the issuer’s assets and liabilities the most – with the end goal of determining its performance outlook.

The successful implementation of AI and alternative data in the investment process will benefit asset managers and issuers by modernizing investment and due diligence processes. The investment community has the resources and expertise to discover an issuer’s tax revenue base shift through advanced data, with those same findings having the potential to help guide municipalities’ financial and policy decisions.

Machine learning has been used by credit card companies for fraud detection for decades, and can potentially be used to identify discrepancies and errors in financial statements, when compared with data sourced outside of the issuer.

Deploying these types of technologies may eventually be a matter of pure survival for money managers, because clients will likely gravitate towards money managers that successfully combine alternative datasets, AI, and sector expertise to identify real-time shifts in credit.

Those who are patient enough to wait until the issuer’s next quarterly or annual report is released will not fare as well.

S&P Global provides industry-leading data, software and technology platforms and managed services to tackle some of the most difficult challenges in financial markets. We help our customers better understand complicated markets, reduce risk, operate more efficiently and comply with financial regulation.

This article was published by S&P Global Market Intelligence and not by S&P Global Ratings, which is a separately managed division of S&P Global.