Updated: Jan 15, 2021
Last week, explained why the pandemic creates a “when bad data happens to good models” situation. To catch up, see: Two Data Deficits and Why Data Deficits Matter. We also de-bunked the misconception that scenario analysis when applied to public policy risks is not the equivalent of playing darts.
Today, we de-bunk the misconception about event risk…..it is NOT a random variable, particularly when applied to public policy risks.
Event Risk 101
In the capital markets, “event risk” is defined as follows:
The risk that the ability of an issuer to make interest and principal payments will change because of rare, discontinuous, and very large, unanticipated changes in the market environment such as (1) a natural or industrial accident or some regulatory change or (2) a takeover, or corporate restructuring.
We used NASDAQ’s definition, but the core component lumps together three kinds of events which have a material impact on a company’s ability to meet its obligations. Those three events are: natural disasters, industrial accidents, and regulatory change.
The definition thus places policy risk as a random exogenous development beyond the company’s control on an equal footing with force majeure events.
The International Risk Management Institute goes farther, making clear that event risk consists of the “risk of loss associated with fortuitous occurrences (e.g., fires, hurricanes, tortuous conduct). Event risk, which is synonymous with pure risk, hazard risk, or insurance risk, presents no chance of gain, only of loss.”
But with all due respect to our esteemed colleagues in risk management, we are going to challenge the assumption that public policy risks (including regulatory policy) are fortuitous, rare, discontinuous or random. And we have the data to prove the point.
Why Public Policy Risk is NOT Random
The reality is that public policy rarely changes over night. Most public policy choices change over a period of time. Even in crisis, temporary solutions are made permanent often with a lag time of years. It is a truism among public policy professionals that anyone surprised by an outcome has not been paying attention to the process.
In keeping with the meteorological metaphors we have been using for this blog series, public policy risks are less like volcanic eruptions (which occur with limited, if any, advance notice) and more like storm systems that can be tracked and whose path can be anticipated.
Moreover, it is not exactly true that natural disasters are purely unforeseen, random occurrences. Exposure to many if not most weather-related and other risks (like fire or accidents) are managed by finding ways to measure potential exposure and implement mitigation techniques (like safety procedures, fire drills, purchasing insurance, hiring external experts to deliver evaluations and recommend policy shifts).
Just like firms hire risk management experts to minimize their exposure to natural disasters, financial firms hire public policy experts, but at large firms those experts tend to focus on advocacy – attempting to influence the outcome of public policy initiatives for the benefit of the firm and/or the industry. They are kept far from the risk management and alpha generation functions in order to avoid accidentally informing investment decisions using materially non-public information. Smaller firms may hire consultants with specialized knowledge of a policy process for their expert opinion, but these individuals also have no operational engagement with alpha generation or risk management.
However, monitoring and measuring exposure to public policy risks until recently has been impossible due to lack of objective, transparent data.
Risk professionals and investment professionals interested in attempting to measure exposure to public policy risks, have had no data to incorporate into modeling systems.
Public policy occurs in the real of language. Policymakers, advocates, and voters express views and take action using words. So in the realm of quantitative finance and modeling systems, public policy risks had to be seen as a random variable from a technical perspective because no data sets existed to incorporate into risk measurement systems. The few sets of structured data that do exist in this arena are famously fickle/unreliable (public polling data) or profoundly backward looking (voting records).
Our patented process makes it possible to convert the words of the public policy process (unstructured data) into integers (structured data) before decisions are made. The resulting time series data makes it possible to identify correlations, covariances, and other relationships important for measuring -- and managing – exposure to what many will consider a new risk class: public policy risk.
With nearly 18 months of data already for top-tier global macro risks (e.g. , trade policy; Brexit; macroprudential policy), we are now starting to identify patterns at monthly and other intervals. Those patters will be particularly helpful for bond and currency investors to identify more precisely their exposure to potential shifts in public policy trajectories. Definable patterns are also visible weekly for COVID-19 and other pandemic-era public policy initiatives, although these data sets understandably only date to February 2020.
The good news is that our objective, transparent data makes it easier for analysts and strategists to make better decisions using concrete data. Early adopters receive significant information advantages relative to their competitors who still mistakenly treat public policy risks as random variables. The bad news for risk managers is that their risk models will require updating and expansion to incorporate more rigorous estimates of public policy risks.
PolicyScope data is available through the Bloomberg Enterprise Access Point.
Customized widgets and dashboards are available via API HERE.
In June 2020, BCMstrategy, Inc. held a webinar with Interactive Brokers that explored the challenges risk managers and strategists face when confronted with low data environments. The current COVID19 period presents unique challenges for scenario analysis at the tail of the distribution, where investment risks are uniquely exposed to public policy risks in addition to traditional financial risks. This blog series is based on that webinar series.