The COVID-19 pandemic has caused devastating financial stress for many companies, while leaving others relatively unharmed and some even better off than before. A Refinitiv StarMine white paper uses hypothesis-driven modeling to predict which companies will best weather a severe financial crisis, and to help investors identify the most resilient companies today.
- COVID-19 has caused extreme financial stress for many companies. This has taken the form of companies issuing profit warnings, missing earnings or declaring bankruptcy, and the shocks have been felt in every sector.
- The Refinitiv StarMine team has produced a white paper that hypothesizes which financial characteristics lead to resilience in companies, and uses those features to predict how companies will fare during a severe economic downturn.
- The analysis formulated specific hypotheses in such areas as return on assets (ROA) and the likelihood of declaring bankruptcy, and back-tested the hypotheses against the Dot-Com Bust and the Global Financial Crisis.
For more data-driven insights in your Inbox, subscribe to the Refinitiv Perspectives weekly newsletter.
In addition to the human tragedy it has unleashed, the COVID-19 pandemic has caused unprecedented financial turmoil and economic instability.
Many more companies than usual have shown signs of extreme stress, from issuing profit warnings, to missing earnings, to declaring bankruptcy.
The effect varies by industry and by company, but shocks have been felt in every sector. For example, according to Ernst & Young, more financial services profit warnings have been issued already in October 2020 than were issued in all of 2019.
Which companies are able to resist the shock of COVID-19?
Predicting which companies will be resilient to such severe shocks is the focus of a new research paper from the Refinitiv StarMine team, Identifying Resilient Companies in a Financial Crisis.
Inspired by the impact of the COVID-19 pandemic, their work aspires to more generally predict resilient companies in any period of severe economic impact. Their goal is to help investors through the current, and future, economic downturns.
They first addressed the question of what it is to be ‘resilient’.
In general terms, they see it as the ability to survive and thrive during and immediately after a significant market downturn.
They explored several quantitative proxies for resilience and settled on a combination of bankruptcy avoidance and relative future profitability as measured by change in ROA at multiple time horizons. The ability of a company to avoid defaulting on its debts, and to continue to generate profits, is key to surviving an economic shock.
How did they identify such companies?
A primary aim of their research was to answer that very question for investors by exploring how StarMine quantitative models and Refinitiv financial data can facilitate this search.
How can StarMine models predict company resilience?
The StarMine suite of equity quantitative models has a track record of excelling at picking winners and losers by ranking companies on a 1-100 percentile scale across quantitative factors.
Similarly, the StarMine suite of Credit Risk models has an excellent track record of identifying companies likely to default in the next year.
However, these models were designed to apply across all economic conditions. They were not optimized solely for turbulent markets. So, a key question is: Can a combined formulation of StarMine models that were designed for generally more stable economic conditions also succeed in extreme economic conditions?
They formulated specific hypotheses regarding predicting the rank of change in return on assets (ROA) and the likelihood of declaring bankruptcy, and ran event studies and back-tested these hypotheses during the two most similar historical periods to today’s conditions: The Dot-Com Bust, and the Global Financial Crisis.
Though the testing period was limited, they were able to use it to rule out some hypotheses as well as to prioritize promising ones.
Table 1. Some hypotheses about resilience
Based on their findings, the team constructed a Resilience Model score that combined several StarMine equity model scores along with the StarMine Combined Credit Risk model score as inputs.
Table 2. Selected event study results showing Spearman rank correlations between the StarMine model score and the change in ROA on a universe of U.S.-traded securities at 90, 180, 365 and 730-day time windows after the market dropped by 20 percent
The results show great promise in predicting change in ROA through a financial crisis. The Resilience Model score also succeeds at helping avoid companies that might go bankrupt.
Due to the limitation of testing on only two historical events, they cannot guarantee these results will generalize, but they see strong evidence suggesting the Resilience Model is worth further investigation. See the research paper for further details of results and methodology.
Watch: Early indicators of Thomas Cook’s bankruptcy — StarMine financial modeling: A Data Moment