Forecasting the future

By Steven W. Easson, FCIA, Chair of the SOA/CIA Project on Real-Time Delphi Study on Economic Variables

Futures research forecasting methods

In forecasting the future, for purposes such as valuations and capital projections, our typical actuarial toolkit includes items such as experience studies, historical market data, complex stochastic modelling, credibility theory, and I go could go on. Very useful tools in relatively stable times. However, along come the “tail events” – the financial crisis and more recently COVID-19. In retrospect, lots of precision in valuation and capital projection assumptions were set before the start of this pandemic. But now, we see lots of work and extraordinary judgement and challenges in refining those assumptions.

In writing this article, I would like to whet your appetite to consider non-traditional tools if you have not already. These are tools to facilitate derivation of plausible future “tail event” scenarios, to facilitate business contingency plans and associated potential adjustments to scenario and assumption setting. These dynamic tools facilitate discussion among (ideally multi-disciplinary) experts to produce plausible futures and deterministic scenarios of what events could happen, their likelihood and the impact: a “fan of plausible futures,” e.g., impacts of the current COVID-19 and identification of future COVID-related events and their impacts. An exact science? Certainly not. A useful forecasting tool? Definitely.

As I mentioned during my April 3 Town Hall remarks, the SOA/CIA will shortly release a report on the use of futures research methods applied to forecasting future values of economic variables. The study consisted of two rounds, both using the futures research tool “Real-Time Delphi” (RTD). RTD#1 gathered opinions from the study’s participants of future values of economic variables, rationales for such opinions, and, relatedly, suggestions for plausible future-shaping developments (economic, political, technological, or social) for use in an impact analysis in RTD#2. Based on 28 developments selected from the feedback in RTD#1, study participants judged the likelihoods and impacts of these events in RTD#2. Noteworthy is the list of these events that were set in 2019, before on onset of COVID-19. Examples include: (i) a catastrophic pandemic; (ii) oil prices dipping below $30; and (iii) U.S. federal debt/GDP reaching 150%. Using the additional futures research tool “Trend Impact Analysis,” along with curve fitting and Monte Carlo simulations, the impact on the extrapolative forecasts derived from RTD#1 were adjusted for each of the developments using likelihood and impact estimates of the developments provided by study participants in RTD#2.

Was the primary purpose of this study to provide actuaries with a source of best estimate assumptions of future values of economic variables? Absolutely not. Was the primary purpose of the study to create an awareness for actuaries of other forecasting methods to consider adding to their toolkit? Absolutely yes. This study was sponsored by the CIA’s Research Council and multiple sources within the SOA including the SOA’s Predictive Analytics and Futurism Section. It was a repeat of the similar study completed in 2005 under the then SOA Futurism Section but it used more advanced versions of futures research tools. Access to these futures research tools will be made available concurrent with the release of the study’s report.

Articles of interest

I would also like to point your attention to three recent articles you may find interesting, under the same banner of “forecasting,” to provide interesting perspectives into potential drivers influencing forecasting assumptions.

Oliver Wyman – Navigating the pandemic’s next wave

This article proposes the need for insurers to have “COVID-19 crisis war rooms” to manage real-time decision-making. It provides insights into: (i) dynamic, multi-country projections of epidemic development; (ii) quantification of evolving scenario impacts on key business and financial drivers; and (iii) contingent business plans for, e.g., cost and capacity management, capital and liquidity planning, operations and technology management, and staffing actions.

And the following two articles give interesting perspectives on short- and long-term challenges, second order impacts, and actions companies can take.

McKinsey – Implications of coronavirus for North American life and annuities writers

McKinsey – Coronavirus response: Short- and long-term actions for P&C insurers

This article is part of a series of practice-specific articles under the Actuarial Guidance Council. Read the next article on mortality by Marie-Andrée Boucher and Steve Bocking.

This article reflects the opinion of the author and does not represent an official statement of the CIA.

This article originally appeared on the CIA COVID-19 Hub.

Follow us

Contact Us

Canadian Institute of Actuaries
360 Albert Street, Suite 1740
Ottawa, Ontario K1R 7X7

Subscribe to our emails