The Role of Economic Scenario Generators in the Age of Covid-19
Covid-19 has tipped our previously held socio-economic orthodoxies on their head but the value of insurance, of all kinds, has been highlighted in sharper relief. Insurance has always been a tough gig, but now more than ever we should be thankful for the experience and talent manifest in the sector. Most of us who work in the financial services industry will know or work alongside someone with FIA after their name. Known, admired and sometimes pilloried for their grit and stoicism, actuaries are valued colleagues. They tackle the mystical arts of mortality, longevity and morbidity, which to the layperson sounds like they come straight from Harry Potter.
Magic is not involved, however. Actuarial science can trace its earliest roots to the age of antiquity. Nonetheless, it was really James Dodson’s ground-breaking development of the concept of long-term insurance in the 18th century that gave birth to the field. Pioneering actuaries performed feats of mathematical brilliance with little more than paper and quill. Mercifully, a raft of technological advances have created an empowered environment for the modern insurer to work, and one in which a myriad of products and services can be offered.
The insurance toolkit today is very significant indeed, solutions are multi-faceted and enterprise enabled. Moreover, the extent to which an entire software industry is now developing adjacent to core Insurance has brought about a new Portmanteau -InsureTech. There are many firms entering this space, all fundamentally aiming to disrupt traditional insurance paradigms and ultimately reduce premiums. Within the pantheon of more traditional insurance solutions we find technology which has been designed to tackle the fundamental challenge of asset-liability management (ALM). Insurance firms create extremely complex liabilities, which tend to be long-dated and inherently contingent in nature. They must ensure, therefore, that the assets they hold on their balance sheet match off against these liabilities for regulatory capital purposes, but also in order to survive commercially.
Economic Scenario Generators (ESGs) are fundamental to the analysis of ALM problems. Oversimplifying, they are software tools that facilitate simulated analysis of economic variables and risk factors. Whereas in fact,they are highly sophisticated models of economies which not only predict the future paths of single factors but capture the inherent interdependence of said factors. This has always been a vital capability for ALM portfolio management, not only on the asset side of the balance sheet but also to enable a proper dissemination of future liabilities. Looking at the world through new eyes as we now must, it’s difficultto overstate how mission-critical such frameworks will become. Nonetheless, for existing paradigms such as LDI, which require a forensic coupling of future assets and liabilities, an ESG in combination with 1’st class analytics is mandatory to prevent mismatches and non-optimal benchmark asset allocations.
6 months ago, no one in the West could have predicted what we are now experiencing. In South-East Asian countries, with their experience of SARS and bird flu, pandemic risk would have not been considered quite such an outlier. Nonetheless we are truly now in un-navigated economic territory globally. Stress-testing and scenario analysis comes in a variety of formats and styles. Many are formulated by benchmarking variability on previous events and crises. None of these would have offered any forewarning of the impending magnitude of Covid-19. Specific predictions vary and are challenging to make, but we can be confident in seeing a record single quarter decline in global GDP. ESGs are not crystal balls and would not, ceteris paribus, have provided any direct mitigation to these challenges. However, as we prepare to make our first tentative steps into the ‘new normal’ we must surely re-evaluate the role that enhanced analytics can provide for asset allocators.
Generically, ESGs typically incorporate three standard inputs: The current market prices of instruments and both historical prices and qualitative forward forecasts of those instruments. Having identified the commonalities, we should also recognise that ESGs tend to come in 2 distinct flavours. Risk-neutral ESGs come equipped with the requisite parameterisation to incorporate derivatives and long dated insurance liabilities which feature embedded optionality. Derivative pricing theory is not supportive of ‘subjective’ forward looking asset projections, so as such risk-neutral ESGs tend to focus on the here and now. Real-world ESGs, on the other hand are very much focussed on creating a comprehensive sandbox for stress testing balance sheets against potential future economic states. Real-world ESGs are therefore vital for regulatory capital purposes, for example for use in standard formula solvency capital ratio (SCR) requirements under Solvency II. The market is well served with incumbent providers who offer powerful solutions which deliver comprehensive out-of-the-box coverage and functionality. However, we must now question whether these solutions go far enough. Business critical use-cases for ESGs tend to focus on their application in support of regulatory capital calculations alone. We now live in aworld in which extreme tail events define the epoch. Asset allocation for life and pension portfolios must adapt to become flexible and reactive with respect to emerging investment strategies as we try to exit the crisis. It will also be vital that the modern ESG frame work can flex accordingly to support risk functions and wider business needs too. An ESG should not simply be a data dump of historical prices but offer enough customisation of that data to enable the client to take control. Features such as flexible model hierarchies, hybrid term structures and dynamic benchmark calibrations should be standard deliverables. In doing so ESG frameworks can re-vitalise their purpose in firms and permeate emerging mission critical workflows for asset allocators.
ESGs will become increasingly vital resources for firms. They will help define new mission critical operational processes within the organisations which rely upon them. Moreover, the nascent trend for ESG model validation will no doubt gather pace. Historically, insurers have not really sought to challenge the premise that a single ESG framework is appropriate. Now, there is clear recognition that legacy operating models which rely on single source ESGs impose very significant modelling reliance, and possible bias, into both upstream and downstream processes. Validation fundamentally ensures that your assumptions are consistent with the historical data, prices and in-house views to which is intended to calibrate, along with sensitivity tests to ensure reliable behaviours. Furthermore, validation ensures your model is able to replicate relevant stylised facts such as heavy tails, tail dependence and volatility clustering. Modern ESG validation solutions are parsimonious by design and can be delivered as light-touch & hosted or as a fully managed service.Super-charging your ESG solution with a comprehensive, auditable and customisable validation overlay is now not an ‘if’ but a ‘how and when’.
The world has changed and actuaries and others involved in complex asset and liability management, will face unique and complex challenges in the coming months and years. However, they are no longer just equipped with pen and paper. Solutions abound in all the fundamental elements of insurance provision. ESGs have been and will surely become an increasingly critical part of that solution architecture.