COVID-19: IMPLICATIONS FOR RISK MODELLING - Eight actions Australian banks must take now to meet future obligations in credit risk modeling ...

Page created by Zachary Gomez
 
CONTINUE READING
COVID-19:
IMPLICATIONS
FOR RISK
MODELLING
Eight actions Australian banks must take now to
meet future obligations in credit risk modeling

James Gordon
David Howard-Jones
Wolfram Hedrich
COVID-19: Implications for risk modelling

            Banks are responding to extensive government and regulatory stimulus
            to protect borrowers and maintain the flow of credit. As they do this,
            they must carefully manage their processes and data so that outcomes
            from this period can be used effectively to forecast future losses. These
            data will provide a crucial benchmark for idiosyncratic portfolio shocks —
            and regulators will expect them to be curated.

            INTRODUCTION
            As the COVID-19 crisis keeps Australians away from their usual activities, banks’ first duty is to
            help clients and society function as well as possible. But lenders must also plan for the future,
            and that task has been made more complicated by emergency measures to deal with the
            economic fallout from the pandemic.

            Australia’s government, central bank, and financial regulator have taken unprecedented
            measures to support individuals and businesses, including loan guarantees and other wage
            subsidies, while banks have introduced their own initiatives, such as repayment holidays and
            forbearance. Plenty of firms and individuals will benefit from government action or exercise a
            forbearance period: while they will not repay their loans according to the original schedule, these
            borrowers will not be in default. So, banks will have to strip out the effects of emergency
            measures from the models they use to forecast risk.

            In some cases, such as use of the Australian government’s Coronavirus SME Guarantee Scheme,
            the effects will be relatively simple to account for. But in other cases the effects will be harder
            to isolate. Income tax rebates, for example, will fundamentally alter cashflow data and financial
            ratios, so banks will need to reconstruct operating cashflows with and without them. Banks that
            have substantially invested in both credit risk and transaction data will have an advantage.

            We suggest eight immediate actions banks should be taking to facilitate loss forecasting in years
            to come.

© Oliver Wyman                                                                                                    2
COVID-19: Implications for risk modelling

            BACKGROUND
            The COVID-19 pandemic is having a substantial impact on people’s lives and the economy. Social
            distancing and supply chain disruption are putting enormous strain on businesses, and small- and
            medium-sized enterprises (SMEs) are on the front line. At the time of writing, Australia’s COVID-19
            daily transmission rate is in decline and the reproduction number has fallen below the critical value
            of 11, however we are likely still far from the peak of the economic crisis. Indeed, exhibit 1 provides
            a simple comparison of the latest forecasts for 2020 Q1 and Q2 GDP growth against prior quarter-
            on-quarter contractions and severe regulatory stress parameters, and well illustrates how very
            sharp the shock is expected to be.

            Exhibit 1. Q2 2020 GDP forecasts compared to regulatory stress scenarios
            and historic contractions2

                 -8                     -6                        -4                     -2                      0
                                                                                                                     Q2 2020 Forecast
                                                                                                                     APRA 2020 COVID-19
                                                                                                                     APRA 2017
                                                                                                                     Q4 1982
                                                                                                                     Q1 1991
                                                                                                                     Q1 1983
                                                                                                                     Q1 1982
                                                                                                                     Q3 1982
                                                                                                                     Q3 1990
                                                                                                                     Q1 2020 Forecast
                                                                                                                     Q4 2008
                                                                                                                     Q4 1981
                                                                                                                     Q4 2000
                                                                                                                     Q1 2011

               Forecast (consensus)          Forecast (bearish)        APRA Stress Scenario (consensus)
               APRA Stress Scenario (bearish)          Historic contraction (1959-2019)

            1 Australian Government, Department of Health
            2 Oxford Economics, APRA, Westpac, NAB, CBA, and Oliver Wyman analysis. Forecast based on range of estimates reviewed
              during April 2020. APRA 2017 stress was incurred over 8 quarters so was not envisaged as a single quarter shock. APRA 2020
              scenario understood not to be finalized at time of writing, but may be reasonably expected to exceed severity of prior shocks.

© Oliver Wyman                                                                                                                                 3
COVID-19: Implications for risk modelling

            Australian banks, with support from the government and regulators, have introduced
            substantial, wide-reaching protection for consumers and SMEs. These are hugely important
            for borrowers under pressure and the economy as a whole, but banks must carefully balance a
            number of priorities:

            1. Doing what is right for clients, society, and the economy in the short term.

            2. Protecting the bank’s balance sheet and managing future losses.

            3. D
                efending medium-term shareholder returns by fairly and transparently structuring and
               pricing for increased risk in the current environment.

            4. Looking beyond the crisis and ensuring that decisions made now support the recovery and
               safeguard against future stress.

            It is easy to place too little emphasis on the final item in this list as banks scramble to extend
            repayment terms to clients and administer new government loan schemes, but it is critical to
            get it right.

            SUPPORT IN AUSTRALIA
            The Australian government, the Reserve Bank of Australia (RBA), and the Australian Prudential
            Regulation Authority (APRA), have taken coordinated action to support the flow of credit in the
            Australian economy. These initiatives are far reaching and provide unprecedented levels of
            assistance in several ways:3

            • Government-guaranteed unsecured lending: Under the Coronavirus SME Guarantee Scheme,
                 banks and other eligible lenders will provide unsecured loans to small businesses of up to
                 AU$250,000 Australian dollars, which will be 50-percent guaranteed by the government.
            • Responsible lending exemptions: The government has exempted lenders from responsible-
                 lending obligations for six months when they provide credit to existing small-business
                 customers. The exemption is aimed at speeding up small businesses’ access to credit.
            • Cash injections: Up to AU$100,000 in rebates on income tax for SMEs with employees.
            • Wage subsidies: Up to AU$1,500 per fortnight per employee for up to six months, and up
                 to 50 percent of an apprentice wage (subject to an AU$21,000 cap per apprentice).
            • Tax reductions: Instant asset write-off threshold increased from AU$30,000 to AU$150,000.
            All Australian banks are scrambling to implement the relief schemes alongside their own
            initiatives, such as repayment holidays and forbearance. The result is enormous pressure on
            institutions that are themselves struggling to adapt to disruption and a surge in demand.
            The speed at which banks respond to client needs will have a huge impact on the economic
            consequences of the crisis.

            3 Australian Government, Business

© Oliver Wyman                                                                                                   4
COVID-19: Implications for risk modelling

            (TEMPORARY) MISSING DEFAULTS
            These support initiatives are intended to protect borrowers from a liquidity squeeze that pushes
            them into arrears and ultimately default. They create three categories of borrowers:

            1. Borrowers that continue repayments according to their existing agreement, with or without
               cash injections or wage relief.
            2. Borrowers that do not repay because they have exercised a forbearance period or otherwise
               restructured their loans under a support scheme.
            3. Borrowers that do not repay and have no forbearance agreement.

            The first and third category of customers require no further elaboration; they are traditional
            performing and non-performing exposures respectively. However, the scale of government
            interventions and banks’ own initiatives put a significant portion of exposures temporarily into
            the second category.

            As the economy deteriorates, customers will migrate between these categories, with their rate
            and timing of migration dependent upon specific characteristics of each borrower and the
            support mechanisms on which they rely. As the peak of the social crisis passes and the economy
            restarts, further changes will be driven by the pace and sequencing of reversion to “business as
            usual” as authorities and banks ramp down their support.

            This creates considerable challenges for future loss forecasting efforts, as the specific benefits
            of government and financial sector response at the present time cannot be relied upon in
            future periods of economic shock. Some of the effects, such as the Guarantee Scheme loans,
            will be explicit and relatively simple to account for; others will be harder to strip out. Changes to
            amortization schedules and income tax rebates, for example, will fundamentally alter cashflow
            data and financial ratios. And this does not just affect loss forecasting; decisioning, pricing, and
            propensity modeling will all have different requirements for identifying and accounting for these
            interventions. Properly accounting for these will require reconstruction of operating cashflows
            as they would have been without the interventions. Banks that invest substantially in both credit
            risk and transaction data will be at an advantage.

            This year’s economic turmoil will provide crucial data for future credit models and stress tests,
            and critical insights into borrower and portfolio reactions, as banks work to improve their
            resilience. However, the unprecedented level of intervention will need to be stripped out of the
            data so that it does not distort base models.

© Oliver Wyman                                                                                                      5
COVID-19: Implications for risk modelling

            IMMEDIATE ACTION
            Banks must make decisions now that will determine how they collect, interpret, and ultimately
            use 2020 credit and default data in the future. Here are eight actions that, if taken now, will
            facilitate loss forecasting in the years to come.

            1. C
                apture borrower financial health: Banks should store a detailed record of the financial
               health of each borrower going into the crisis. In addition to the typical financial metrics, this
               should include transaction-level data to allow reconstruction of operating cashflows and their
               underlying growth and volatility, along with details of depreciation and other actions that are
               not captured in the top and bottom lines.

            2. C
                ollect business restructuring data: Banks should be capturing structural information
               about businesses before and during the crisis, such as the number of operating locations
               and employees, and legal structures. It will be important to be able to distinguish between
               temporary changes to these, such as furloughed or wage-reduced staff, versus permeant
               changes to the cost-base and operating structures of the organization coming out of the crisis.

            3. Track cashflows during support: Government cash injections and wage subsidies must be
               tracked carefully so that the underlying net operating cashflow, both before and during stress,
               can be extracted and analyzed at a later date. This will enable fair, accurate reconstruction of
               cost-to-income ratios and liquidity measures, as well as underwriting and decisioning with and
               without support.

            4. S
                yndicate data: Banks will need a way to identify and track borrowers that received support
               that is otherwise invisible, for instance where it flowed into an account at a different bank or
               was in the form of a rebate or deduction. This will require a level of data sharing that has yet
               to be solved. One option is syndication of the new reporting on support dissemination, for
               example, sharing back data captured in the ARF 920.04 reporting standard. This would provide
               banks with a system-wide view of the level and allocation of support, and potentially a way
               to track support across their portfolio at the borrower level. An alternative would be for the
               bureaus to capture and centralize this information, but this raises further questions on the
               level and limit of reporting.

            5. U
                nderstand sector-level idiosyncrasies: Restrictions on business activity, and targeted
               government support, are affecting different sectors in different ways. These must be identified
               and understood. For example, grocery cashflows will be significantly altered as a result of
               social distancing, panic buying, and wage relief. The impression this leaves on their cashflows
               will look very different from those of non-essential retail outlets that have pivoted to an online-
               only service.

            6. T
                rack distress and capital treatments in parallel: Regulators have allowed for special capital
               treatment for exposures restructured or otherwise modified under the support packages. This
               creates a trichotomy in banks’ data systems; an exposure must be treated as performing for
               capital purposes, is experiencing a payment holiday for operations, and has been restructured

            4 https://www.legislation.gov.au/Details/F2020L00434

© Oliver Wyman                                                                                                       6
COVID-19: Implications for risk modelling

               for credit purposes. Banks must have a way to append appropriate metadata for each of these
               use cases such that these exposures can be appropriately segmented and understood in
               the future.

            7. D
                igitize origination and restructuring: It is imperative that Australian banks standardize
               their data collection processes for the new lending conditions, whether new originations or
               adjustments to an existing exposure, so that in the future they can reconstruct and analyse
               the way decisions were made. Some banks are using new digital platforms to support the
               government schemes, increasing the chance that data captured are of good quality and
               complete. However, those without ready access to such channels must rapidly install a
               tactical solution.

            8. S
                tart loss forecasting now: Banks must quickly assess whether the incremental data being
               captured are sufficient to support future modelling efforts. The best way to do this is to
               put them to use in loss forecasting, which will and help identify gaps whilst there is time to
               remediate them.

            CONCLUSION
            These are highly challenging times for the Australian banking system, and banks will have to
            make a great effort to balance competing priorities. As they take urgent action to deal with the
            severe economic downturn, they should not forget that they will, in future need, to account for
            the substantial echo that will be left in default and financial data. This implies careful capture of
            data today that will let them understand the myriad idiosyncratic aspects of the crisis. If banks do
            this, they will emerge from the crisis stronger, and be better able to prepare for the next one.

© Oliver Wyman                                                                                                      7
Oliver Wyman is a global leader in management consulting that combines deep industry knowledge with specialized
expertise in strategy, operations, risk management, and organization transformation.

For more information please contact the marketing department by email at info-FS@oliverwyman.com or
by phone at one of the following locations:

Americas                               EMEA                                   Asia Pacific
+1 212 541 8100                        +44 20 7333 8333                       +65 6510 9700

AUTHORS

James Gordon, Principal, Digital
james.gordon@oliverwyman.com

David Howard-Jones, Partner, Finance & Risk and Public Policy
david.howardjones@oliverwyman.com

Wolfram Hedrich, Partner, Finance & Risk and Public Policy
wolfram.hedrich@oliverwyman.com

Copyright © 2020 Oliver Wyman
All rights reserved. This report may not be reproduced or redistributed, in whole or in part, without the written permission of Oliver Wyman
and Oliver Wyman accepts no liability whatsoever for the actions of third parties in this respect.
The information and opinions in this report were prepared by Oliver Wyman. This report is not investment advice and should not be relied on
for such advice or as a substitute for consultation with professional accountants, tax, legal or financial advisors. Oliver Wyman has made every
effort to use reliable, up-to-date and comprehensive information and analysis, but all information is provided without warranty of any kind,
express or implied. Oliver Wyman disclaims any responsibility to update the information or conclusions in this report. Oliver Wyman accepts no
liability for any loss arising from any action taken or refrained from as a result of information contained in this report or any reports or sources
of information referred to herein, or for any consequential, special or similar damages even if advised of the possibility of such damages. The
report is not an offer to buy or sell securities or a solicitation of an offer to buy or sell securities. This report may not be sold without the written
consent of Oliver Wyman.

Oliver Wyman – A Marsh & McLennan Company                                                                                www.oliverwyman.com
You can also read