` General Insurance Stress Test 2015 Scenario Specification, Guidelines and Instructions 1 July 2015 1 CONTENTS Introduction.............................................................................................................................................. 3 1. European Windstorm and flood set of events .................................................................................. 6 2. US Hurricane set of events ............................................................................................................ 10 3. Synchronous set of terrorism events ............................................................................................. 13 4. Motor liability stress test................................................................................................................. 16 5. Economic shock ............................................................................................................................. 18 6. Supply Chain disturbance .............................................................................................................. 25 7. Liability / Rserve stress test ........................................................................................................... 26 8. Solar Flare / Geomagnetic storms ................................................................................................. 28 9. Cyber Loss ..................................................................................................................................... 29 10. 1-in-200 insurance loss stress test (insurer specific) ..................................................................... 31 11. Reverse stress test (insurer specific) ............................................................................................. 32 12. Definitions ...................................................................................................................................... 33 13. Further notes for firms with defined pension schemes – only applicable for scenario 5 ............... 36 2 INTRODUCTION This document provides details of the stress tests to be evaluated by firms that are within the scope of the PRA’s General Insurance Stress Test exercise in 2015 (’GIST 2015’). This document also provides notes and instructions as to how to complete the Microsoft Excel workbook “GIST 2015.xls” which firms have been provided to record the results of each stress test. Stress tests The stress tests have been developed to address potential market-wide and firm-specific issues. The below grouping illustrates the primary purpose for each of the stress tests. Note: stress tests within Section B do not imply that establishing a market-wide stress would be inappropriate – rather that these are scenarios which the PRA has not defined in sufficient depth to ensure reasonable consistency between firms. A: Understanding how specific events impact the UK insurance industry (consistency between firms is important) 1. 2. 3. 4. 5. European windstorm and flood set of events US Hurricane set of events Synchronous terrorism set of events Motor Liability Stress Test Economic shock B: Understanding the extent to which firms have considered specific potential stresses and the impact on their Solvency II Balance Sheet 6. 7. 8. 9. 10. 11. Supply chain disturbance Liability / Reserve stress test Solar Flare / Geomagnetic Storms Cyber Loss Worst case own – 1-in-200 insurance loss stress test (insurer specific) Reverse stress test (insurer specific) In addition, for each stress test firms are required to provide details of the expected reinsurance recoveries split by reinsurers, and their expectation of the likelihood of such an event. Firms are not required to recalculate their full Solvency II Balance Sheet after each stress – however, firms are required to provide the gross and net loss and where appropriate any other mitigating impacts that would impact the size of the underwriting loss – for instance an allowance for tax (see subsequent section below: “Allowance for tax after stress event”). High level notes are provided in this document, and firms are encouraged to apply their own methodologies to calculate the expected loss under each described scenario, reflecting their specific risk profile and insurance coverage provided. 3 Coverage for the stress test submission Where firms are submitting an internal model application the stress test submission must be aligned to that application. For the avoidance of doubt for UK Groups that are applying for a group application the submission should cover all group-wide operations (including those outside the UK). For those firms not in the internal model approval process (IMAP) a separate submission will be required for each legal entity, unless your supervisor has communicated otherwise in the covering letter for this submission. Where firms are uncertain as to the scope required please email: PRA_GIStressTesting2015@bankofengland.gsi.gov.uk Opening balance sheet: Standard formula vs internal model Solvency Capital Requirement Firms are only required to calculate their capital requirements before the stress event. For firms in IMAP the capital requirement should be provided on both an internal model and standard formula basis. Firms not currently in IMAP should at a minimum provide capital requirements using the standard formula basis, and we encourage firms to provide the internal model results where available. Materiality Firms should complete all scenarios unless they can demonstrate that given their specific risk coverage the impact is immaterial. In this case immateriality is defined as less than 5% of total net written premium. Firms should include a breakdown of all reinsurers where expected recoveries are more than 2% of the total recoverable. Emergence of risk In each case the shock is assumed to apply reasonably instantaneously, so firms should not consider the emergence of risk; it is only the ultimate view of the loss event that should be recorded. Allowance for tax after the stress event The aim of the stress test is to understand the quantum of the risks on the insurer’s balance sheet and in aggregate across the industry. The PRA recognises that each insurer’s tax position will differ; as a result we require all stresses before any allowance for tax. However, firms may wish to provide details of the anticipated tax impact for each scenario within the template (under the free-form text box titled: “What would be the consequences of this scenario, and how would it impact your business model”). The PRA will take this into consideration when assessing the overall results. 4 Reporting and sign-off requirements This exercise is to be carried out only by firms selected by the PRA. Firms who have not received a request are not required to complete the workbook. For each stress test firms are required to submit a number of outputs that are standard across scenarios, as well as additional information specific to each scenario that will allow the PRA to understand the impact of each stress in greater detail. All parts of the template are required to be completed. On submission a senior executive is required to confirm they are satisfied with the completion of the template for each of the relevant stress tests. Deadline for submission Submission of the completed excel template is required by 5pm on 1 October 2015. The Excel work book should be saved ensuring that Firm Name and FRN number are contained within the file name and the subject of the email. Submissions should be sent to PRA_GIStressTesting2015@bankofengland.gsi.gov.uk Queries The stress tests have been developed in conjunction with a number of industry participants. Any queries should be submitted to PRA_GIStressTesting2015@bankofengland.gsi.gov.uk. Please ensure that the Firm Name and FRN number is included within the subject of the email. 5 1. EUROPEAN WINDSTORM AND FLOOD SET OF EVENTS 1.1 EVENT DEFINITION This stress test is for a cluster of two severe European windstorms followed by a severe flood. The windstorm tracks have been based on the previous two storms that occurred in 1987 and 1990. Firms are to assume that the events are due to different weather systems and are sufficiently separated in time to be considered three separate events for the purposes of reinsurance recoveries. 1.2 ASSUMPTIONS Firms are asked to estimate the size of the loss per event and in aggregate using their natural catastrophe modelling capabilities. In estimating the gross loss, firms should allow for storm surge but not demand surge or post loss amplification. For demand surge or post loss amplification, an uplift of 5% should be applied to all three events. Should the firm not have access to suitable modelling capabilities, a set of damage ratios has been provided by the PRA for each windstorm and the flood event in Annexes 1, 2 and 3 of the GIST 2015.xls workbook. Firms may use these to estimate the loss. However, firms are strongly encouraged to carry out their own modelling to estimate the size of potential losses. Firms should consider what management actions including changes to their reinsurance programmes they may take following the events. These should be described with the estimated associated costs, if any, disclosed and allowed for post event in the above calculations. Firms should assume events fall under the same treaty year, that any changes made to the reinsurance programme do not incept before the events occurred, and should include the impact of both inwards and outwards reinstatement premiums. 1.2.1 First windstorm The map below illustrates the footprint for the first windstorm event which is assumed to match the characteristics of windstorm 1987J with top gust speeds greater than 45m/s. Loss Amounts United Kingdom France Norway Netherlands Belgium Denmark Germany Sweden Switzerland Luxembourg Ireland Austria Europe Source: AIR 6 4,009,688,681 1,199,769,711 513,515,946 123,222,591 64,506,803 61,603,946 29,788,771 21,032,732 16,891,912 828,281 66,709 54,925 6,040,971,008 At today’s values, this is estimated to cause GBP 4.0 billion of losses in the UK and GBP 6.0 billion of 1 losses across Europe, split as in the table above . The matching AIR Event ID would be 410000003 (i.e. event 3 from the European Wind Historical Catalogue) and the matching RMS Event ID would be 2899004 (we are aware that the industry loss estimates between AIR and RMS differ). 1.2.2 Second windstorm The map below illustrates the footprint for the second windstorm event which is assumed to match the characteristics of windstorm Daria with top gust speeds greater than 40 m/s. United Kingdom Netherlands Belgium Germany France Denmark Sweden Ireland Switzerland Norway Luxembourg Poland Austria Czech Republic Estonia Lithuania Latvia Finland Europe Loss Amounts 5,901,536,317 1,387,715,983 803,632,543 699,549,253 566,145,010 268,619,076 172,152,565 110,298,135 46,195,209 31,766,937 10,329,130 2,051,204 1,279,040 1,033,491 300,264 129,439 129,125 13,904 10,002,876,625 Source: AIR At today’s values, this is estimated to cause GBP 5.9 billion of losses in the UK and GBP 10.0 billion of losses across Europe, split as in the table above. The matching AIR Event ID would be 410000004 (i.e. event 4 from the European Wind Historical Catalogue) and the matching RMS Event ID would be 2899009 (we are aware that the industry loss estimates between AIR and RMS differ). 1 Industry loss estimates for the events are based on AIR Worldwide’s Industry Exposure Database 7 1.2.3 Third event: UK flood For the third event, firms are to assume precipitation induced flooding in the south of England with the most impacted areas in decreasing size of loss being London East, Peterborough, Oxford and Bristol. The map below illustrates the footprint for the third event, with some of the characteristics of the flood detailed in the table alongside. Event Duration (hours) Average Precipitation Rate (mm/hr) Average Excess Run-off (mm/hr) Average Discharge relative to 2 Year Flow Flood 144 1.22 0.12 2.29 Source: AIR The flood event is assumed to result in severe flooding with the event lasting 144 hours across the south of England leading to an industry loss of GBP 4.8 billion. The closest matching AIR Event ID would be 920017260. Alternatively, firms using RMS may model this event using RMS flood event ID 1945288. The PRA is aware that the footprints are not spatially exactly similar although the main areas impacted are broadly the same and the industry loss estimates correspond. Source: RMS 8 1.3 REPORTING Firms should provide separate gross estimates for each event, work out how the reinsurance programmes would respond, and estimate the net loss to the firm. Data assumptions, where made, should be disclosed including for example: the allowance made for locations not geo-coded or insufficiently accurately geo-coded, commercial policy deductibles and sub limits where not recorded, and data fields such as construction type, occupancy, and others where not recorded. Firms are also asked to disclose their estimates of the secondary uncertainty around their loss estimates, the vendor model and version used or whether they have used the PRA supplied factors, as well as any other assumptions made in the loss estimation. The gross loss estimate should break down the loss between: residential property damage losses commercial property damage losses (including industrial & agricultural) business interruption losses contingent business interruption losses motor losses marine and energy losses liability losses and other type of losses. 9 2. US HURRICANE SET OF EVENTS 2.1 EVENT DEFINITION This stress test is for a Katrina, Rita and Wilma (2005) type of scenario where a series of three major US hurricanes occur in the same year. 2.2 ASSUMPTIONS As for the European scenario, firms are expected to carry out their own modelling to estimate the impact of the losses. In estimating the gross loss, firms should allow for storm surge but not demand surge or post loss amplification. For demand surge or post loss amplification, an uplift of 25% should be applied. Should the firm does not have access to suitable modelling capabilities, a set of damage ratios has been provided by the PRA for each hurricane in Annexes 4, 5 and 6 of the GIST 2015.xls workbook which firms may use to estimate the loss. Firms should consider what management actions including changes to their reinsurance programmes they may take following the events. These should be described with the estimated associated costs, if any, disclosed and allowed for post event in the above calculations. Firms should assume all three hurricanes fall under the same reinsurance treaty year and should include the impact of both inwards and outwards reinstatement premiums. 2.2.1 First hurricane through Florida before making landfall in Texas The map below illustrates the track of the first hurricane of category 3 on the Saffir-Simpson scale making landfall in Palm Beach, Florida. The hurricane is assumed to cause losses across the Gulf of Mexico before making landfall again as a Category 4 hurricane in Chambers, Texas. It will also create some losses across the Caribbean. The table provides details of the hurricane’s US landfalls. Saffir-Simpson Category Central Pressure (mbar) Maximum Windspeed (mph) Maximum Radius (miles) Speed (mph) Angle (degrees) Longitude (degrees) Latitude (degrees) State County Source: AIR 10 US Landfall 1 3 952.2 119.5 25 6.5 -35.4 -80.089 26.382 Florida Palm Beach US Landfall 2 4 929.6 139.5 23 5.7 -16.3 -94.200 29.579 Texas Chambers 2 The resulting industry loss is assumed to be some USD 56.0 billion , with the AIR Event ID being 270133233 and the closest matching RMS ID being 2864983 (we are aware that the industry loss estimates between AIR and RMS differ). 2.2.2 Second hurricane hitting the US North East The map below illustrates the track for the second category 3 hurricane making landfall in New Jersey, and causing losses across the north-eastern US states of New York, Connecticut, Pennsylvania and Delaware. Details of the hurricane’s landfall are provided in the table. Saffir-Simpson Category Central Pressure (mbar) Maximum Windspeed (mph) Maximum Radius (miles) Speed (mph) Angle (degrees) Longitude (degrees) Latitude (degrees) State County US Landfall 1 3 956.3 111.1 38.1 33 -49.2 -74.456 39.397 New Jersey Atlantic Source: AIR The resulting industry loss is assumed to be some USD 24.4 billion, with the AIR Event ID being 270093160 and the closest matching RMS Event ID being 2851343 (we are aware that the industry loss estimates between AIR and RMS differ). 2 Industry loss estimates for the events are based on AIR Worldwide’s Industry Exposure Database 11 2.2.3 Third hurricane going through Florida before drifting north The map below illustrates the track for the third category 4 hurricane making landfall in Lee, Florida before making landfall again as a category 3 hurricane in South Carolina. Details of the hurricane’s landfalls are provided in the table. Saffir-Simpson Category Central Pressure (mbar) Maximum Windspeed (mph) Maximum Radius (miles) Speed (mph) Angle (degrees) Longitude (degrees) Latitude (degrees) State County Landfall 1 4 941.8 Landfall 2 3 953.6 133.7 25.6 14.1 48.7 -82.244 26.626 121.4 18.3 12.1 22.9 -78.709 33.797 South Carolina Horry Florida Lee Source: AIR The resulting industry loss is assumed to be some USD 37.7 billion, with the AIR Event ID being 2700163397 and the closest matching RMS Event ID being 2850375 (we are aware that the industry loss estimates between AIR and RMS differ). 2.3 REPORTING Firms should provide separate gross estimates for each event, work out how the reinsurance programmes would respond, and estimate the net loss to the firm. Data assumptions, where made, should be disclosed including for example: the allowance made for locations not geo-coded or insufficiently accurately geo-coded; commercial policy deductibles and sub limits where not recorded; and data fields such as construction type, occupancy, and others where not recorded. Firms are also asked to disclose their estimates of the secondary uncertainty around their loss estimates, the vendor model and version used or whether they have used the PRA supplied factors, as well as any other assumptions made in the loss estimation. Results provided should break down the gross loss estimates at least between: residential property damage losses commercial property damage losses (including industrial & agricultural) business interruption losses contingent business interruption losses motor losses marine and energy losses liability losses and other type of losses 12 3. SYNCHRONOUS SET OF TERRORISM EVENTS 3.1 EVENT DEFINITION This stress test is for a synchronous set of three terrorism events in London, New York and a third city of firm’s choice, each using a 2 ton bomb dissimulated in a medium sized box van. The terrorist attacks are assumed to be coordinated by one terrorist organisation and detonated outside the entrance or service bay of each target location on a Monday at 3 pm UK time for the London event, 10am EST for the New York event, and peak business hours on the same day for the third event of firm’s choice. 3.2 ASUMPTIONS 3.2.1 General Firms should provide separate gross estimates for each terrorism event; estimate the benefit of recoveries from Pool Re, from the Terrorism Risk Insurance Program Reauthorization Act (TRIPRA) 2015, from any other government or pool provided cover or from any reinsurance purchased if relevant; and quantify the net loss to the firm. Recoveries from pools should be listed using the same format to capture expected recoveries from reinsurers. Firms are encouraged to carry out their own modelling of losses using the methodology or modelling framework they consider most appropriate. In assessing the potential loss, firms should consider fire following, business interruption costs and allow for clean-up costs. Firms should consider death and injuries including total and partial disability. In the absence of any such modelling, firms may default to the following PRA provided damage, 3 fatality and disability ratios within circular concentric zones centred on the detonation location. For property damage and business interruption, firms should assume a 60% damage ratio within a circular zone of 100m radius, a 20% damage ratio beyond 100m up to 200m, and a 5% damage ratio beyond 200m up to 400m. For fatalities, firms should assume a 10% fatality rate within a circular zone of 100m radius and a 1% fatality rate ratio beyond 100m up to 200m. For disabilities, firms should assume a 15% disability rate within a circular zone of 100m radius, a 5% disability rate ratio beyond 100m up to 200m, and a 1% disability rate beyond 200m to 400m. Firms may assume the disabilities are split equally between total and partial disabilities. Firms will also need to consider their own operational risk losses where relevant, including any retained losses under their own insurance programmes. Firms should assume events fall under the same treaty year and that no changes are made to the government covers or reinsurance programmes protecting them at the time of the loss. Firms should estimate, and disclose, the costs of any management actions including changes to their reinsurance programmes they would expect to take post the loss. 3 Damage ratios and fatality rates have been set after discussions with RMS’ terrorism practice. 13 3.2.2 Terrorism event on Lime Street Assume the first event in the set is a 2 ton bomb dissimulated in a medium sized box van and detonated on a Monday at 3pm UK time next to the Lloyd’s building as shown on the map below. The o polar coordinates of the location of the bomb are assumed to be (latitude 51.513558 , longitude o 0.081547 ). 14 3.2.3 Terrorism event at the Nasdaq Stock Market Assume the second event in the set is a 2 ton bomb dissimulated in a medium sized box van and detonated on a Monday at 10am EST at the Nasdaq building in Manhattan as shown on the map o below. The polar coordinates of the location of the bomb are assumed to be (latitude 40.756043 , o longitude -73.985804 ). 3.2.4 Terrorism loss For the third event, firms are to choose a location in Paris, Frankfurt, Hong Kong, Singapore or Sydney depending on where they feel they have the most significant exposures. Firms should assume this third event in the set is also for a 2 ton bomb dissimulated in a medium sized box van, detonated at peak time on the same day as the first two events. 3.3 REPORTING Results provided should break down the gross loss estimates at least between: Commercial property damage losses including buildings, contents and business interruption. If terrorism coverage is excluded, fire following loss should be included to the extent cover is provided or required by law, Residential losses, Personal accident losses, Employer’s Liability/Workers Compensation losses, Stand-alone political risk, war or terrorism specific covers provided, any other insurance class of business, and own operational risk losses. 15 4. MOTOR LIABILITY STRESS TEST 4.1 EVENT DEFINITION This is a two part scenario combining a change in the legal environment surrounding future PPOs (driving more and bigger PPOs) plus the imposition of a retrospective review, with associated costs. This scenario is only relevant for insurers who have a material UK Motor insurance portfolio. 4.2 ASSUMPTIONS 4.2.1 Part I Following a new legal precedent, settling large motor claims as a PPO award becomes the normal position. Furthermore, settling the lost earnings element of the claim with a PPO, rather than with a traditional lump sum, also becomes the standard practice. In assessing the first part of this scenario firms should assume that their own propensity for large motor claims (excess of £1m) to settle including a PPO award rises to 90%, regardless of experience to date. The definition of large claims can be assumed to be based upon that used by the IFOA PPO Working Party. (£1 million in 2011 values indexed at 7% per settlement year.) The £1m definition applies to the bodily injury element of the claim only and is per claimant. The element of the claim settled as a PPO (rather than as a lump sum) is assumed to have been in respect of future care costs only prior to this scenario occurring. You should assume that this scenario applies to all unsettled large claims (IBNR as well as known reported claims), defined as above, as at 31.12.2014, and that the valuation discount rate assumptions are: Gross nominal investment return 3.5% per annum; Average weekly earnings 4.0% per annum; Constant universal (overall percentiles) rate for ASHE 6115 = 4.5% per annum; and Real discount rate = 1.045/1.035 – 1 = -0.966% per annum Unsettled and IBNR claims should be discounted to the assumed future date of settlement. Please use your usual PPO mortality and longevity assumptions. You should make no allowance for any additional Reinsurance credit risk. 4.2.2 Part 2 This part of the scenario is intended to evaluate the potential impact of a retrospective legal change, which even seeks to unpick historical settlements. It is retrospective only and does not apply to future claims. The scenario assumes that historically settled motor claims above £1m, settled after Jan 1 2008, could be “re-opened” and the outstanding liability be settled on the current PPO basis. This represents 7 years of claims settlements (as at end of 2014). It quantifies a one off payment with the aim of addressing the economic gap between the lump sum that a claimant did receive and the economic value that they would have received had the claim settled as a PPO award (i.e. ensuring the claimant is indemnified for any investment, inflation or mortality risk), thus including a PPO income stream in respect of the care costs that they claimed for, instead of the equivalent lump sum that was provided. To assess the potential retrospective impact of this scenario, firms are required to estimate the difference between the lump sum that was paid and the economic cost that would have been incurred if the claim had settled with a PPO. Definitions of large claims and valuation discount rates should be assumed to be the same as in part 1. 16 Firms should assess the difference between the care cost element of the claim that was settled using a lump sum and the economic value of those future care costs using the economic assumptions set out above together with your usual PPO mortality and longevity assumptions. The estimate should be as at the date of settlement and inflated to 31.12.2014 at 7% per annum. Firms should assume that there are no margins available within any lump sum that was paid that could be used to offset any element of this scenario and that reinsurance does not respond to this compensation payment, Please provide impact assessments for Parts 1 and 2 of the scenario separately. 4.3 REPORTING In addition to the standard Balance Sheet impact from this stress test we require separate gross loss impacts for each of Part 1 and Part 2. For Part 1 we also require your expected loss ratio in respect of your motor portfolio for the 2014 4 accident year (on both a gross and net of reinsurance basis), as well as details of the PPO loss ratio both before and after the stress test. Note we recognise that this will not directly correspond to the Part 1 stress, as it will only cover PPOs attached to the 2014 accident year. In the notes for Part 1 please provide details of your normal valuation basis, including mortality and longevity assumptions. Please also provide details of the split of the gross loss between the IBNR claims and known claims, and explain your approach to estimating the impact of IBNR claims. 4 PPO loss ratio means the part of the overall Motor Loss Ratio that is in respect of the PPO element of the large claims 17 5. ECONOMIC SHOCK 5.1 EVENT DEFINITION The economic shock scenario is consistent with the 2015 Financial Policy Committee Stress Test of the UK Banking system. A brief outline of this scenario is provided below: Globally area stress Global growth disappoints materially relative to expectations and disinflationary pressures build up. This triggers a rapid deterioration of market sentiment globally. Risk appetite abruptly diminishes and market participants attempt to de-risk their portfolios, generating safe-haven capital flows to highquality US assets. The dollar appreciates against a wide range of currencies, especially those of emerging market economies. Liquidity in some markets becomes seriously impaired and credit risk premia rise sharply. Commodity prices fall further, putting additional downward pressure on global inflation. This leads to falls in consumption, investment and property prices. The deterioration of global financial market sentiment is also evident in the UK. In this scenario, it is assumed that policymakers observe these developments as a series of unexpected shocks. Additional monetary policy stimulus is pursued, which has the effect of lowering the yield curve. Economic output falls across a number of regions including the euro area, emerging market economies and the UK. The marked reduction in global nominal demand causes a further reduction in commodity prices. The oil price troughs at US$38 per barrel and remains low throughout the scenario. Other commodity markets also see price falls and remain very weak throughout the scenario. This puts additional downward pressure on global inflation. Falls in commercial property prices are more pronounced, reflecting the larger average overhang of unsold property in that market. This is associated with sharp falls in real estate investment and industries associated with construction. Euro area stress In the euro area, output growth slows due to a combination of international spillovers and domestic amplifications. Slower world trade results in materially lower demand for exports, a channel that acts more strongly for the euro area ‘core’ given the greater trade linkages with Asian economies. As elsewhere, the weaker economic conditions in the euro area lead to higher risk premia. This risk aversion causes the euro to depreciate by around 25% against the US dollar and by 15% against sterling. Within the euro area, risk premia rise most strongly in asset markets for the more highly indebted sovereigns, households and firms, given the effect that falling nominal GDP has in increasing the real burden of debt. In the UK, output growth turns negative as export demand falls sharply and there are spillovers through confidence effects. Financial linkages provide another channel of transmission. The reduction in inflationary pressures from the slowing economy and falling commodity prices results in inflation turning negative during the scenario. This constitutes the largest fall in the price level in the UK for over 80 years. As elsewhere, additional monetary policy stimulus is pursued and Bank Rate is reduced to zero. Operating conditions are particularly challenging for UK corporates in this scenario. If insurers require additional detail and colour around the scenario then refer to the Bank of England Banking Stress Test 2015. http://www.bankofengland.co.uk/financialstability/Documents/stresstesting/2015/keyelements.pdf Insurers are not expected to implement the full banking stress test which considers a 5 year horizon, nor apply the full economic and asset stress factors required by banks and building societies. Instead, we provide the relevant factors that insurers need to apply in Section 5.2 below 18 Please note it is inevitable that these factors will not capture the breadth of all firms’ assets, geographies or products. Where this is the case we expect insurers to consider suitable factors and make their own judgments / assessment that are appropriate and within the spirit of the wider economic stress illustrated above. When applying expert judgment in relation to the asset shocks we would expect insurers to consider the worst market moves observed in the historical periods per region detailed in the table below. Geographical region of positions Asia and Emerging Markets Europe and the United States Historical period 2008 H2 2011 H2 and 2012 H1 The economic assumptions firms should apply under stress are set out in 5.2. Definitions are available in the spreadsheet used to collect the scenario feedback. Firms should assess the extent to which this would impact their insurance operations as well as the impact on their investments. When considering changes in Pension Scheme commitments, for firms with defined benefit schemes, we expect insurers to assess the impact of the instantaneous asset shock on their pension scheme investments and report the extent to which this results in a surplus or deficit. We acknowledge that firms have a number of options in which to manage funding pension schemes under stress and this will be considered when evaluating the scenario outcome. Further notes as to how firms should apply the economic scenario to pensions projections are provided in Section 13. 5.2 ASSUMPTIONS Macroeconomic variables (Base and after Stress) UK area Base Stress % Change UK real GDP 432,814.0 422,690.0 -2.3% UK nominal GDP 452,391.3 444,767.9 -1.7% 128.3 127.1 -0.9% 5.7 7.7 35.1% UK corporate profits 208,952.0 199,767.0 -4.4% UK household income UK CPI (assume same change for RPI) UK unemployment rate 313,560.0 316,468.0 0.9% UK residential property price index 100.0 87.6 -12.4% UK commercial real estate price index - aggregate 100.0 83.0 -17.0% UK commercial real estate price index - prime 100.0 80.1 -19.9% UK commercial real estate price index - secondary 100.0 84.1 -15.9% UK equity prices 100.0 64.0 -36.0% 0.5 0.0 -100.0% Sterling Investment Grade corporate bond spread 141.0 452.0 220.6% Sterling High Yield corporate bond spread 441.0 1,762.0 299.5% PPP-weighted World real GDP 100.0 99.3 -0.7% Oil price 76.1 43.3 -43.1% Volatility index 16.0 44.9 180.6% GBP-EUR exchange rate index 100.0 115.2 15.2% GBP-USD exchange rate index 100.0 88.7 -11.3% G20 emerging economy dollar exchange rate index 100.0 75.2 -24.8% Bank Rate 19 Euro area aggregate (including periphery) Real GDP Base Stress % Change 100.0 98.5 -1.5% Consumer price inflation 0.2 -1.3 -750.0% Unemployment rate 11.4 12.8 12.3% ECB policy rate 0.1 0.0 -100.0% Residential property price index 100.0 88.2 -11.8% Commercial real estate price index 100.0 89.0 -11.0% Asset Class Region Risk Factor Liquidity Horizon EURUSD -0.2 GBPUSD -10% USDCHF 5% USDTRY 44% 1y FX Equities Credit Rates Europe Europe Europe Europe USDRUB 30% FTSE100 INDEX -36% EUROSTOXX50 -26% MSCI EMU -28% ITRAXX 260% ITRAXX XOVER 240% ITRAXX SNR FIN 240% GER GOV 1Y -1600% GER GOV 5Y -300% GER GOV 10Y -2300% EUR SW 1Y 6400% EUR SW 5Y 1800% EUR SW 10Y EUR PERIPH EX GR 1Y EUR PERIPH EX GR 5Y EUR PERIPH EX GR 10Y GBP GOV 1Y -1200% (1) The convention of 'Ccy1Ccy2' represents the number of Ccy2 per Ccy1. (2) These are relative percentage shifts. For example, if USDJPY spot rate on 20Feb-2015 is 120, the 1d 5% shock will take the spot rate to 120 x (1 + 5%) = 126 These are relative percentage shifts. If the FTSE100 index on 20-Feb-2015 is 6500, a 1d -5% shock will take the index value to 6500 x (1 - 5%) = 6175. These are relative percentage shifts to the 5yr credit spreads. (also to be applied to all Corporate credit including structured products & loans) 19600% 392 298 -42 GBP GOV 5Y -78 GBP GOV 10Y -87 GBP GOV 20Y -108 GBP SW 1Y -3 GBP SW 5Y -74 GBP SW 10Y -74 GBP SW 20Y -73 20 These are absolute basis points shifts to the annual interest rate. For example, if the 1-year CN government yield is 500 bps, the -20 bps 1d shock will take the rate to 480 bps. Euro area aggregate (excluding periphery) Base Stress % Change 100.0 98.9 -1.1% Consumer price inflation 0.4 -1.2 -400.0% Unemployment rate 7.3 8.0 9.6% ECB policy rate 0.1 0.0 -100.0% Residential property price index 100.0 87.6 -12.4% Commercial real estate price index 100.0 88.9 -11.1% Real GDP Euro area periphery Base Stress % Change 100.0 97.6 -2.4% Consumer price inflation -0.3 -1.4 366.7% Unemployment rate 18.0 20.5 13.9% ECB policy rate 0.1 0.0 -100.0% Residential property price index 100.0 89.5 -10.5% Commercial real estate price index 100.0 89.5 -10.5% Base Stress % Change 100.0 99.8 -0.2% 5.7 7.1 24.6% 100.0 78.4 -21.6% 0.3 0.3 0.0% US residential property price index 100.0 93.9 -6.1% US commercial real estate price index 100.0 90.1 -9.9% US dollar IG corporate bond spread 133.0 261.0 96.2% US dollar HY corporate bond spread 468.0 1,208.0 158.1% Real GDP US US real GDP US unemployment rate US equity prices US policy rate 21 Asset Class Region Risk Factor Liquidity Horizon 1y Equities US Credit US S&P 500 -0.2 CDX HY 130% CDX IG 100% USD GOV 1Y -3500% 11900% 10600% 2200% 10200% -9500% USD GOV 5Y Rates US USD GOV 10Y USD SW 1Y USD SW 5Y USD SW 10Y 22 These are relative percentage shifts. If the FTSE100 index on 20-Feb-2015 is 6500, a 1d -5% shock will take the index value to 6500 x (1 - 5%) = 6175. These are relative percentage shifts to the 5yr credit spreads. These are absolute basis points shifts to the annual interest rate. For example, if the 1-year CN government yield is 500 bps, the -20 bps 1d shock will take the rate to 480 bps. Asia Region FX Equities Credit Rates Asia Asia Asia Asia Risk Factor Liquidity Horizons 1y USDJPY 0.2 USDCNY 10% USDCNH 10% USDHKD 0% USDINR 21% USDKRW 44% USDMYR 15% USDIDR 31% USDTWD 12% USDSGD 13% AUDUSD -35% NIKKEI225 HANG SENG INDEX SENSEX -38% KOSPI -55% ITRAXX JAPAN ITRAXX EX JAPAN IG ITRAXX EX JAPAN HY SG GOV 1Y 227% SG GOV 3Y 100 HK GOV 1Y 100 HK GOV 3Y 110 CN GOV 1Y -100 CN GOV 3Y -100 TRY GOV 1Y 500 TRY GOV 3Y 500 SG SW 1Y 130 -65% -53% 400% (1) The convention of 'Ccy1Ccy2' represents the number of Ccy2 per Ccy1. (2) These are relative percentage shifts. For example, if USDJPY spot rate on 20Feb-2015 is 120, the 1d 5% shock will take the spot rate to 120 x (1 + 5%) = 126 These are relative percentage shifts. If the FTSE100 index on 20-Feb-2015 is 6500, a 1d -5% shock will take the index value to 6500 x (1 - 5%) = 6175. These are relative percentage shifts to the 5yr credit spreads. 750% 100 SGSW 3Y 112 HK SW 1Y 180 HK SW 3Y 120 CN SW 1Y 0 CN SW 3Y 0 TRY SW 1Y 400 TRY SW 3Y 400 23 These are absolute basis points shifts to the annual interest rate. For example, if the 1-year CN government yield is 500 bps, the -20 bps 1d shock will take the rate to 480 bps. A copy of these assumptions is provided in the Annex of the GIST 2015.xls workbook. 5.3 REPORTING Firms should assess the impact on both their investments and their underwriting activities. Consideration should be given to lines of business such as credit insurance that are directly related to economic conditions, but firms should also consider other lines of business that could be indirectly impacted by the wider economic climate described above. Note: Within the template for this particular stress test Gross and Net aggregate loss should reflect the total loss arising from both investment and underwriting activities. 24 6. SUPPLY CHAIN DISTURBANCE 6.1 EVENT DEFINITION The earthquake & tsunami in Japan followed by the floods in Thailand in 2011 resulted in a significant portion of the world’s suppliers of hard drives to be out of business for several months. The knock-on impact to PC manufacturers around the world illustrated the interconnectivity risk that can arise from supply chain disturbances, as well as the complexities in identifying concentrations and aggregations of risk. In recognising that insurers are assisting industry to mitigate this risk, this stress test is aimed at understanding the ability of insurers to identify the potential for aggregations in terms of the contingent business interruption across different insureds. 6.2 ASSUMPTIONS A supply chain loss could arise if a leading Supplier of a key product with few substitutes suffers a risk or catastrophic loss leading to a significant disruption of supply. Given the complexity of global supply chains, the varied nature of different commercial portfolios and differences in coverage provided by insurers, we have not identified or named a Supplier or defined a loss scenario. Instead, firms are asked to identify the five suppliers to whom they have the largest exposures under the Contingent Business Interruption (CBI) cover included in their commercial policies. For the stress test scenario, firms are to assume a Gross Loss corresponding to their largest exposure i.e. full limit losses arising from the Supplier where they have the largest exposure being inoperative indefinitely. Firms should consider not only Suppliers named in their commercial policies but should also make some allowance for exposures which might arise under Unnamed Suppliers CBI coverage provided. 6.3 REPORTING Firms are required to provide details of the top 5 CBI exposures, noting that only the largest exposure will be included for the purposes of the stress test. Firms should disclose any additional exposures they may have to these Suppliers from other policies they write e.g. if they also participate on the Supplier’s commercial insurances. Firms are also required to state the method used to identify their top 5 suppliers. 25 7. LIABILITY / RESERVE STRESS TEST 7.1 EVENT DEFINITION The intention of this stress test is to capture the potential for systemic losses arising from product or process activity (for instance a product recall loss impacting an industry) or from professional / service work (for instance professional negligence claims arising following falls in asset values). These claims often take long to settle, can span more than one underwriting year, result in accumulations across different sectors, and typically impact insurers through reserve deteriorations. Currently, the most common methods for capturing such events are via techniques that exploit past claim payment and reserve development. This stress test is intended to move beyond consideration of pure historic events, and to leverage exposure information. As a result, we have not prescribed an event definition, instead, we require firms to provide details of their largest sector exposures and to apply specific “damage ratios” to determine the gross aggregate loss following the stress. Firms have the ability to provide feedback on the plausibility of the loss calculated under this event (by providing the return period for such an event), as well as a description of the event that could give rise to such a loss. 7.2 ASSUMPTIONS In discussion with some insurers it is clear that historic exposure data may not be readily available for older underwriting years – as a result we have currently restricted this stress to events impacting the 2014 underwriting year only. Firms are required to identify their Total Exposed Aggregate Limit (TEAL) for each of their largest sectors for which they provide liability insurance cover. Whilst firms are free to decide on the appropriate sector groupings, we expect these would generally follow some standardised groupings such as those provided by the traditional SIC groupings. Our preferred method is to use the Companies House Standard Industrial Classification (SIC) of Economic Activities (2007). For example, Manufacture of Electrical Equipment [Section C Division 27], Civil Engineering [F42], or Real Estate Activities [L68] would be acceptable sectors to use. We do 5 not expect sectors to be further sub-divided into more granular areas (for instance, Manufacture of wiring and wiring devices [C27.3]). We will accept other groupings that adhere to the same principles of grouping common exposures as per the linked SIC document. Firms using alternative groupings should explain how they have decided on these sectorial groupings. In establishing the groupings it is important that this is linked to the firm’s view of emerging risks – examples may be nanotechnology. 5 As we are targeting firms’ Casualty accounts, firms should not use ‘Motor Liability’ as a sector, unless this is the primary source of exposure from sectors such as Manufacture, Sale, Repair (etc.) of Motor Vehicles. 26 The aggregate gross loss should be calculated as: 1) 3% of TEAL for your top 2 sectors combined (as defined above); less 2) Current estimate ultimate claims relating to these policies (to avoid double counting) In deriving the net losses firms will need to consider the events that will cause the implied stress events and how their specific reinsurance will correspond. 7.3 REPORTING In addition to the standard scenario information, which will allow firms to feedback on the likelihood of this scenario, we require firms to provide information in relation to their top 3 sectors as defined by the 6 TEAL . Further, and in order for us to understand the way in which firms view their liability exposures, we require firms to state what proportion of their liability exposures are coded as ‘Miscellaneous’ (or ‘Other’). Firms are encouraged to provide descriptive information on the plausibility of an event that would result in the aggregate loss as determined using these assumptions set out in this stress test. Firms are required to provide the methodology and rationale for their selected groupings. 6 We acknowledge that for the purposes of the gross aggregate loss only the top 2 sectors are required. 27 8. SOLAR FLARE / GEOMAGNETIC STORMS 8.1 EVENT DEFINITION The purpose of this stress scenario is to assess the extent to which insurers have considered or can readily assess the impact of an earth directed solar storm that damages satellites and causes geomagnetic storms on earth that result in power outages & loss of certain communication systems. The scenario should also consider the potential vulnerability of space weather on the financial services sector. Elements of this sector’s operations depend on accurate timing (using space based technologies) which are space-weather-event-exposed. We acknowledge that the event description and the assumptions detailed below are broad, and as a result, we also expect a wide variation in the level of depth and information that firms provide, that will inevitably reflect the extent to which the firm has already considered a stress event of this nature. 8.2 ASSUMPTIONS Given the potential complexity of this scenario we have kept the definition relatively free-form and are relying on insurers to provide the details needed to translate this into a loss event. However, to ensure some consistency in the level of severity that insurers use we provide the following notes: Assume that the solar flare impacts multiple infrastructures (including telecommunications, transport and power grids) simultaneously in the US (along the Atlantic corridor between Washington DC and New York City) and the UK. Assume that a sufficient number of power transformers (i.e. those using alternating current power transmitting high voltage and with low loss) are knocked out such that power outage occurs and that power can only be restored once enough damaged equipment has been repaired or replaced. In this event the length of repairs or replacement should be at least 1 month and linked to particular regional areas in the US and the UK. It is at a firm’s discretion which area of the US and UK to select for outage purposes. In the UK, the main electric power plants are widely spread across England, Wales, Scotland, and Northern Ireland. However, we suggest that firms stress those regions where they have the highest exposures. In deciding which power grids are assumed to fail firms can use their own knowledge of the vulnerabilities / age of the power grids to which an insurance claim could occur. Where relevant firms should also consider the impacts on space-borne technology and the extent to which satellites in orbit can be damaged or the extent to which other covers could be impacted – such as satellite launches, and breakdown in communication systems with knockon consequences on the financial sector. The extent to which consideration and credit is taken for multilateral and international collaboration response to cope with this extreme event this should be set out in the spreadsheet under Management Actions. 28 9. CYBER LOSS 9.1 EVENT DEFINITION A series of simultaneous cyber-attacks are launched on large multinational organisations and discovered across the Retail sector (SIC code 47 – including all sub codes) with the intention of causing major disruption and financial loss to organisations. During the attacks, customer data (e.g. IP, credit card/bank details and/or other highly confidential information) is taken without authorisation. The attacks target vulnerabilities in the operating systems, web applications and/or software used by these organisations. For the purposes of this exercise it is assumed that multiple systems and/or multiple organisations using the same systems/software are affected. The hacking attacks are likely to take the form of a virus but equally could take an alternative vector of attack. Similarly for the purposes of this exercise it is assumed that multiple organisations across the world in the Retail sector come under attack at the same time (i.e. a macro event). As a result of the breach, customer management and trading systems, networks and supply chains are disrupted at these organisations for a minimum duration of 24 hours. The organisations affected have adopted reasonable network security processes, including anti-virus software and patching. 9.2 ASSUMPTIONS The assumption is that your fifteen largest clients (within the Retail sector – SIC code 47, including all sub codes) worldwide - based on exposure to policies (including, but not limited to, cyber liability), are targeted. Assume that all client data at these Retail organisations is lost (i.e. assume losses are a minimum of 90% of relevant policy limits for the top fifteen companies). Assume that class actions are pursued and you will face third party liability claims. For reinsurance purposes please calculate separately on the basis that these attacks are deemed both as one event and as fifteen separate events, returning whichever causes the largest net loss. Where an Electronic Data Exclusion or a Cyber Attack Exclusion clause has been consistently applied to a non-cyber specific (Other) policy, a zero gross loss incurred may be assumed for that line of business. If however such clauses have NOT been consistently applied across that line of business, a minimum 90% limit loss must be assumed. 29 9.3 REPORTING The estimated losses should consider and report the breakdown of the loss into the following lines of business: CYBER POLICY LOSSES o First party loss notification, associated costs and breach management costs, including crisis management o Business Interruption (excluding physical damage) o Contingent business interruption o Third party liability losses o Regulatory defence, legal fees and fines covered amounts o Other losses OTHER POLICY LOSSES o Crime o E&O policies with cyber endorsements o Technology E&O o D&O o GL / failure to supply o Other policies that may respond Firms are required to identify the cumulative gross loss incurred from the fifteen clients for each of the sections/heads of policy coverage above, as well as the anticipated reinsurance recoverable also split by the same policy section/head of coverage. 30 10. 1-IN-200 INSURANCE LOSS STRESS TEST (INSURER SPECIFIC) 10.1 EVENT DEFINITION This stress test is unique to each firm. As a result, no event definition is provided. Instead you are required to provide a scenario that corresponds to an estimated 1-in-200 year net aggregate 7 insurance loss over a one year period. We are aiming to gauge the types of events that are most relevant to your firm around this level of likelihood. Where the internal model is used as an input for this stress test, please provide commentary explaining how these events could arise and how these would impact your firm’s business. 10.2 ASSUMPTIONS You should report details of an extreme underwriting, reserving or a combined underwriting/reserving based scenario, not listed elsewhere in this exercise, which represents your net insurance losses with a probability of around 0.5% over the next 12 months. 10.3 REPORTING A detailed scenario description should be provided that corresponds to real world events. For instance it is not sufficient to state losses that will increase by £Xm under this event. Instead a description of the scenario is required – for instance under this scenario claims inflation is assumed to increase by 2% per annum above the best estimate; or claims deteriorate due to a major product recall by a pharmaceutical company after the discovery of harmful side-effects. In addition, please provide a breakdown of the increase in the gross and net technical provisions and the associated return period for the change in net technical provisions using the Solvency II lines of business classifications. For the avoidance of doubt the return period per line of business will not necessarily be 1-in-200; rather the net aggregate of all the losses will represent the 1-in-200 loss scenario. 7 (i.e. both underwriting and reserving) 31 11. REVERSE STRESS TEST (INSURER SPECIFIC) 11.1 EVENT DEFINITION Reverse stress tests are stress tests that require a firm to assess scenarios and circumstances that would render its business model unviable, thereby identifying potential business vulnerabilities. Reverse stress testing starts from an outcome of business failure and identifies circumstances where this might occur. This is different to general stress and scenario testing which tests for outcomes arising from changes in circumstances. Note: a firm's business model is described as being unviable at the point when crystallising risks, including loss of key staff or infrastructure or counterparty, cause the market to lose confidence in the firm. A consequence of this would be that counterparties and other stakeholders would be unwilling to transact with or provide capital to the firm and, where relevant, existing counterparties may seek to terminate their contracts. Such a point could be reached well before a firm's regulatory capital is exhausted. 11.2 ASSUMPTIONS For this scenario, you should use the event identified that is most likely to crystallise, drawing as 8 appropriate on any work already carried out as part of the SYSC 20.2 requirements. 11.3 REPORTING In addition to the standard stress test information we require completion of a number of questions specific to reverse testing that are aimed at enabling us to understand 8 your assessment of what constitutes business model failure; how this assessment feeds into your risk appetite and risk mitigation actions in place; the limitations of your assessment and sensitivities; and the governance arrangements you have in place. https://fshandbook.info/FS/html/handbook/SYSC/20/2 32 12. DEFINITIONS VARIABLE NAME DEFINITION SOURCES FOR HISTORICAL DATA UK real GDP Bank estimates of past gross domestic product (chained volume [1] measure, seasonally adjusted, £m). ONS and Bank calculations. UK nominal GDP Gross domestic product (current prices, seasonally adjusted, £m). [1] ONS code: YBHA. ONS. UK CPI UK Consumer price index (All items index, non-seasonally adjusted, indexed to 2014Q4 = 100). ONS code: D7BT. ONS. UK unemployment rate UK unemployment rate (LFS definition, all, aged 16 and over, seasonally adjusted, %). ONS code: MGSX. ONS. UK corporate profits UK corporate nominal profits (Nominal GDP minus pre-tax labour income, includes self-employment income, £m). Derived from ONS codes YBHA, ROYJ, ROYH and ROYK. ONS and Bank calculations. UK household income UK household nominal income (Nominal disposable income, adjusted for pensions contributions, £m). ONS code: RPQK. ONS. UK residential property price index Quarterly average of Halifax and Nationwide measures (indexed to 2014 Q4 = 100). Halifax, Nationwide and Bank calculations. UK commercial real estate price index aggregate Quarterly average of UK commercial property price index from the Investment Property Databank (the average of all UK properties in the IPD sample, indexed to 2014 Q4 = 100). Investment Property Databank (IPD UK) and Bank calculations. UK commercial real estate price index prime UK commercial real estate price index secondary UK equity prices Bank Rate Sterling IG corporate bond spread Sterling HY corporate bond spread IPD series for prime CRE (representing all properties in the lowest yielding quartile of its sample, indexed to 2014 Q4 = 100). Prime CRE is generally considered to refer to larger properties often located in London or other large cities, often with strong leases that create an investment similar to a bond. The IPD series for prime represents all properties in the lowest yielding quartile of its sample. IPD series for secondary CRE (representing all properties in the highest yielding quartile of its sample, indexed to 2014 Q4 = 100). Secondary property refers to all commercial real estate property that is not prime. Investment Property Databank (IPD UK) and Bank calculations. Investment Property Databank (IPD UK) and Bank calculations. Quarterly average of FTSE all-share price index (indexed to 2014 Q4 = 100). Thompson Reuters Datastream. Quarterly average of official Bank Rate (%). Bank of England. Quarterly average option adjusted spread over maturity-matched government spot curve on GBP denominated investment grade corporate and securitized debt publicly issued in the eurobond or UK domestic market (basis points). Quarterly average option adjusted spread over maturity-matched government spot curve on GBP denominated below investment grade corporate debt publicly issued in the eurobond or UK domestic market (basis points). BofA Merrill Lynch Global Research and Bank calculations. BofA Merrill Lynch Global Research and Bank calculations. PPP-weighted world real GDP PPP-weighted world real GDP, based on IMF World Economic Outlook data (indexed to 2014 Q4 = 100). IMF World Economic Outlook and [3] Bank calculations. Oil price Crude Oil-Brent Dated FOB USD per barrel. Thompson Reuters Datastream. 33 Volatility index Quarterly average of the VIX index. Thompson Reuters Datastream. GBP-EUR exchange rate index EUR/GBP exchange rate, defined as euros per pound (quarter-end, indexed to 2014Q4 = 100). Thompson Reuters Datastream, Bloomberg and Bank calculations. GBP-USD exchange rate index USD/GBP exchange rate, defined as US dollars per pound (quarterend, indexed to 2014Q4 = 100). Thompson Reuters Datastream, Bloomberg, and Bank calculations. G20 emerging economy dollar exchange rate index GDP-weighted basket of G20 emerging economy exchange rates vs the US dollar (quarter-end, indexed to 2014Q4 = 100). Thompson Reuters Datastream, Bloomberg, IMF WEO, and Bank calculations. Euro area excluding periphery Euro area countries excluding Cyprus, Greece, Ireland, Italy, Portugal and Spain. Real GDP Aggregated series of individual countries' real gross domestic product (seasonally adjusted, indexed 2014 Q4 = 100). Eurostat and Bank calculations. Consumer price inflation Estimated annual growth in euro area excluding periphery harmonised index of consumer prices (HICP). Eurostat and Bank calculations. Unemployment rate Estimated unemployment rate for euro area excluding periphery. Individual countries' unemployment rates are weighted by labour force. Eurostat and Bank calculations. Residential property price index Residential property prices (indexed to 2014 Q4 = 100). Individual countries' price indices have been weighted by nominal gross domestic product. OECD Housing Prices Database and Bank calculations. Commercial real estate price index Commercial property prices (indexed to 2014 Q4 = 100). Individual countries' price indices have been weighted by nominal gross domestic product. Investment Property Databank (IPD UK) and Bank calculations. Euro area periphery Cyprus, Greece, Ireland, Italy, Portugal and Spain. Real GDP Aggregated series of individual countries' real gross domestic product (seasonally adjusted, indexed 2014 Q4 = 100). Eurostat and Bank calculations. Consumer price inflation Estimated annual growth in euro area periphery harmonised index of consumer prices (HICP). Eurostat and Bank calculations. Unemployment rate Estimated unemployment rate for euro area periphery. Individual countries' unemployment rates are weighted by labour force. Eurostat and Bank calculations. Residential property price index Residential property prices (indexed to 2014 Q4 = 100). Individual countries' price indices have been weighted by nominal gross domestic product. OECD Housing Prices Database and Bank calculations. Commercial real estate price index Commercial property prices (indexed to 2014 Q4 = 100). Individual countries' price indices have been weighted by nominal gross domestic product. Investment Property Databank (IPD UK) and Bank calculations. US real GDP US real gross domestic product (seasonally adjusted, indexed to 2014 Q4 = 100). Thompson Reuters Datastream. US unemployment rate US unemployment rate (total unemployment, seasonally adjusted, %). Thompson Reuters Datastream. US equity prices Quarterly average of S&P 500 price index (indexed to 2014 Q4 = 100). Thompson Reuters Datastream. 34 US policy rate US Federal Funds Rate. Thompson Reuters Datastream. US residential property price index Case-Shiller house price index (seasonally adjusted, indexed 2014 Q4 = 100). Thompson Reuters Datastream. US commercial real estate price index US commercial property prices (indexed 2014 Q4 = 100). Federal Reserve Board. US dollar IG corporate bond spread US dollar HY corporate bond spread Quarterly average option adjusted spread over maturity-matched government spot curve on US dollar denominated investment grade corporate debt publicly issued in the US domestic market (basis points). Quarterly average option adjusted spread over maturity-matched government spot curve on US dollar denominated below investment grade corporate debt publicly issued in the US domestic market (basis points). BofA Merrill Lynch Global Research and Bank calculations. BofA Merrill Lynch Global Research and Bank calculations. UK government bond yields Zero coupon nominal government bond yields (%). Bloomberg and Bank calculations. Euro government bond yields Zero coupon nominal government bond yields (%). Bloomberg and Bank calculations. US government bond yields Zero coupon nominal government bond yields (%). Bloomberg and Bank calculations. 3 month sterling Libor 3m sterling Libor spot rates (%). Bloomberg and Bank calculations. 3 month Euribor 3m Euribor spot rates (%). Bloomberg and Bank calculations. 3 month dollar Libor 3m dollar Libor spot rates (%). Bloomberg and Bank calculations. Sterling Libor swap rates Euribor swap rates US dollar Libor swap rates Zero coupon interest rate swaps, derived from Libor fixings, Libor futures, Libor Forward Rate Agreements and Libor interest rate swaps (%). Zero coupon interest rate swaps, derived from Libor fixings, Libor futures, Libor Forward Rate Agreements and Libor interest rate swaps (%). Zero coupon interest rate swaps, derived from Libor fixings, Libor futures, Libor Forward Rate Agreements and Libor interest rate swaps (%). Bloomberg and Bank calculations. Bloomberg and Bank calculations. Bloomberg and Bank calculations. Footnotes [1] The data for real GDP reflects the likelihood of revisions to the data over the past, as embodied in the February Inflation Report. The data for nominal GDP does not. [2] The published series for lending to UK PNFCs may differ slightly from the series available on the Bank of England website. This is due to an updated methodology for calculating lending growth rates. [3] The historical data in this series is a simple quarterly interpolation of annual data, and is different to the historical data presented in Chart 1 in the Key Elements of the 2015 stress test publication. [4] These representative spreads are intended to provide notes to firms on the broad movement of euro area countries' bond yields relative to German nominal government bond yields. [5] The representative spread for the periphery excludes Greece whose bond yields have behaved differently to those of other euro area periphery countries in the recent past. 35 13. FURTHER NOTES FOR FIRMS WITH DEFINED PENSION SCHEMES – ONLY APPLICABLE FOR SCENARIO 5 Pensions Projections in the 2015 Stress Test may require additional variables that are not in the assumptions set out in scenario 5. In order to help firms assess proportionality, we set out below some additional assumptions to ensure greater consistency: a. Discount Rate Changes: We decompose the starting discount rate into ‘risk free’ and ‘credit spread’ components. We assume the ‘risk free’ curve is the gilt curve in the scenario (available up to 20 years). If extrapolation is required, we will assume the risk-free rate at 25 years is: Stress Scenario: calculated as the 20-year rate times the ratio of the 31 Dec 2014 25year and 20-year spot rates. Baseline Scenario: equal to the 20-year rate. For the ‘credit spread’ component, we use the methodology described below. b. RPI Inflation Changes: we assume that 40% of the change in the risk free rate (at each term) is passed on to a change in real yields on index-linked gilts. This gives a change in implied RPI, which is used to adjust the starting assumption. c. CPI vs RPI: We assume the same changes in term-related CPI expected inflation as the changes in the RPI inflation curve of the same term. d. UK and US Credit Spreads: Spreads are provided for Investment Grade and NonInvestment Grade Bonds. We assume the spreads on corporate bonds (including the IAS 9 discount rate) increase pro rata – i.e. by I(t) / I(0) where I(t) is the spread Index at time t. We assume that the change in credit spreads applies across the yield curve for that rating category of bond. When valuing corporate bond assets we ignore the impact of default and downgrades – i.e. the change in value can be based on the changing risk-free rate and credit spread. We do not value bonds individually and we expect firms to use representative bonds in each of the rating buckets. e. UK and US Equity assets: Equity assets are assumed to move with the equity returns provided in the scenario, ignoring basis risk. The index supplied is an equity price index only, i.e. does not include dividends. The gross dividend yield at time zero on the relevant index is assumed to remain constant following the stress. f. Euro, Asian and other overseas assets: For material holdings, movements in equity prices or credit spreads can be estimated using a combination of: UK and US equity paths A comparison of the overseas GDP paths in the scenario with UK / US GDP paths. The one-year stresses provided in the “Traded Risk Scenario” document. Where there is no appropriate GDP index, or the holding is not material, the remaining assets are assumed to be invested in the relevant UK index and to be denominated in sterling. g. Property assets: The property index is assumed not to include rental income. A rental yield of 5% is assumed to be applicable throughout the projection. h. Alternative Assets: Hedge funds, private equity and other alternative assets without an index are assumed to be invested in equities. 9 The starting credit spread plus the starting risk free rate used in bullet point (a) should add up to the discount rate at 31 December 2014. 36 We understand that firms may have their own methodology in place, and may use their own consistent assumptions. In order to help our analysis, we expect any material differences in firms’ methodology to be explained and justified. PRA 20 Moorgate London EC2R 6DA 37