Session 19 - Pension Crisis: Lessons and Implications for Risk Management (Practice Area: Investment/Pension) Thursday, June 21, 2012, 14:30 16:00

Transcription

1 Session 19 - Pension Crisis: Lessons and Implications for Risk Management (Practice Area: Investment/Pension) Thursday, June 21, 2012, 14:30 16:00 Moderator: Ross Dunlop, Cardinal Capital Management Speaker: Vincent Morin, Air Canada Pension Plan

2 HOW LOW CAN RATES GO? Japan 1995 Canada 2010 Japan 2000 Japan 2010 Source: Bloomberg 2

3 HOW LOW CAN RATES GO? Source: Bloomberg 3

4 HOW LOW CAN RATES GO? Unseen territory Negative real yields in many parts of the world Negative nominal yields in Switzerland Source: Bloomberg 4

5 HOW LOW CAN RATES GO? The risk-free asset pool (trillions of US$) US Treasuries US Agencies MBS/ABS UK gilts Germany+France government debt Italy+Spain government debt Japan government debt Risk-free assets available Less debt held by central banks Less international reserves Risk-free available to investors The pool of risk-free assets available to investors has almost decreased by half over the last 4 years Canada is believed to be a healthy country. Drives demand and prices up for Canadian bonds Sources: JPMAM, Federal Reserve, UK DMO, Euro area finance ministries and central banks, Japan MoF, BoJ, IMF 5

6 HOW LOW CAN RATES GO? Canadian Bond Market $1,200B of outstanding issues (DEX Universe index) $340B are Long-Term issues (>10 years) $260B are LT Governments issues $87B are LT Government of Canada issues (small!) Nominal Canadian Bond Allocation of 147 Large Plans (2010 PIAC Survey) is $260B 22% of all issues, assuming they are all invested in the DEX Universe index Real Return Bond Allocation of $45B Approx. 80% of the total Real Return Bond issues! Actual float is small (i.e. amount of bonds actually traded on the market) Insurance companies, mutual funds, ETFs, DC plans all are very large relatively passive holders of LT government bonds More than $1,250B of liabilities valued on small portion of $87B of Canada GVT bonds yields 6

7 IMPACT FOR PENSION PLANS If 50% of pension funds switch from DEX Universe to DEX GVT LT, it represents additional LT GVT Bond Purchase of >$70B, or 28% of total LT GVT Bond Issues If 25% of funds add Bond Overlay in DEX GVT LT for 20% of their AUM, it represents additional LT GVT Bond Purchase of >$50B, or 19% of total LT GVT Bond Issues Considering that, worldwide, investors are looking for risk-free assets, imbalance in supply/demand is likely to drive yields lower in times of market stress Doubles the stress for defined benefit pension plan sponsors need to cope with potential recession (less revenues) + pension deficits (higher contributions) 7

8 IMPACT FOR PENSION PLANS Solvency pension liabilities increased to unsustainable levels Annuity Proxy of 3.3% as of Risky assets still under the shadow of the 2008 credit crunch, with systemic risk priced in However risk premiums are at historical highs for many asset classes Risk premiums can be reflected in going concern discount rates, but not for solvency 8

9 IMPACT FOR PENSION PLANS Since 2008, markets react as if there is a binary potential outcome Scenario A - Catastrophic The end of the Euro Banking system fails, with worldwide contagion Series of sovereign defaults, with contagion to the US The great reset Low probability, but large damage Scenario B Most likely Europe solve its problems (not an easy task) and gets back on low growth Banking system recapitalized Small GDP growth (diminished by fiscal austerity) Slight increase in yields Hopefully, Scenario B most likely to happen No plan sponsor can manage a plan as if Scenario A happens Yields driven by likelihood of Scenario A, as are Solvency valuations 9

10 RISK MANAGEMENT AT AIR CANADA Focus on solvency deficit, but managed as a going-concern Approach changed to Liability Driven Investments in 2009 Goal to reduce Surplus Risk, to mitigate volatility of past service contributions LDI does not mean immunization the plan is a going-concern and will still be invested in risky assets Investment strategy strongly linked to evolution of LT yields main driver of volatility of contributions Significant use of derivatives and leverage 10

11 RISK MANAGEMENT AT AIR CANADA Expected VAR 95 1-Year Return Asset Only Surplus/Deficit Traditional approach 7% -9% -15% Immunized approach 4% -7% -3% LDI approach 7% -10% -7% Value to optimize based on traditional approach Value to optimize based on LDI approach 11

12 RISK MANAGEMENT AT AIR CANADA Investment team needs strong understanding of 1. dynamics of actuarial liabilities, and 2. good understanding of market dynamics (especially bonds) Two general set of responsibilities regarding investment decisions 1. Invest assets of Canadian pension plans 2. Propose changes to long-term asset mix policy Various hurdles Communication (LDI strategy may lag in bull markets) Governance (need delegation to manage derivatives) Actuary must understand strategy to set discount rate appropriately Need expertise 12

13 RISK MANAGEMENT APPROACHES Limited Many pension plans had none Others, limited to efficient frontiers Traditional Asset only Separate department (police) Hard-preset VaR guidelines But did not work Emphasis on recent past data PM arbitraged risk measures Air Canada Pension Plan Emphasis on pension liabilities and total fund; no silos Integrated to portfolio management, thus part of the investment team Used in every single investment decision 13

14 IMPORTANT RISK MEASURES RISK REPORT as of Dec Annualized Standard Deviation Monthly 95% VaR SURPLUS RISK Benchmark portfolio 8.73% -4.81% Current portfolio (all risks) 9.41% -5.05% ACTIVE RISK Total 1.64% -0.67% Internal mandates (1) 1.29% -0.50% External managers (2) 1.01% -0.46% Additional risk management : Traditional min/max weights Limits on notional Counterparty risk Operational risk Nominal duration Credit duration CASH-FLOW RISK -1.57% (1) Portable alpha, credit, asset mix deviations, etc. (2) Excludes portable alpha 14

15 FOOD FOR THOUGHTS Actuarial valuation assumptions based on belief that government bonds are risk-free assets This may not be the case anymore What is more risky in the current economic context? A. A very conservative and well diversified portfolio of assets (for example, corporate bonds and limited use of alternative investments, such as high yield bonds and real estate) B. A portfolio 100% invested in a single Sovereign Government debt 15

16 FOOD FOR THOUGHTS Is solvency valuation efficient in making sure pension plans are fully funded in case of termination? Plans terminated in recent years were in deficit Burden made employers switch to DC plans Allowing the use of a low risk portfolio rather than annuity purchase proxy, as the basis for valuation would help reduce the liability PBSA defines solvency liabilities as the liabilities of a plan that relate to defined benefit provisions and are determined on the basis that the plan is terminated. Actuarial assumptions must reflect that fact, thus use the annuity proxy If terminated would be changed to... have a strong likelihood to have sufficient assets to pay for all future benefits in the law, could result in a higher discount rate assumption i.e. around 5%, using a conservative diversified portfolio under current market conditions Strong likelihood could be defined such as having 80% chance of being solvent in 3 to 5 years (assuming the plan s solvency ratio was 100% at inception) 16

17 FOOD FOR THOUGHTS What serves a pension plan member best? A. Requiring a plan sponsor to pay past service contributions following a solvency valuation, knowing it will trigger the actual termination of the plan, with large deficit B. Requiring funding of past service on a going concern basis, as the plan sponsor is expected to remain solvent, should the plan assets generate returns in line with going concern assumption, but with the risk the plan sponsor could at some point in the future become insolvent, triggering the actual termination of the plan, with large deficit Over the short term, having benefits as secured as possible as prescribed by law, forcing plan sponsors to inject significant amount of money in pension funds, may be desirable for plan participants, if it does not lead the plan sponsor to bankruptcy Over the long term, side effect is decreasing number of employees having DB Plans burden of an increasing number of Canadians not having sufficient assets at retirement will ultimately fall on the governments Need to re-establish the fairness between DB and DC plans, by helping plan sponsors in funding DB schemes, at reasonable risk and cost 17