INTEGRATIVE INVESTMENT MANAGEMENT. and THE COST OF PENSION FUND CONSTRAINTS. KEY FORUM INSIGHTS AND POSSIBLE ACTION STEPS June 2008

Transcription

1 INTEGRATIVE INVESTMENT MANAGEMENT and THE COST OF PENSION FUND CONSTRAINTS KEY FORUM INSIGHTS AND POSSIBLE ACTION STEPS June 2008 A Discussion Forum Sponsored by Rotman International Centre for Pension Management Rotman School of Management University of Toronto Held June 3-4, 2008 Toronto, Canada Prepared by Keith Ambachtsheer Adjunct Professor of Finance Director, Rotman International Centre for Pension Management Rotman School of Management, University of Toronto

2 NEW PERSPECTIVES ON INSTITUTIONAL INVESTING DISCUSSION FORUM AGENDA JUNE 3-4, 2008 Rotman School of Management, University of Toronto 105 St. George Street, Room 133 Tuesday, June 3 (Day 1) 8:30-8:35 Welcome and Forum Overview Keith Ambachtsheer, Rotman ICPM and Ann Henhoeffer, Rotman ICPM (Canada) 8:35-9:40 The Cost of Active Investing Kenneth French, Dartmouth College (USA), and Past President of the American Finance Association (USA) Prof. French will deliver an ICPM-tailored version of his recent Presidential Address to the American Finance Association. We have budgeted time for post-address discussion. 9:40-10:30 Delegated Decision-Making in Institutional Investing Michael Brandt, Duke University (USA) This will be the first presentation of the results of this important ICPM-sponsored research project (with Jules van Binsbergen and Ralph Koijen) 11:00-Noon 1:00-1:45pm 1:45-2:30pm 2:45-3:30pm 3:30-4:15pm 4:15-5:00pm 5:30pm onward Delegated Decision-Making in Institutional Investing Discussion Jack Gray, Sydney University of Technology (Australia), Brett Hammond, TIAA-CREF (USA), and Rob Bauer, University of Maastricht (The Netherlands). Jack, Brett, and Rob will lead off as discussants of the Brandt et al. paper. General forum participant discussion on the practical applicability of the paper will follow. Stock Market Performance and Pension Fund Investment Policy: Rebalancing, Free Float, or Market- Timing? Dirk Broeders, De Nederlandsche Bank (The Netherlands) Dirk will present the findings of a study (with Jacob Bikker and Jan de Dreu) geared to answer the above question. The study findings should spark some lively reactions. Measuring and Managing Return, Risk, and Incentive Compensation at CPP Investment Board John Ilkiw, Don Raymond, CPP Investment Board (Canada) John and Don will recount the evolution of the current CPPIB investment decision-making structure and the challenges that still lie ahead. Measuring and Managing Return, Risk, and Incentive Compensation at APG Investments Stefan Lundbergh, Onno Steenbeek, APG Investments (The Netherlands) Stefan and Onno will recount the evolution of the current APG Investment decision-making structure and the challenges that still lie ahead. Measuring and Managing Return, Risk, and Incentive Compensation at TD Bank Alan Jette, TD Bank (Canada) What can pension funds learn from other financial institutions? As the key risk manager at one of North America s best-run commercial banks, Alan brings a unique perspective to this question. What Were the Day s Most Important Insights? Group Discussion led by Keith Ambachtsheer Group Discussion is a powerful way to distill forum participants best thinking on what the most important insights of the day were for them. As importantly, are there things that we can do collectively (e.g., through ICPM) that could produce better outcomes? Cocktails and Networking Dinner ICPM Rotman International Centre for Pension Management Page 2 of 17

3 NEW PERSPECTIVES ON INSTITUTIONAL INVESTING DISCUSSION FORUM AGENDA JUNE 3-4, 2008 Rotman School of Management, University of Toronto 105 St. George Street, Room 133 Wednesday, June 4 (Day 2) 8:00-8:30 Continental Breakfast 8:30-8:45 Restrictions on Pension Investing: Why Is This An Important Topic? Keith Ambachtsheer, Rotman ICPM (Canada) 8:45-9:15 Pension Investing and Regulation: A Legal Perspective Poonam Puri, Osgoode Law School, York University (Canada) Prof. Puri will bring us up to date on the changing regulatory environment in which pension funds have to operate. 9:15-10:00 Restrictions on Pension Investing: A Cost Assessment Philip Davis, Brunel University (UK) Prof. Davis has conducted research on this question for many years, and will give us an assessment of what we know today about the cost (and the benefits?) of various types of investment restrictions, and on what can be done to reduce those costs. 10:00-10:20 Restrictions on Pension Investing: A Canadian Perspective Michael Nobrega, Ontario Municipal Employees Retirement System (Canada) 10:20-10:40 Break 10:40-11:00 Restrictions on Pension Investing: An Australian Perspective Leo de Bever, Victorian Funds Management Corporation (Australia) 11:00-11:20 Restrictions on Pension Investing: A Swedish Fund Perspective Tomas Franzén, AP2 Fund (Sweden) 11:20-11:40 Restrictions on Pension Investing: A US Fund Perspective Joe Dear, Washington State Investment Board (USA) 11:40-Noon Restrictions on Pension Investing: Optimal Collaboration Strategies Participant discussion session led by Keith Ambachtsheer Once again, this final morning session will focus on brainstorming on what could be done collectively (e.g., through ICPM) to produce better outcomes. 12:15-1:30 Celebrating Our Third Birthday: Where Does ICPM Go From Here? Lunch for all attendees with the newly-expanded ICPM Board of Directors at Royal Canadian Yacht Club located at 141 St. George Street (main floor dining room). ICPM Rotman International Centre for Pension Management Page 3 of 17

4 Rotman ICPM June 2008 Discussion Forum Attendees List Douglas Stratton Alberta Finance - AIM Canada douglas.stratton@aimco.alberta.ca Ron Liteplo Alberta Local Authorities Pension Plan Corp Canada ron.liteplo@gov.ab.ca Meryl Whittaker Alberta Local Authorities Pension Plan Corp Canada meryl.whittaker@gov.ab.ca Tomas Franzén AP 2 Sweden tomas.franzen@ap2.se Onno Steenbeek APG Finance The Netherlands onno.steenbeek@apginvestments.nl Stefan Lundbergh APG Investments The Netherlands stefan.lundbergh@apginvestments.nl Chresten Dengsoe ATP Denmark cd@atp.dk Michael Preisel ATP Denmark mip@atp.dk Paul Costello Australia Future Fund Australia paul.costello@futurefund.gov.au Katrina Ellis Australian Prudential Regulation Authority Australia katrina.ellis@apra.gov.au Paul Gauthier BIMCOR Inc. Canada gauthier.p@bimcor.ca Philip Davis Brunel University United Kingdom philip.davis@brunel.ac.uk Richard Guay Caisse de dépôt et placement du Québec Canada rguay@lacaisse.com Jean-François L'Her Caisse de dépôt et placement du Québec Canada jlher@cdpcapital.com Patrick Savaria Caisse de dépôt et placement du Québec Canada psavaria@lacaisse.com Terrie Miller CEM Benchmarking Inc. Canada terrie@cembenchmarking.com Tom Scheibelhut CEM Benchmarking Inc. Canada tom@cembenchmarking.com Jan Fokkens Cordares The Netherlands j.fokkens@cordares.nl Alwin Oerlemans Cordares The Netherlands a.oerlemans@cordares.nl Stephen Burnie CPP Investment Board Canada sburnie@cppib.ca John Ilkiw CPP Investment Board Canada jilkiw@cppib.ca Steven James CPP Investment Board Canada sjames@cppib.ca Don Raymond CPP Investment Board Canada draymond@cppib.ca Ken French Dartmouth College USA kenneth.r.french@tuck.dartmouth.edu Dirk Broeders De Nederlandsche Bank The Netherlands d.w.g.a.broeders@dnb.nl Sandra Wesseling De Nederlandsche Bank The Netherlands a.a.t.wesseling@dnb.nl Michael Brandt Duke University USA mbrandt@duke.edu Tony Montag Georgia State University USA tony.montag@amontag.com Virginia Atkin KPA Advisory Services Canada virginia@kpa-advisory.com Rob Bauer Maastricht University The Netherlands r.bauer@finance.unimaas.nl Patrick Janssen MN Services The Netherlands patrick.janssen@mn-services.nl Tim Mitchell New Zealand Superannuation Fund New Zealand tmitchell@nzsuperfund.co.nz Sadayuki Horie Nomura Research Institute Japan s-horie@nri.co.jp John Grace Office of the Superintendent of Financial Institutions Canada john.grace@osfi-bsif.gc.ca Rob Brown Ontario Expert Commission on Pensions Canada rlbrown@uwaterloo.ca Athos Sani Ontario Ministry of Finance Canada athos.sani@ontario.ca Bruce MacNaughton Ontario Ministry of Finance Canada bruce.macnaughton@ontario.ca Irene Barone Ontario Municipal Employees Retirement System Canada lbarone@omers.com Ana Caçoilo Ontario Municipal Employees Retirement System Canada acacoilo@omers.com Philip Haggerty Ontario Municipal Employees Retirement System Canada phaggerty@omers.com Michael Nobrega Ontario Municipal Employees Retirement System Canada mnobrega@omers.com Graham Pugh Ontario Municipal Employees Retirement System Canada gpugh@omers.com Barbara Zvan Ontario Teachers' Pension Plan Board Canada barbara_zvan@otpp.com Robert Bertram Ontario Teachers' Pension Plan Board Canada robert_bertram@otpp.com Andrew Spence Ontario Teachers' Pension Plan Board Canada andrew_spence@otpp.com Poonam Puri Osgoode Hall Law School Canada ppuri@osgoode.yorku.ca Jeroen Tielman PensionSummit The Netherlands jeroen.tielman@pensionsummit.com Jaap van Dam PGGM The Netherlands jaap.van.dam@pggm.nl Rene van de Kieft PGGM The Netherlands rene.v.d.kieft@pggm.nl Don Guy Pollara Canada dguy@pollara.com Keith Ambachtsheer Rotman ICPM Canada icpm@rotman.utoronto.ca Ann Henhoeffer Rotman ICPM Canada icpm@rotman.utoronto.ca ICPM Rotman International Centre for Pension Management Page 4 of 17

5 Feng Chi Rotman School of Management Canada Siu Kai Choy Rotman School of Management Canada Alexander Dyck Rotman School of Management Canada Paul Halpern Rotman School of Management Canada Ali Emre Konukoglu Rotman School of Management Canada Fatma Sonmez-Saryal Rotman School of Management Canada Xiaolu Wang Rotman School of Management Canada Jun Zhou Rotman School of Management Canada Sandy Matheson State of Washington USA David Hartley SunSuper Australia Jack Gray Sydney University of Technology Australia Alan Jette TD Bank Canada Douglas Fore TIAA-CREF Investment Management USA Brett Hammond TIAA-CREF Investment Management USA David Blitzstein United Food & Commercial Workers - International Union USA dblitzstein@ufcw.org Peter Moon Universities Superannuation Scheme Limited United Kingdom pmoon@uss.co.uk David Russell Universities Superannuation Scheme Limited United Kingdom drussell@uss.co.uk Karen Coll University of Toronto Asset Management Corp. Canada karen.coll@utam.utoronto.ca Bill Moriarty University of Toronto Asset Management Corp. Canada bill.moriarty@utam.utoronto.ca Leo de Bever Victorian Funds Management Corporation Australia ldebever@vfmc.vic.gov.au Joseph Dear Washington State Investment Board USA jdear@sib.wa.gov Stephen Lerch Washington State Investment Board USA stevel@dor.wa.gov Theresa Whitmarsh Washington State Investment Board USA twhitmarsh@sib.wa.gov Cynthia Williams York University Canada cwilliam@law.uiuc.edu ICPM Rotman International Centre for Pension Management Page 5 of 17

6 Post-Forum Survey Feedback Attendees were happy with the Forum: out of 34 responses, there were 25 excellent ratings, 9 good ratings, and zero poor ratings. Out of a range of 1.0 (poor) to 2.0 (good) to 3.0 (excellent), the average ranking for the 12 individual sessions was 2.3. All 12 individual sessions had an average ranking of at least 2.0, with one session actually achieving an average 3.0. Some specific comments: I think everything went well. Good speakers. Ample time for discussion. the reality was even better than what was promised in the agenda well-executed Excellent topics, excellent participation and discussion, good mix of people. Good blend of theory and practice The plenary discussion format worked well.better than small-group breakouts followed by report-backs Maybe the program was a bit too full? There were points where there didn t seem to be enough time.maybe plan for longer breaks and/or less content. The other major question in the survey related to future Forum topics. Suggestions included: Pension Design Study of national retirement income system design Managing longevity risk Managing inflation risk Movement towards individual pension accounts/dc-based systems Regulatory Issues The solvency measurement debate Appropriate disclosure/accounting rules Regulatory oversight in individual pension accounts/dc-based systems Organization Design & Management Organizing the risk management function at the total fund/balance sheet/enterprise level Benchmarking/performance measurement (especially in the private markets area) Collaboration strategies among funds (especially in the private markets area) In- vs. out-sourcing Investment Beliefs & Strategies Market dysfunctionality Infrastructure investing The SRI question GOOD NEWS! Many of these topics will be covered in the Fall 2008 Discussion Forum! See the preliminary October 2008 Discussion Forum Agenda at the end of this report. ICPM Rotman International Centre for Pension Management Page 6 of 17

7 Theme #1: Integrative Investment Management June 3, 2008 This session had five modules: 1. A study by Ken French measuring the cost of equity investing in the USA. 2. An ICPM-funded study by Michael Brandt, Ralph Koijen, and Jules van Binsbergen, laying out a normative theory of delegated investment decision-making, followed by comments from Jack Gray, Brett Hammond, and Rob Bauer. 3. A study by Dirk Broeders, Jacob Bikker, and Jan de Dreu on the relationship between the asset mixes of Dutch pension funds and equity markets behavior. 4. Three cases on delegated decision-making in action at CPP Investment Board (Don Raymond and John Ilkiw), APG Investments (Onno Steenbeek and Stefan Lundbergh), and TD Bank (Alan Jette). 5. Key insights and possible action steps based on general participant discussion. The Cost of Active Investing presented by Ken French addressed two questions: 1. How much does active investing under-perform passive investing? 2. How much does society pay for price discovery? He answered the questions by adding up the fees and commissions paid by individuals directly, through retail funds (e.g., mutual funds, ETFs), wholesale funds (e.g., pension funds, endowment funds), and hedge funds. Key findings: Since 1980, the proportion of US equities held directly by individuals halved from 50% to 25%, while the proportion held through mutual funds increased from 5% to 35% (from some $50B to almost $5T). Over the same period, open-ended mutual fund fees halved from 200bps to 100bps because of a material decline in sales costs (loads). Expressed relative to the market capitalization of total US equities, the fees in this sector increased from 10bps in 1980 to 25bps in They have stayed at that level the last 10 years. The active fees in DB funds have fluctuated between 35bps and 45 bps, while passive fees declined from 8bps to 3bps. At the same time, the proportion passively invested increased significantly (e.g., to 50% of all US equity holdings in public funds). Expressed relative to market capitalization of total US equities, the fees in the entire wholesale funds sector have halved from 17 bps to 9bps. Exposure through hedge funds has gone from a few $billion in 1990, to over $500B in 2007 (almost half through fund-of-funds). Management fees in 2007 amounted to 450bps (125bps base fees plus 325bps performance fees). The total was 650bps for fund-of-funds component. Expressed relative to total US equities market capitalization, total 2007 hedge fund fees amounted to 15bps. Expressed relative to total US equities market capitalization, trading costs have declined from 60bps in 1980 to 20bps in 2006, mainly due to a dramatic fall in trading commission rates. ICPM Rotman International Centre for Pension Management Page 7 of 17

8 Overall costs (fees plus commissions) have fluctuated around 75bps of total market capitalization, with rising costs due to increased mutual and hedge fund exposures offset by falling trading costs and pension/endowment fund management fees. The cost of active investing relative to passive investing increased over the observation period (to an incremental 67bps or $100B in 2006). In conclusion, Prof. French asked: are these findings evidence on the part of active investors of (a) confusion?, (b) overconfidence? or (c) superior skill? (Keith Ambachtsheer s own answer to this question is: lots of (a) and (b), and some (c) ). Optimal Decentralized ALM presented by Michael Brandt examined the situation where financial institutions decentralize investment decision-making in order to take advantage of specialized external investment skills. However, this strategy can lead to suboptimal diversification, confused risk preferences, and too much mismatch risk vs. the institution s liabilities. The paper addressed three key questions: 1. What is the cost of decentralization? 2. Is there an optimal benchmarking strategy that minimizes that cost? 3. How can the suggested approach be operationalized in practice? The context is a pension fund with bond-like liabilities, seven available asset classes, and two available active strategies (one in equities and one in fixed income). The pension fund expresses its reward/risk preferences in A/L terms. The two active managers express their reward/risk preferences in A1/B1 and A2/B2 space, where B1 and B2 are equity and fixed income market benchmarks respectively. The cost of decentralization (i.e., everybody tries to maximize their own welfare, independent of anyone else) can be expressed as: Cost Suboptimal = Diversification + A/L Mismatch Risk-Alpha Conclusions included: Under reasonable assumptions, the implication is that the active strategies must have very high Information Ratios (e.g., in excess of 1.0) in order for the decentralized decision structure including active management to be superior to a passive approach using the seven available asset classes. The paper then goes on to show how optimal benchmarks can be constructed for the active managers, which lead to the same outcome as if the investment process was centralized at the pension fund level. These benchmarks result in jointly attaining the full benefits of diversification, and simultaneously achieving the optimal alpha overlay/assetliability mismatch risk combination. An important model implementation challenge is to get the right numbers (orders of magnitude?) for the assumptions required to drive the model. Two key required assumptions are (1) the risk tolerances of the active managers, and (2) their predictive abilities (i.e., information coefficients). There are two estimation approaches are: (1) go directly to best estimates, or (2) develop best estimates indirectly through the study of relevant active management return databases. ICPM Rotman International Centre for Pension Management Page 8 of 17

9 The paper discussants Jack Gray, Brett Hammond, and Rob Bauer made the following points: Quantifying the potential cost of uncoordinated decentralized asset management is an important research contribution to pension fund management, as is recognizing to potentially conflicting interests of pension plan participants (good risk-adjusted returns) and external asset managers (growing assets under management). The paper redirects the attention of the internal pension fund CIO function towards understanding the preference functions of external managers (e.g., their culture, incentives, behaviors, failures, and organization structures). A key result of the paper is that external managers must manage to liability-tailored benchmarks. The practical question of getting external managers to accept this reality remains as a very real obstacle. Stock Market Performance and Pension Fund Investment Policy: Rebalancing, Free Float, or Market Timing? was presented by Dirk Broeders. Based on eight years of information on 748 funds available to it, De Nederlansche Bank noted significant variation in both the actual asset mixes of Dutch pension funds, and in the indicated target strategic mixes. The goal of the study was to try to infer the possible causes for these fluctuations by seeing how they related to movements in the equity and bond markets themselves. They found that relative stock market performance influences the asset allocation of pension funds in a number of ways: In the short term, actual allocation changes were driven by a combination of free floating (passive management) and market timing (active management, with large funds being most active). Only 39% of excess returns are rebalanced (12% in up-markets, 49% in down-markets). In the medium term, stock market performance seems to lead to changes in strategic asset allocation (1% relative outperformance leads to 0.01% increase in strategic equity allocation, on average). On average, changes in asset allocation over time have not generated additional returns. Measuring and Managing Return, Risk, and Incentive Compensation at CPP Investment Board was presented jointly by Don Raymond and John Ilkiw. The presentation had three parts: the Early Years ( ), becoming an Active Global Investor (2006-now), and the Road Ahead: At first, the decision to start by investing passively and the Canadian pension fund 30% limit on foreign investing (Foreign Property Rule) imposed significant constraints. In preparation for the removal of these constraints, performance measures which adjusted returns for cost and risk relative to a minimum risk portfolio (now called Net Liability Mimicking Portfolio or NLMP) were considered (e.g., the RANVA metric). Conceptually, preparing for active management reduces to (a) accepting a necessary amount of mismatch risk vs. the NLMP through building an implementable, low-cost, low-complexity Reference Portfolio with a target net real return of 4.2%, and (b) building an arsenal of value-adding strategies (i.e., better betas, unique alphas, taking advantage of CPPIB s unique situation), and (c) implementing those strategies by taking positions that differ from those of the Reference Portfolio (i.e., active risk ). Organizationally, the decision-making process flows from the BofD (Reference Portfolio and active risk budget), to the Investment Planning Committee (IPC) and Portfolio Design and Risk Management (PDRM) ( better betas ), to the investment teams (Private ICPM Rotman International Centre for Pension Management Page 9 of 17

10 Real Estate, Public Markets, Private Investments) ( alpha ). The entire process is supported by an internal Corporate Legal/Financial/IT group. Explicit risk budgeting process promotes accountability, transparency, comparability, and efficiency. The incentive compensation framework is based on a series of principles (e.g., competitive, aligned, etc.), choices (e.g., benchmarks, hurdles, caps/floors, time frame, etc.), and applications (e.g., total fund, strategies, groups, portfolios, etc.). Compensation framework related to doing a $400M private equity buyout as an example, and to oversight by the Board HRC Committee. Road Ahead issues (current assets about $125B): refining the Reference Portfolio, improving ALM modeling, new decision metrics, explaining the continuously evolving process to stakeholders. Measuring and Managing Risk, Return, and Incentive Compensation at APG was presented by Onno Steenbeek and Stefan Lundbergh. Once the merger with Cordares occurs in late 2008, APG will be a pension services provider for 3.9 million people, have assets under management of 244B, and will have 3,500 employees. APG operates under the supervision of DNB, which requires a high funded ratio with a high level of probability. Key risk management tools include conditional pension indexation (i.e., no indexation below 100% funded ratio, full indexation above 140% funded ratio), strong diversification, liability-related duration management, and a new LHP-ROP portfolio management structure (Liability-Hedging Portfolio and Risk- Optimizing Portfolio). Specific points made included: Significant evolution in asset mix to non-traditional investments (e.g., indexed loans, infrastructure, private equity, commodities, hedge funds, new ideas/innovation). Dutch funds are challenged with dual LHP anchoring (i.e., regulator focus: 1-year nominal; client focus: longer term, inflation-adjusted). There is continuous evolution of the package of financial services offered towards individualization/personalization. Building a competitive ROP is an increasingly challenging proposition. Dutch culture constrains the ability to provide highly competitive financial remuneration packages. So creating an innovating, exciting environment within which to work is especially important. The 2% allocation (i.e., almost 5B) to New Ideas/Innovation is especially important in this regard. A stepwise, inviting decision-making process has been set up for evaluating New Ideas/Innovation proposals and (if accepted) implementing them. All of APG s 450+ investment professionals are invited to participate. Examples were provided in the timberland, music rights, and landfills areas. Fortunately for APG, the following equation is working: Complex + Challenging + Difficult = Interesting ICPM Rotman International Centre for Pension Management Page 10 of 17

11 Creating Shareholder Value was presented by Alan Jette, TD Bank s SVP Treasury & Balance Sheet Management. The reason for inviting him was to provide the Forum participants with the perspective of a non-pensions person on the question of risk measurement, management, and compensation. Listening to the risk manager of a major bank which managed to avoid the subprime debacle is especially timely: The goal at TD is to create shareholder value which it defines as {net income-after-tax minus a risk-based capital charge}. TD does this through the ruthless, no-exceptions, transparent application of its capital allocation model with four components: (1) funds transfer pricing (fully hedged, optionality-adjusted, liquidity premium-adjusted), (2) expected credit loss assessments (using full credit cycle), (3) risk-based capital allocations (including tail risk component, and (4) clear differentiation between undertaking productive risks (e.g., credit) and unproductive risks (e.g., equity market). This discipline keeps TD away from carry-trade strategies (interest rate or credit-based) and structured products with embedded tail risks. Banks without this kind of discipline did get into carry-trade and structured product strategies, with serious financial consequences for their shareholders, and for the global economy. Both regulatory capital rules and internal incentive compensation schemes created incentives to engage in carry-trade and structured product strategies. In the end, success requires two things: (1) getting the economics of the business right, and (2) building and fostering an internal culture to stay the course. What Did We Learn Today? was a group discussion around two questions: (1) what where the AHA! moments for you today? and (2) What are the possible related action steps for when you get back to the office? Comments included: Understanding a pension fund s total cost structure is very helpful. We need to think carefully about whether we can cost-justify the money we are spending, especially in high-cost areas such as private equity and hedge funds. Delegated decision-making structures are naturally high-cost structures because of a number of disconnects between various inside and outside agents. Only integrated decision structures can successfully deal with these agency costs. Many attendees liked the New Ideas/Innovation Investing concept, and we need to think about how to apply it in our organization. Many attendees liked the ideas presented around creating a thoughtful incentive compensation structure, especially building the right Reference Portfolio and pricing embedded options as part of the cost of capital hurdle rate used to trigger incentive compensation. Again, we need to think about how to apply these ideas in our organization. The discussions on risk management today have made me wonder whether our current risk management methods could really withstand close scrutiny. We need to look into this. Maybe we need to change the standard pension investment industry question from What would Dave do? (i.e., David Swenson at the Yale Investment Office) to... What would Alan do? (i.e., Alan Jette at TD Bank). ICPM Rotman International Centre for Pension Management Page 11 of 17

12 Theme #2: Investment Constraints June 4, 2008 This session had two set-up presentations, one on the legal perspective by Poonam Puri and the other on the cost perspective by Philip Davis. Four mini-case studies followed on Canada (Michael Nobrega), Australia (Leo de Bever), Sweden (Tomas Franzén), and the USA (Joe Dear). The session finished with group discussion on the questions of (1) what further research is required to understand the investment constraints issue in the fullest sense, and (2) what are possible collaboration strategies pension funds can employ to deal with removing investment constraints that are costly, and serve no clear purpose. Legal and Regulatory Framework for Pension Fund Investments: Issues and Challenges presented by Poonam Puri set out some regulatory principles, and compared the merits and demerits of rules-based and principles-based approaches to pension regulation. This was followed by an examination of enforcement issues, and some thoughts about the future. Key points included: Important regulatory principles are clear policy goals, a sound theoretical and researchbased foundation, benefits must exceed costs, and they must promote competition, efficiency, and innovation. Rules-based regulation leads to predictability and ease of monitoring. Its downsides include a one size fits all mentality, loophole searching, innovation dampening, and costliness. Principles-based regulation is outcome-focused, has a partnership rather than adversarial orientation, is more pragmatic, interactive, and flexible. Potential downsides include a lack of certainty, is subject to different interpretations, is more challenging for regulators (e.g., how to enforce?), and is better suited to large, rather than small funds. Current issues in regulation: do you remove all rules, or keep some and offer discretionary relief? How to deal with the small funds problem? Restrictions on Pension Fund Investment: A Cost Assessment was presented by Philip Davis and focused on determining if principles-based regulation (i.e., Prudent Person Rule or PPR) produced better pension fund investment results than rules-based regulation (i.e., Quantitative Asset Restrictions or QAR). Key conclusions: Measurement is difficult because many real world regulatory regimes are blends of PPR and QAR. Also, there are other factors at work that may impact comparative investment results by country, such as financial structure, taxation, and governance quality. Finance theory favors PPR over QAR. QAR will hamper efficiency, flexibility, limit diversification benefits, and foster an asset-only management mindset. ICPM Rotman International Centre for Pension Management Page 12 of 17

13 Economic theory also favors PPR over QAR because PPR fosters competition and innovation. QAR may lead to inefficient allocation of capital and increase costs of providing pensions. Empirical findings generally support the theory-based prediction of PPR s superiority. In a 2002 study involving 7 OECD countries, Davis found that pension funds in PPR countries had better reward/risk characteristics than those in QAR countries. The same result was found when the pension fund results were measured against own-country passive benchmarks. In a 2007 study, Hu looked at 22 advanced countries and 17 emerging countries, and produced similar findings. Removing constraints on international diversification was found to be especially important. In a 2008 study, Davis and Hu analyzed Canadian pension fund results versus those of US and UK funds in light of the remaining QAR aspects of Canadian regulation. Once again, the QAR aspect was found to have a cost (note: the Canadian 30% FPR was recently lifted, while some other QARs still remain). Regulatory momentum favoring the PPR philosophy is on the rise, as it should be. Restrictions on Pension Investing: A Canadian Perspective was presented by Michael Nobrega. As the CEO of a major Canadian pension fund, he expressed concern about the remaining Canadian QARs embedded in pension regulations. He pointed out that the average $1 pension benefit in a funded plan is 70% funded by investment income. Specifically: Participation in private markets (e.g., infrastructure, real estate, private equity) is becoming increasingly important element in generating a competitive rate of return on pension assets. This need is making the remaining QARs increasingly troublesome. Examples are 5% limit on individual parcels of real estate or resource properties, 10% limit in any single investment, 15% limit in Canadian resource properties, 25% limit in the combination of real estate and resource properties, 30% limit on voting shares in a single corporation. Canada is lagging in the international movement towards a principles-based regulation philosophy, and has now fallen behind leading countries such as the USA, Netherlands, UK, Australia, and even Japan. This lag is costly in terms of lower investment returns (estimated between 30-90bps), lost opportunities, and higher compliance costs. These drags in turn impact negatively on pension plan sustainability and fairness. There are hopeful signs that Canada will soon deal with its remaining QARs. OMERS specific recommendations include: exempting plans like OMERS from QARs, and making Canadian pension legislation and regulations truly PPR-based. Restrictions on Pension Investing: An Australian Perspective was presented by Leo de Bever. He painted a broad picture of positives and negatives in the current Australian situation: Rapid growth in pension savings continues (currently AUS$1 trillion, 100% of GDP). There is a relatively high-cost retail sector, and a lower-cost wholesale sector. ICPM Rotman International Centre for Pension Management Page 13 of 17

14 Even in the wholesale sector, excess returns have been hard to come by for a number of related reasons. While there are no QAR-type constraints in Australia, there are many others. They include: Australia still has too many funds that are too small; there is a manage by rote mentality that includes weak internal governance, broad use of external consultants and multiple external investment mandates; a focus on what worked in the past; and investment horizons that are too short. A macro consequence of these structural shortcomings is a retirement income system that has pension adequacy problems (i.e., systemic under-saving), longevity risk problems (i.e., a lump sum rather than a life annuity focus), and cost problems (i.e., too high). Restrictions on Pension Investing: A Swedish Perspective was presented by Tomas Franzén. Sweden now has a modest guaranteed minimum pension, and a large Notional Defined Contribution system based on an 18.5% of pay contribution rate (16% goes through four National Reserve Funds (aggregate assets about $150B), and 2.5% goes into a personal pension account). The 4 National Reserve Funds are arms-length government agencies with independent Board of Directors. There is a transparent performance assessment process for the Funds. The Funds are subject to a series of QARs: At least 30% of assets must be in high-grade debt. There is a 5% ceiling on private markets exposure (ex real estate), and commodities investing is totally disallowed. There is a 10% ceiling on investment in the securities of a single issuer/counterparty. Foreign currency exposure must be at least 60% hedged. There are significant limitations on investing in Swedish companies: no more than 10% of voting rights; no more that 2% of market capitalization. At least 10% of assets must be managed externally. Restrictions on Pension Investing: a US Perspective was presented by Joe Dear, who focused on the potential for political interference in the operational and investment decisions of public pension funds. This interference goes beyond enforcing legitimate fiduciary standards. For example: WSIB faced demands for divestment from firms doing business in Sudan and Iran in It was able to achieve an acceptable outcome by agreeing to spell out the actions it would take as an active owner of corporate securities. It also agreed to engage and listen to interest groups. At the same time, WSIB made it clear that it would oppose any kind of mandated restrictions. WSIB had a similar experience with proposals to mandate in-state investments (ETIs). Here the pressure came from elected officials and local investment/venture fund managers. Again, a positive engagement strategy was pursued, pointing out that, ultimately, it was trust money with return targets attached to them. However, holdings in state-based companies and real estate are regularly disclosed. The current backlash against globalization in the USA could bring on new demands for investment restrictions. Outrage at current CEO compensation levels further aggravates the situation. What can be done? Ironically, the answer is skillful engagement in the political process. In politics, when you need a friend, it s too late to get one. Like spiritual masters, pension funds need to be in the world of politics, but not of it. ICPM Rotman International Centre for Pension Management Page 14 of 17

15 What Did We Learn Today? was a group discussion around two questions: (a) what kind of research projects might ICPM sponsor to throw further light on the constraints issue?, and (2) are there collaborative strategies that ICPM Research Partners should be considering to possibly alter constraint regimes that are costly, and no longer serve a useful purpose? Ideas: The variety of constraint experiences recounted this morning pointed to the reality that they come in very different shapes and sizes. Sometimes they are very specific QAR-type constraints still embedded in regulations. Sometimes the constraints arise from the internal governance weaknesses within the funds themselves. Sometimes they arise from the fact that pension funds are large, visible, and accessible to external special interest groups. Measuring the cost of the constraints (whatever their source) is always an important prelude to fostering change. However, it is often not enough. The specific process through which positive change can most likely be effected also needs to be studied and planned. This in turn means pension funds must study how to best execute their role as change agents. Sometimes a fund can do this alone. More often, positive change is most efficiently and effectively attained through collaborative efforts at the national and international levels. An example is to mount a concerted push for principles-based regulation in all jurisdictions. This implies working with international, as well as national regulatory organizations and their overseers to effect change over time. This approach overcomes the them versus us problem that sometimes overshadows relations between the pension management and regulatory communities. ICPM Rotman International Centre for Pension Management Page 15 of 17

16 ROTMAN ICPM OCTOBER 2008 DISCUSSION FORUM Co-hosted by Rotman ICPM and Washington State Investment Board October 21-22, 2008 Carwein Auditorium, Keystone Building (look for T-31 on the building) 1754 Commerce Street, Tacoma Campus, University of Washington (USA) DRAFT - As at June 24, 2008 Day 1 New Developments in Pension Design and Delivery 8:30-8:45 Welcome and the Day 1 Agenda Keith Ambachtsheer, Rotman ICPM (Canada) 8:45-9:30 Individual vs. Collective Elements in Pension Design: New Evidence Eduard Ponds, APG and University of Tilburg (Netherlands) 9:30-10:15 Individualizing Collective Pensions in The Netherlands Theo Nijman, University of Tilburg, and Alwin Oerlemans, Cordares (The Netherlands) 10:15-10:45 Break 10:45-11:30 Estimating Fair Value in National Pension Plans: The Case of Canada Steven James, CPP Investment Board (Canada) 11:30-12:15 The Longevity Annuity: Rethinking Annuitization Jason Scott, Financial Engines Inc. (USA) 12:15-1:15 Lunch 1:30-2:15 The Danish Pension System: A Study in Innovation Chresten Dengsoe, ATP (Denmark) 2:15-3:00 A Template for Pension Reform in Canada, the UK, and the USA Keith Ambachtsheer, Rotman ICPM (Canada) 3:00-3:30 Break 3:30-4:15 Implementing Pension Reform in the UK: A Progress Report Brendan Mulkern, Universities Superannuation Scheme (UK) 4:15-5:00 Pension Reform in the USA: Initiatives at the State Level Sandy Matheson, State of Washington (USA) 5:30 onward Cocktails and pre-dinner remarks by Don Ezra, Russell Investments (USA) followed by Networking Dinner at the Museum of Glass located at 1801 Dock Street minute walk from the campus to the Museum ICPM Rotman International Centre for Pension Management Page 16 of 17

17 Day 2 Designing and Implementing Successful Collaborative Strategies for Pension Funds 8:30-8:45 Theory and Practice of Designing Successful Collaboration Strategies: A Review Keith Ambachtsheer, Rotman ICPM (Canada) 8:45-9:45 Developing Collaboration Strategies at the Washington State Investment Board: A Strategic Initiative Stephen Lerch, Washington State Investment Board (USA) 9:45-10:30 Collaboration Case Study #1: The CPPIB and OTPP FountainVest Joint Venture in China Mark Wiseman, CPPIB (Canada) 10:30-11:00 Break 11:00-11:45 Collaboration Case Study #2: The APG and PGGM Alpinvest Joint Venture Else Bos, PGGM (Netherlands) 11:45-12:30 Collaboration Case Study #3: A Private Markets Collaboration Clearing House for Pension Funds Jeroen Tielman, PensionSummit (The Netherlands) 12:30-1:30 Lunch 1:45-2:30 Collaboration Case Study # 4: The Australian Industry Funds Holdings Joint Venture Kevin Davis and Christine Brown, Melbourne Centre for Financial Studies (Australia) 2:30-3:15 Collaboration Case Study #5: Making Danes Pension-Literate Ole Beier Sørensen, ATP (Denmark) 3:15-3:45 Break 3:45-4:45 Collaboration Case Study #6: The Network for Sustainable Financial Markets Raj Thamotheram, AXA Investment Management (France and UK) 4:45-5:30 Where Do We Take Collaboration Strategies From Here? Group Discussion 6:00 onward Cocktails followed by Dinner. Location TBA This is a by-invitation-only event. For more information please contact Ann Henhoeffer icpm@rotman.utoronto.ca ICPM Rotman International Centre for Pension Management Page 17 of 17