MULTI-ASSET SOLUTIONS OVERVIEW AND PROCESS RISK BUDGETING WITHIN A MULTI-ASSET PORTFOLIO. Nick Samouilhan, PhD FRM CFA.

Transcription

1 RISK BUDGETING WITHIN A MULTI-ASSET PORTFOLIO Nick Samouilhan, PhD FRM CFA Fund Manager, Multi-Asset Funds Aviva Investors February 2013 MULTI-ASSET SOLUTIONS OVERVIEW AND PROCESS 2

2 RISK BUDGETING A PRIMER i) Aim is to maximise the return for a given risk taken, and ii) Take risk up to the point that this is rewarded For most funds (including multi-asset funds) The relevant risk measure is standard deviation to a benchmark (ω) Tracking Error The relevant return measure is return relative to the benchmark (α) Alpha 3 MULTI-ASSET RISK ASSET ALLOCATION AND SECURITY SELECTION Two ways of introducing tracking error, and alpha, to multi-asset portfolios Asset Allocation Security Selection I) Asset Allocation For example, the overweight UK equities, underweight Gilts This will lead to a tracking error And an assumed alpha II) Security Selection For example, the UK exposure is through an active manager, who deviates from UK benchmark This will lead to a tracking error And an assumed alpha 4

3 TYPICAL MULTI-ASSET FUND RISK DECOMPOSITION Benchmark Fund Active Asset Allocation Security Selection Active Balanced Managed Weight Weight Weight Risk Contribution Risk Contribution Risk UK Equity 33.28% 39.71% 6.43% North America Equity 11.78% 7.98% -3.80% Japanese Equity 3.06% 2.13% -0.93% European Equity 11.93% 9.21% -2.72% Pacific Basin Equity 4.42% 6.86% 2.44% Emerging Markets 1.48% 2.55% 1.07% Global Convertibles 0.00% 3.24% 3.24% UK Government 5.66% 2.72% -2.94% UK Credits 3.76% 7.10% 3.34% UK Index-Linked 0.00% 2.14% 2.14% Overseas Fixed 9.22% 5.76% -3.46% Property 0.31% 4.10% 3.79% Alternative Assets - Other 8.64% 2.78% -5.86% Cash Assets 6.46% 3.73% -2.73% Active Total Risk % % 0.00% Source: Aviva Investors, Illustrative Only MULTI-ASSET SOLUTIONS RISK BUDGETING ASSET ALLOCATION 6

4 MULTI-ASSET FUNDS INVESTMENT PROCESS Target Return Cautious fund Volatility Target Returns Volatilities Correlations Manager Record Manager Style Opportunity Set Scenarios and Returns Client Sensitivities Portfolio Optimisation Implementation Risk Monitoring Performance Reporting Europe Equities Quant Passive Active US Equities Passive Returns, Risks etc EM Equities Gov Bonds Active Passive Active Credit Active Commodities External ETF Portfolio Client Requirements Strategic Asset Allocation Vehicle Selection Manager Selection Tactical Asset Allocation Fund Management Multi-Asset Funds 7 ASSET ALLOCATION PROCESS FOUNDATION OF THE PORTOLIO

5 LONG RUN RETURNS FORECASTS US EQUITIES Strong Negative correlation (-0.81) Strong Positive correlation (0.81) 25% US Equities:Realised vs. Average Historical Returns 25% US Equities:Realised vs. Ex-ante Sustainable Returns 20% 20% 15% 15% 10% 10% 5% 5% 0% 0% -5% % yearsubsequent return 20-yearaverageof historical returns 10-year subsequent return Sustainable return Forecasting using sustainable returns works much better than using past returns. Validated approach for over 100 asset classes and alternative beta investments. Source: Aviva Investors Strategy team based on data from Global Financial Database and DataStream, September EXPECTED RETURNS TIME VARYING 4% Sustainable returnover risk-free rate, versus average 3% 2% 1% 0% -1% -2% -3% UK Equities:LTaverage=1.74% UK Property: LT average=1.58% UKCredit: LT average=1.27% EM Equities: LT average=3.58% -4% Source: Aviva Investors Strategy team based on data from Global Financial Database and DataStream, September

6 LONG-TERM FORECASTED ANNUALISED RETURN VS. RISK 10% Pacific equities (ex Korea & Taiwan) EM equities* 8% UK equities Asia-Pacific equities EU equities Global equities ex. UK Sustainable return 6% 4% UK IG Credits Global Converts EMBonds (local currency) European Property UK Property NA equities JP equities Global HY Commodities EM Bonds (hard currency) 2% Europe IG Credits EU Bonds Global Aggregate UK Gilts US IG Credits UK index-linked UK Cash USBonds JP Bonds 0% 0% 5% 10% 15% 20% 25% 30% 35% Expected Volatility (incl. fat-tail adjustment) Source: Aviva Investors Strategy team based on data from Global Financial Database and DataStream, September 2010 ILLUSTRATIVE ASSET ALLOCATIONS LOW, MEDIUM AND HIGH TRACKING ERROR Alpha: 0.5% Alpha : 0.8% Alpha : 1.0% Tracking Error: 1.5% Tracking Error : 3.0% Tracking Error : 5.0% 12

7 MULTI-ASSETS RISK ASSET ALLOCATION RISK (AA) FRONTIER 13 ASSET ALLOCATION HOW MUCH RISK TO TAKE? 14

8 RISK BUDGETING DYNAMIC CHANGES IN THE SAA FRONTIER Frontier will increase if: Increase in Beta Returns Increase in Beta Vols Change in Beta Correls Target IR Take greater risk Its rewarded 15 MULTI-ASSET SOLUTIONS INCORPORATING TACTICAL SCENARIOS 16

9 WHAT ABOUT THE SHORT RUN IN THE LONG RUN, WE HAVE NO CLIENTS Valuation based models focus on delivering returns through full cycle Typically 7 years plus Clients almost always have shorter time frames In addition, there may be short term (tactical) information to take advantage of Increase, temporarily, the equity allocation, for example Scorecard or Compelling Forces approach Unidimensional Scenario based approach Addresses uncertainty (partly) For richer and broader debate on portfolio position Beyond simple equity positioning Significant focus on end client expectations 17 SCENARIO BASED OUTLOOK BOUNDED UNCERTAINTY Source: Aviva Investors Strategy team, Illustrative only 18

10 FROM SCENARIOS TO MARKETS BROAD PATTERNS Source: Aviva Investors Strategy team, Illustrative only 19 EXAMPLE START OF

11 WHAT ABOUT THE SHORT RUN IN THE LONG RUN, WE HAVE NO CLIENTS Optimising for all scenarios likely to result in very little risk being taken The nature of having broadly encompassing views Importance to portfolios based on: Probability of scenario Magnitude of impact on portfolio Client expectations (perhaps most important) Significant difference between Cautious investor Aggressive / Dynamic investor Not always captured in investment policy documents 21 TESTING THE PORTFOLIOS WHAT DO CLIENTS WANT? Portfolio Hard Times Great Expectations Financial Crisis P&L P&L P&L Cautious 4.85% 5.62% -2.79% Dynamic 8.11% 11.33% % Source: BarraOne and Aviva Investors, Illustrative only 22

12 QUARTERLY UPDATE: AVIVA IRELAND MULTI-ASSET FUNDS FUND PERFORMANCE ROBUST TO SCENARIO 23 MULTI-ASSET SOLUTIONS RISK BUDGETING SECURITY SELECTION 24

13 ACTIVE VERSUS PASSIVE SOME BROAD POINTS Market Efficiency Avoid lazy decisions Cost Headline costs are partly irrelevant Index funds vs ETF s vs Futures Regimes Cross-Sectional Volatility Time frames Tactical vs Strategic holding Sometimes Beta is Better 25 ACTIVE VERSUS PASSIVE MARKET EFFICIENCY 26

14 ACTIVE VS PASSIVE MARKET REGIME UK Equity Managers US Equity Managers European Equity Managers Active return Tracking Error Active return Tracking Error Active return Tracking Error Median: Full Sample Median: 1st Quartile Median: 4th Quartile Top Decile of Cross Sectional periods Median: Full Sample Median: 1st Quartile Median: 4th Quartile Bottom Decile of Cross Sectional periods Median: Full Sample Median: 1st Quartile Median: 4th Quartile Full Sample Difference: 1st and 4th Difference: 1st and 4th Difference: 1st and 4th Source: Aviva Investors, 27 MULTI-ASSETS RISK SECURITY SELECTION RISK (SS) Optimisation problem Various combinations of active funds will result in the same tracking error at portfolio level 100% Active UK equity, passive everything else 50% Active UK equity, 75% active credit etc Note the correlation between managers is important Styles matter here Which combination should you pick? Combination that maximised the expected (overall) alpha per tracking error 28

15 MULTI-ASSETS RISK SECURITY SELECTION RISK (SS) Equity W E (50%P, 50% A) Bond W B (25%P, 75% A) Equity W E (65%P, 35% A) Bond W B (25%P, 75% A) αp = W E x [wa,e α Active,E + wp,e α Passive,E +wf,eαfwi,e] + W B x [ wa,b α Active,B + wp,b α Passive,B + wf,b α Fund,B] ω P = [W 2E x [w 2A2,Eω2Active,E + w2p,e ω 2Passive,E +w2f,e ω 2WI,E] + W 2B x [ w2a,b ω 2Active,B + w2p,b ω 2Passive,B + w2f,b ω 2Fund,B] + ij 2W iw j ωi ωjρi,j]^1/2 29 RISK BUDGETING DYNAMIC CHANGES IN THE SELECTION FRONTIER Frontier will increase if: Increase in Alpha Increase in Tracking Error Change in Manager Correls Target IR Take greater risk Its rewarded 30

16 RISK BUDGETING THEORETICALLY COMBINING SS AND AA 31 MULTI-ASSET RISK BUDGETING QUESTIONS For a given risk budget (say, 2% tracking error), how do we efficiently allocate this to AA and/or SS risks? If it is a combination of these risks, what is the best combination? How do we efficiently make use of SS risk in the portfolios given that there are multiple ways of getting the same tracking error from different combinations of active and passive investments in different asset classes? How does the budgeting between the risks change over time? How do we manage this allocation dynamically? On a broader level, how much of the tracking error range should we use, and how does this change dynamically? When should we be at the top end of the range and when should we be at the bottom end? 32

17 HOLISTIC RISK BUDGETING THE OPPORTUNITY FRONTIER Always take risk where it pays off the most Marginally, assign to the risk (SS or AA) with the highest expected alpha/te (IR) If done incrementally, this results in an Opportunity Frontier Showing combination that gives the highest expected alpha per unit of risk % 1.0% Expected Tracking Error 2.0% Security Selection 1.0% Expected Tracking Error Opportunity Frontier 1.0% Expected Excess return Asset Allocation Expected Excess return 2.0% Expected Excess return HOLISTIC RISK BUDGETING THE OPPORTUNITY FRONTIER Expected Tracking Error 34

18 Expected Excess return 2.2% 2.0% Expected Excess return HOLISTIC RISK BUDGETING THE OPPORTUNITY FRONTIER Asset Allocation Security Selection 1.0% 1.0% 2.0% 1.0% Expected Tracking Error Expected Tracking Error 30% 2.0% Expected Excess return Opportunity Frontier 1.0% 2.0% Expected Tracking Error 35 OPPORTUNITY FRONTIER HOW MUCH RISK TO TAKE? 36

19 HOLISTIC RISK BUDGETING DYNAMIC CHANGES IN THE OPPORTUNITY FRONTIER Frontier will increase if: Increase in AA payoff Increase in SS payoff Change in correlation structure Beta and Alpha Target IR Take greater risk Its rewarded 37 MULTI-ASSET RISK BUDGETING THE WHAT, HOW AND WHEN For a given risk budget (say, 2% tracking error), how do we efficiently allocate this to AA and/or SS risks? By allocating to highest IR If it is a combination of these risks, what is the best combination? In a way that overall portfolio alpha is maxed per unit of risk How do we efficiently make use of SS risk in the portfolios given that there are multiple ways of getting the same tracking error from different combinations of active and passive investments in different asset classes? By allocating to combination with highest alpha per unit of risk How does the budgeting between the risks change over time? How do we manage this allocation dynamically? Dynamically, as SS and AA (and hence Opportunity Frontier) will self adjust On a broader level, how much of the tracking error range should we use, and how does this change dynamically? Take highest risk consistent with minimum IR target 38

20 Q&A 39 BIOGRAPHY Nick Samouilhan Fund Manager, PhD, CFA, FRM Main Responsibilities Nick is responsible for responsible for managing a range of multi-asset funds at Aviva Investors Experience & Qualifications Prior to joining Aviva Investors, Nick worked at Investec Asset Management on their Multi-Asset desk. His specific focus was on cross-asset research, quantitative modelling and strategic asset allocation. Before joining Investec, Nick worked in academia, where he lectured economics and finance, and has published multiple academic papers on volatility modelling, macro-economic forecasting and portfolio diversification. Nick holds a PhD degree in Economics from the University of Cape Town and the Investment Management Certificate. Nick is also both a FRM charterholder and a CFA charterholder. 40

21 IMPORTANT NOTES This document is intended for institutional or professional investors and experienced advisors only. Private investors and scheme members should seek professional advice before making an investment decision. The information provided in this document and any appendix is confidential and should only be used for the purposes requested. Except where stated as otherwise, the source of all information is Aviva Investors as at 31 December 2012 Any future returns and opinions expressed are based on the internal forecasts of Aviva Investors and should not be relied upon as indicating any guarantee of return from an investment managed by Aviva Investors. No part of this document is intended to constitute advice or recommendations of any nature. Where performance has been illustrated it is not intended to be a guide to the future. Other information sourced from Aviva Investors. The value of an investment in the fund can go down as well as up and can fluctuate in response to changes in exchange rates. Past performance is not a guide to the future. Full terms and conditions of the Aviva Investors products and services are available on request. Aviva Investors Global Services Limited, registered in England No Registered Office: No. 1 Poultry, London EC2R 8EJ. Authorised and regulated in the UK by the Financial Services Authority and a member of the Investment Management Association. Contact us at Aviva Investors Global Services Limited, No. 1 Poultry, London EC2R 8EJ. 41