1 / 29

Ex Post Relative Valuation Process Honing

Sector Optimization for Fixed-Income Portfolios Constrained By Value-at-Risk and Traditional Risk Measures Ron D’Vari, Juan C. Sosa, Kishore Yalamanchili State Street Research and Management 8TH ANNUAL IAFE CONFERENCE New York City October 14-15, 1999.

missy
Download Presentation

Ex Post Relative Valuation Process Honing

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Sector Optimization for Fixed-Income Portfolios Constrained By Value-at-Risk and Traditional Risk Measures Ron D’Vari, Juan C. Sosa, Kishore Yalamanchili State Street Research and Management 8TH ANNUAL IAFE CONFERENCE New York City October 14-15, 1999

  2. Risk-constrained Optimization Facilitates Integration of Various Sector Views In Portfolio Construction • Research • Macro • Quantitative • Credit • Ex Ante • Expectations • Markets • Spreads • Risks • Portfolio Synthesis • Maximize Return • Minimize Risk • Ex Post • Monitoring • Attribution • Ex Post • Relative Valuation • Process Honing State Street Research & Management

  3. Risk-Constrained Fixed-Income Sector Optimization Incorporating VaR and Traditional Risk Measures • Objectives • Risk Models • Risk-Constrained Optimization • Results • Conclusions State Street Research & Management

  4. Objectives • Validated, Comprehensive, and Flexible Risk Model • Tactical Sector Allocation and Optimization Model: • Integration of tactical views of all research teams • Incorporation of risk explicitly in the investment process • Comprehensive Tool to Synthesize Fixed-Income Portfolios: • Maximize return under a set of probability weighted scenarios • Constrain risk • Traditional measures such as relative duration, sector weights, duration contribution • Stress-scenarios incorporating outliers and extreme observations • Flexible historical Value-At-Risk (VaR) allowing for non-normal, time-variant distributions with fat tails State Street Research & Management

  5. Value-At-Risk Models • Objectives: Calculate distribution of returns and downside risk • Comprehensive - include interest rate risk, curve and spreads, for all major fixed-income sectors • Flexible - allow specification of time window, decay factor, and confidence level • Accurate - account for non-normally distributed asset classes such as MBS, high yield, and emerging market debt State Street Research & Management

  6. Methodology • Domestic High Grade • Weekly derived spread data from individual securities in the Government/Corporate/Mortgage universe • Variance/Covariance • High Yield • Weekly aggregate spread data for subsectors • GARCH with shocks • Emerging Markets • Weekly aggregate spread data for subsectors • GARCH with shocks • Portfolio VaR and Relative VaR estimated via Structured Monte Carlo simulation with rolling correlation matrix State Street Research & Management

  7. Methodology Work-in-Progress • Asset Backed Securities • Nondollar • Alternative specification for domestic assets (e.g. AR(1) to address mean reversion) State Street Research & Management

  8. Why Not Variance-Covariance? • Model choice has significant effect on the estimation of risk-return trade-offs, hence the optimal choice of portfolios • Variance-Covariance VaR underestimates risk of non-normal assets (e.g. High Yield and EMBI) State Street Research & Management

  9. Certain Fixed Income Sectors Exhibit Strongly Non-Normal Behavior State Street Research & Management

  10. 4-week 95% VaR for Government/Corporate/Mortgage Plus Alternative Sector, August 27, 1999 State Street Research & Management

  11. Ratio of 4-week Simulated Expected Return to 4-week 95% VaR of Gov./Corp./Mtg. Plus Alternative Sector, August 27, 1999 State Street Research & Management

  12. 4-week 95% VaR for Government/Corporate/Mortgage Plus Alternative Sector, August 27, 1999 State Street Research & Management

  13. Ratio of 4-week Simulated Expected Return to 4-week 95% VaR of Gov/Corp/Mtg Plus Alternative Sector, August 27, 1999 State Street Research & Management

  14. Constrained Optimization Test Case • Return optimized over 6 month horizon • Unchanged term structure and spreads returning to their mean • Choice of Constraints • Traditional: Duration • Value at Risk (4 Week, 95 Percentile) • Five Stress Scenarios: • Unchanged term structure and spreads (UNCH) • Unchanged term structure and max spreads (MAXSPD) • Unchanged term structure and min spreads (MINSPD) • Max treasury yields and corresponding spreads (MAXTSY) • Min treasury yields and corresponding spreads (MINTSY) • Time Period: July 5, 1996 to Aug 27, 1999 State Street Research & Management

  15. State Street Research & Management

  16. Optimization Test Case- Initial Data State Street Research & Management

  17. Optimization Test Case - Scenario Set State Street Research & Management

  18. Optimization Test Case - Scenario Set State Street Research & Management

  19. Under No Constraints Emerging Market is Asset of Choice State Street Research & Management

  20. Duration Constraint Alone Is Inadequate State Street Research & Management

  21. VaR Constraint Leads to Reasonable Allocation State Street Research & Management

  22. Addition of Duration Constraint Modifies Solution Modestly State Street Research & Management

  23. Duration and Scenario Constraints AloneCould Lead to Extreme Solutions State Street Research & Management

  24. VaR and Stress-Scenario Constraints Combined Lead to Reasonable Overall Allocation State Street Research & Management

  25. VaR and Stress-Scenario Constraints Combined Lead to Reasonable Overall Allocation State Street Research & Management

  26. VaR-Constrained Efficient Frontier State Street Research & Management

  27. VaR-Constrained Efficient Frontier State Street Research & Management

  28. Conclusions • VaR Model Choice Is Significant in Assessing Risk-Reward in Portfolios • Risk-Constrained Optimization Facilitates Integration of Various Sector Views In Portfolio Construction • Experience Is Required to Select Suitable Model Parameters and Choice of Constraints State Street Research & Management

More Related