Migration Tracking and Dynamic Weighting February 28, 2005 Team Members Dae Jin Choi Nicolas Paleokrassas Juliet Xu John Duval Wei King Ng Global Blue.

Slides:



Advertisements
Similar presentations
Quantitative Stock Selection Portable Alpha Gambo Audu Preston Brown Xiaoxi Li Vivek Sugavanam Wee Tang Yee.
Advertisements

FACTORIAL ANOVA Overview of Factorial ANOVA Factorial Designs Types of Effects Assumptions Analyzing the Variance Regression Equation Fixed and Random.
Chapter 11 Optimal Portfolio Choice
Optimal Portfolio Choice and the Capital Asset Pricing Model
Principles of Corporate Finance Session 29 Unit IV: Risk & Return Analysis.
1 Fin 2802, Spring 10 - Tang Chapter 24: Performance Evaluation Fin2802: Investments Spring, 2010 Dragon Tang Lectures 21&22 Performance Evaluation April.
Diversification, Beta and the CAPM. Diversification We saw in the previous week that by combining stocks into portfolios, we can create an asset with.
INVESTMENTS | BODIE, KANE, MARCUS ©2011 The McGraw-Hill Companies CHAPTER 7 Optimal Risky Portfolios 1.
Economy / Market Analysis
1-1. Copyright © 2005 by The McGraw-Hill Companies, Inc. All rights reserved.McGraw-Hill/Irwin 1 A Brief History of Risk & Return.
Information Analysis. Introduction How can we judge the information content of an idea? More concretely: –Is this a signal we can profitably deploy.
Return, Risk, and the Security Market Line
International Fixed Income Topic IVC: International Fixed Income Pricing - The Predictability of Returns.
© 2003 The McGraw-Hill Companies, Inc. All rights reserved. Return, Risk, and the Security Market Line Chapter Thirteen.
Information in the term structure of variance swaps and CFO predictions of volatility Whit Graham, Josh Kaehler, Matt Seitz.
Dynamic Factor Weights Red Devil Partners Joon Seong Choi, Youngjun Yoo, Richard Park, YK Kim.
1 ASSET ALLOCATION. 2 With Riskless Asset 3 Mean Variance Relative to a Benchmark.
Quantitative Stock Selection Campbell R. Harvey Duke University National Bureau of Economic Research Global Asset Allocation and Stock Selection.
Predictive versus Explanatory Models in Asset Management Campbell R. Harvey Global Asset Allocation and Stock Selection.
GLOBAL ASSET ALLOCATION AND STOCK SELECTION ASSIGNMENT # 1 SMALL CAP LONG-SHORT STRATEGY FIRST-YEAR BRAVES Daniel Grundman, Kader Hidra, Damian Olesnycky,
Global Asset Allocation and Stock Selection Assignment #1
Long/Short Trading Strategy Cam’s Crazies Global Asset Allocation February 2005.
Presentation by: Bryan Durand Josh Amoss Suri Thummala Steve Beuchaw Matthew Malouin Global Asset Allocation February 28, 2005.
Long/Short Sector-based Trading Strategy Emergent Asset Management, LLC Konstantin Savov Scott Smith Pin-Yew Wong Vaswar Mitra Vinaya Jain February 27,
1Capital IQ, A Standard & Poor’s Business Variations on Minimum Variance March 2011 Ruben Falk, Capital IQ Quantitative Research.
Chapter McGraw-Hill/Irwin Copyright © 2009 by The McGraw-Hill Companies, Inc. All rights reserved. 13 Performance Evaluation and Risk Management.
13-1. Copyright © 2005 by The McGraw-Hill Companies, Inc. All rights reserved.McGraw-Hill/Irwin 13 Performance Evaluation and Risk Management.
Optimal Risky Portfolios
Investment Analysis and Portfolio management Lecture: 24 Course Code: MBF702.
Empirical Financial Economics Asset pricing and Mean Variance Efficiency.
 Lecture #9.  The course assumes little prior applied knowledge in the area of finance.  References  Kristina (2010) ‘Investment Analysis and Portfolio.
Growth vs. Value Trading Strategies Global Asset allocation John O’Reilly Sebastian Otero Barba Nikolay Pavlov Franck Violette.
Chapter 3 Delineating Efficient Portfolios Jordan Eimer Danielle Ko Raegen Richard Jon Greenwald.
Chapter 17 Partial Correlation and Multiple Regression and Correlation.
Chapter 06 Risk and Return. Value = FCF 1 FCF 2 FCF ∞ (1 + WACC) 1 (1 + WACC) ∞ (1 + WACC) 2 Free cash flow (FCF) Market interest rates Firm’s business.
A 1/n strategy and Markowitz' problem in continuous time Carl Lindberg
Chapter 10 Capital Markets and the Pricing of Risk.
Chapter 10 Capital Markets and the Pricing of Risk
A Multi-Factor Residual-Based Trading Strategy Finance 453 Adrian Helfert Terry Moore Kevin Stoll Ben Thomason February 26, 2004.
1 Mutual Fund Performance and Manager Style. J.L. Davis, FAJ, Jan/Feb 01, Various studies examined the evidence of persistence in mutual fund performance.
1 The Allocators Presented By: Ainsley Fuhr Mike Gabriel Nate Rozof Graig Saloom Greg Williamson February 27, 2006 Investigating “Innovation Factors” for.
0 Presentation by: Austin Applegate Michael Cormier Paul Hodulik Carl Nordberg Nikki Zadikoff Global Asset Allocation February, Granite Investments.
CHAPTER SEVEN Risk, Return, and Portfolio Theory J.D. Han.
Chapter Performance Evaluation and Risk Management McGraw-Hill/IrwinCopyright © 2012 by The McGraw-Hill Companies, Inc. All rights reserved. 13.
Growth Investing: Growth at a “reasonable” price Aswath Damodaran.
Quantitative Stock Selection: Dynamic Factor Weights Campbell R. Harvey Duke University National Bureau of Economic Research.
0 Presentation by: Sanjun Chen Noah Harris Haley Lai Nicolas Tollie Global Asset Allocation Alpha.
Quantitative Stock Selection: Practical Insights Campbell R. Harvey Duke University National Bureau of Economic Research.
Travis Wainman partner1 partner2
Market Cycle Varying Multifactor Strategies Cyclical Analysts Noah Harris Juliet Xu JJ Haines.
1 Conditional Weighted Value + Growth Portfolio (a.k.a MCP) Midas Asset Management Under the instruction of Prof. Campbell Harvey Feb 2005 Assignment 1.
CHAPTER 9 Investment Management: Concepts and Strategies Chapter 9: Investment Concepts 1.
The Alpha Team How to Construct a Global Portfolio After a Yield Curve Inversion.
Behavioral Finance Fama French March 24 Behavioral Finance Economics 437.
Behavioral Finance Economics 437.
Portfolio Management Portfolio Evaluation March 19, 2015 Slide Set 2 1.
1 Mutual Fund Performance and Manager Style. J.L. Davis, FAJ, Jan/Feb 01 Various studies examined the evidence of persistence in mutual fund performance.
EQUITY-PORTFOLIO MANAGEMENT
ETF Consumer Cyclicals
What Factors Drive Global Stock Returns?
Portfolio Risk Management : A Primer
…A Quantitative Approach
Overview Hypothesis: Develop a market-neutral long/short strategy
Bright Sun Asset Management
Performance Evaluation and Risk Management
Overview : What are the relevant factors? What are the
Small Cap Quantitative Factors Analysis
Informing the Choice of Fractile Usage in Factor-Based Stock Screening
Index Models Chapter 8.
Behavioral Finance Economics 437.
Presentation transcript:

Migration Tracking and Dynamic Weighting February 28, 2005 Team Members Dae Jin Choi Nicolas Paleokrassas Juliet Xu John Duval Wei King Ng Global Blue Devil Partners

- 1 - Global Blue Devil Partners CONTEXT AND PURPOSE Context Global Blue Devil Partners wanted to explore: mechanics of factor selection and evaluation migration tracking static and dynamic weighting Purpose of this document Explain methodology employed Summarize key results Evaluate migration tracking, static and dynamic weighting

- 2 - Global Blue Devil Partners Analysis Road Map: The Different Returns Analyzed Long-Short Migration Tracking Static Weights Optimization Dynamic Weights Optimization Optimized score Long-Short Complexity X XX X X X XX X X X X XX XX XX X XXX X X X X X X

- 3 - Global Blue Devil Partners Started By Examining a Large Number of Factors Retained Earnings Growth/ΔMarket Cap Current Price / 52 Week High Price CFO/Price EPS Growth-Price Growth EPS Growth* Price Growth* (E/P) / (Book/Price) *Also Examined by Industry Only Looked at Largest 500 Compustat Companies, Best Factors: Retained Earnings Growth/ Δ Mkt Cap CFO/Price Price Growth Examined Heat Maps and chose based on consistency of returns, average return, Sharpe Return, Maximum Drawdown Kept Fractiles 1 and 5 for each Factor Factors ExaminedResults Chosen Factors a combination of Price Momentum and Value strategies

- 4 - Global Blue Devil Partners Value Weighted Individual Screen Results Returns (%) Measures Sharpe Ratio Maximum Negative Excess Return (LS) Max No. Consecutive Down Periods Retained Earning Growth/Market Cap CFO/PricePrice Growth Fractile 1Fractile 5Fractile 1Fractile 5Fractile 1Fractile % Period Negative Returns (LS) (17.7%) 4 (12%) 9 (8.7%) 5

- 5 - Global Blue Devil Partners But can Migration Tracking Improve the Results? Rank T Rank T-1 Rank T-2 Rank T-3 Created New Factors based on lagged variables Computationally intensive to do for all time periods, so chose a few discrete intervals Combined new factors with subjective score giving more weight to more recent ranks Not possible if variables hard to lag Important to remember difference in rank of average factor versus average factor rank! Methodology

- 6 - Global Blue Devil Partners Value Weighted Migration Tracking Individual Screen Results Returns (%) Measures Sharpe Ratio Maximum Negative Excess Return (LS) Max No. Consecutive Down Periods Price Growth Earnings/Price & Book/Price CFO/Price % Period Negative Returns (LS) Retained Earnings Growth/ Market Cap 0.02 Fractile 1,Fractile (18%) Fractile 1,Fractile (10%) Fractile 1,Fractile (24%) Fractile 1,Fractile (24%) 4

- 7 - Global Blue Devil Partners Summary of Migration Tracking CFO / Price: Worse Price Growth: Better Earnings/Price / Book/Price: Mixed Results Retained Earnings Growth / Market Cap: Better Unclear whether or not should expect better results Expected Migration Tracking to work better for some factors than others: Persistent effects modeled better with migration tracking Long-lived rankings may already be priced into market CFO / Price: Worse Price Growth: Better Earnings/Price / Book/Price: Mixed Results Retained Earnings Growth / Market Cap: Better Unclear whether or not should expect better results Expected Migration Tracking to work better for some factors than others: Persistent effects modeled better with migration tracking Long-lived rankings may already be priced into market

- 8 - Global Blue Devil Partners After Migration, Optimization Looked at Simple Long-Short Portfolios & Long-Short Migration Tracking Portfolios Result: Optimized 3-factor Migration tracking portfolio beats 3-factor No-Migration Tracking portfolio. Expected Return: 25.33%, Sharpe Ratio: 1.24 Optimizing with multiple fractile always creates a portfolio with a higher Sharpe Ratio and Expected Return, and beats any single factor. The weights from the optimization are static portfolio weights that we can apply to the fractiles Max(w 1 r 1 + w 2 r 2 + w 3 r 3 + w x r x ) Var<= Variance of S&P 500 Total Returns w 1 + w 2 + w 3 + w 4 = 0 Still use fractiles 1 and 5 Fractile returns assumed to equal mean historical- no re-sampling Results

- 9 - Global Blue Devil Partners Putting Optimization Weights into a Scoring Model: c1c2....cnc1c2....cn Factor 1 Fractile 1 W1W1 c1c2....cnc1c2....cn W2W2 c1c2....cnc1c2....cn W3W3 c1c2....cnc1c2....cn …W x 1. Apply weights to Companies… 2. Add weights together to create a score for each company across fractiles… For company 1, ∑ (W 1c1 + W 2c1 + W 3c1 + W x,c1 ) = Score C1 Factor 1 Fractile 5 Factor 2 Fractile 1 Factor X Fractile 1

Global Blue Devil Partners Putting Optimization Weights into a Scoring Model: (2) Fractile 1 Score Factor 3. Use score as a new factor to sort and rank: Fractile 2 Fractile 3 Fractile 4 Fractile 5 4. Go long fractile 1, short fractile 5

Global Blue Devil Partners Optimized weights scoring model are not necessarily better! Returns (%) Measures Sharpe Ratio Maximum Negative Excess Return (LS) Max No. Consecutive Down Periods MT Price Growth & E/P with B/P Price Growth & E/P with B/P MT Price Growth & CFO/Price % Period Negative Returns (LS) Price Growth & CFO/Price 0.16 Fractile 1,Fractile (11%) Fractile 1,Fractile (14%) Fractile 1,Fractile (12%) Fractile 1,Fractile (14%) 7

Global Blue Devil Partners Dynamic Weighting r1r2....rtr1r2....rt Fractile 1 = W 1 r1r2....rtr1r2....rt Fractile 2 +W 2 r1r2....rtr1r2....rt Fractile 3 +W 3 r1r2....rtr1r2....rt Fractile X …+W x Optimized weights were fixed over time: Portfolio Return Not optimal if correlations or expected returns differ by period Creates need to find a way to dynamically change weights

Global Blue Devil Partners Dynamic Weighting r1r2r2...rtr1r2r2...rt Fractile X ΔTerm Structure Slope 0r2r2....0r2r2.... Fractile X (Dynamically Weighted) x Interacting Fractile Returns with Change in Slope of Term Structure Created Dynamic Weights The new fractile returns: Underperformed other portfolios Lost ability to invest when down market predicted Next Step: Only Interact Fractile 1 = Insight: Changing the returns dynamically effectively changes the weights dynamically:

Global Blue Devil Partners Dynamic Weighting with Logit and Probit Models r1r2r2...rtr1r2r2...rt Factor 1 Fractile 1 - r1r2r2...rtr1r2r2...rt Factor 1 Fractile 5 1. Create a dependent variable that is equal to 1 if fractile 1 beats 5, 0 otherwise 2. Regress Binary Variable on Predictive Variable: Lagged Yield Spread Υ = α + β 1 Χ 1 + ε Υ = 0 or 1 Must use logistic or probabilistic distribution: bounded by 1 and zero Interpretation of result as probability of a positive long- short, dynamically weight based on probability Best if done by company, not fractile

Global Blue Devil Partners SUMMARY AND KEY LEARNINGS Migration tracking allows fine-tuning of factors, and may be especially valuable for use with persistent factors. Scoring model results in inconsistent improvements, but appears to work better for factors that incorporate migration tracking. Value of dynamic weights are highly dependent on power of predictive variables, and may be less valuable with long-short models that consistently provide returns in both up and down markets. Further Areas of Potential Study: -Incorporate trading costs into optimization by penalizing returns -Put dynamic returns into scoring model -Analyze longer time periods -Calculate Migration Tracking variables over different time periods or with different subjective weights