Adaptive Asset Allocation Butler V1.3 AMM

Inspired by Butler (2012) Adaptive Asset Allocation: A Primer (http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2328254)

"The paper addresses flaws in the traditional application of Modern Portfolio Theory related to Strategic Asset Allocation. Estimates of parameters for portfolio optimization based on long-term observed average values are shown to be inferior to alternative estimates based on observations over much shorter time frames. An Adaptive Asset Allocation portfolio assembly framework is then proposed to coherently integrate portfolio parameters in a way that delivers substantially improved performance relative to SAA over the testing horizon."

- Top5 out of 11 diversified assets selected by 6m Momentum
- Minumum Variance Optimization based on 6m Covariance matrix
- Volatility Targeting, optionally leverage

Asset classes: U.S.Stocks, Internaonal REIT, European Stocks, U.S.Intermediate Treasuries, Japanese Stocks, U.S.long term Treasuries, Emerging Market Stocks, Commodites, U.S. REITs, Gold (needs non-public data for full backtest)

Paper Results with 8% volatility target, weekly rebalancing:
- CAGR 16.91%
- Sharpe 2.15
- MaxDD -9.56%
- MAR 1.77

V1.3:
- Finished Excel output for detailed analysis, see comments in QS language part
- Added Dynamic Asset list, e.g. ETF after inception date, MFs / indices before (12 ETF / 12 MFs), see comments in QS language part
- Some cleaning and debugging

V1.2:
Update: Most cleaning done, improved price series acuracy before passing to solver

- Moved completely to AMM, code still improvable (bit messy, ok)
- Added Minimum Variance Optimization through MS Solver Foundation (needs some configuration, see gdrive)
- Added Portfolio Volatility Target triggered either by historical or expected volatility
- Added leverage up/down to portfolio volatility target, either by leverage factor or by changing weight constraints
- Added different volatility metrics for visual analysis
- Tweaked Excel and QS output window interface for stats and debugging

Now very stable over parameters, Sharpe up to 1.8, with Volatility <12% and +20% performance level (depending on vola target and leverage).
Hope this is a good foundation to further develop, some ideas:
- Covariance shrinkage and using "crisis" / "bull" moment covariance overlay (good results in Excel)
- Tweak trend indicator, e.g. different lookbacks in bear and bull markets, consider volatility
- Check different asset sets, e.g. more diversity and set weighting constraints depending on market environment

V1.1:
- Changed symbols to 7 mutual funds available in yahoo
- Added further descriptions to AMM code for easier reference and changes

More on our strategies here:
http://www.logical-invest.com

Technical Analysis

• Number of Times an Event Occurred in the First N-Minutes of a Trading Session
• Negative trend breakout list (SMA200-based)
• Correlation Matrix - 41 Major ETFs
• Directional Trend Index Strategy
• Advancing-Declining-Unchanged volume for NYSE, AMEX and NASDAQ
• Forecast Oscillator
• Williams' VIX Fix - WVF
• Butler Adaptive Asset Allocation Model V1.0
• Hull Moving Average Strategy - Multi Time Frame Trading System
• Ascending, Descending and Symmetrical Triangle Chart Patterns
• More trading objects...

Fundamental Analysis

• AAII Sentiment Survey - Historical Data
• Industry Relative Strength
• Earnings after and before market opens
• Yale - Valuation Confidence Index
• Currency Pairs Calendar Data
• Federal funds effective rate
• Consumer Credit Historical Data
• Industry Equity
• Index Investment Data for U.S. Futures Markets
• Short Selling Data
• More trading objects...