Monthly Rebalancing System
A long-only, ETF based, systematic, monthly momentum system with risk management and exposure to global markets and commodities
Ken Long
Ver 1.1
5/01/2012
Please review disclaimer, page 2
Disclaimer:
* Please remember that these stock screening services are for educational purposes to show possible tactics that can be used in the markets. Trading stocks has significant profit potential and also carries a significant risk of loss. You are responsible for your own trading decisions! Obviously, you are responsible for your own trading results and individual results vary. The information contained here is believed to be accurate, but no claims or warranties implied or otherwise are made pertaining to the accuracy of this information and we are not responsible for errors or omissions. There is a very high degree of risk involved in stock trading. Posted past results are hypothetical trades. Results do not reflect slippage, commissions and other market risk factors inherent in actual trades in the stock market. Past results are not indicative of future performance. Tortoise Capital Management., Ken Long, or anyone affiliated with this educational screening assume no responsibilities for your trading results. Tortoise Capital Management, Inc., Ken Long, or anyone affiliated with this educational screening, including friends and relatives, may or may not take a position or have a position in these stocks, based on their trading discretion at the time. You should assume that these positions may be substantial enough to be material. We will not intentionally scalp or front run on any recommendations; The strategies, lessons and all other features are for educational purposes only and should not be construed as investment or trading advice. Tortoise Capital Management, and Ken Long are not offering professional trading advice in this educational service. You must assess the risk of any trade with your broker or financial professional and make your own independent decisions regarding any trades. Copyright 2011 Tortoise Capital Management, Inc. All Rights Reserved. Protected by copyright laws of the United States and international treaties. This newsletter, e-letter, or promotional material may only be used pursuant to the subscription agreement and any reproduction, copying, or redistribution (electronic or otherwise, including on the world wide web) in whole or in part, is strictly prohibited without the express written permission of Tortoise Capital Management, Inc.
System Name: Blended monthly rebalancing
a. Version: 1.1
b. Date of last update: May 01, 2012
c. Summary of change:
i. 1.01: fixed numbering of appendices; fixed cash ETF to SHY; added FXF to portfolio 22
ii. 1.1: clarified recommendation for a 10% trailing stop from highest high while in the trade
2. System description:
a. A momentum-based, asset allocation system designed for monthly reallocation of money in a set of ETF's that cover the world market model. Incorporates 2 look backperiods (3 month and 6th month) and a moving average filter that acts as a stop-loss and as a buying filter that directs money to cash in weak markets.
3. System logic, concepts and definitions:
a. applies the principles of modern portfolio theory, momentum and asset allocation and low-cost ETF's
b. relies upon three-month and six month momentum, which have a persistent edge found in scholarly literature, backtesting and forward testing
c. Uses a blend of 70% of the 3 month lookback and 30% of the 6 month lookback to generate the RS score. (Formula: (.7 x 3 mo return%) + (.3 x 6 mo return%) = blended return%; Calculate blended return% for each member of the portfolio universe, then rank them 1-n.
d. Seeks to outperform the market by being fully invested among the top performers in a population of ETF's, or in cash positions when the market is sufficiently weak.
e. an extension of countless momentum conceptions tested in scholarly literature; (See work by David Aronson for a good literature review, and French’s work on intermediate and short term momentum)
f. Popular literature featuring sound work has been done by Gerald & Marvin Appel, Robert Miner, David Vomund, Vijay Singal)
4. Targeted stocks/ETFs/markets:
a. broad market indexes, represented by low-cost ETF's
b. version 1 (13 ETFs): DIA, EEM, EFA,EPP, EWA, EWJ, ILF, IWM, IYR, MDY, QQQ, SPY, TLT, and SHY as the cash equivalent
c. version 2 (22 ETFs): DBA, DBC, DIA, EEM, EFA, EPP, EWA, EWJ, EWZ, FXF, FXI, GLD, IEV, ILF, IWM, IYR, MDY, QQQ, SPY, TLT, XLE, XLF
5. Historical performance: 7 yr back test: 01/2003 to 06/2011
a. See Appendix 1-3 for details on the variety of backtests conducted at ETFreplay.com to compare a variety of parameters which including varying:
i. ETF populations (13, 22, 23, 26 ETF populations)
ii. A variety of lookback periods to compute Relative Strength
iii. A variety of blends of lookback periods
iv. The number of best performing ETFs to hold each month (1-6)
v. A monthly moving average to act as a buying filter (0-12)
vi. A choice of ETF to use as a cash position (SHY, TLT)
b. Summary of tests: a robust set of parameters with a limited risk of curve fitting include the following reasonable set outlines in the Entry rules below
6. Setups/Filters:
a. Each month the ETFs in the population will be ranked 1-n by blended Relative Strength (RS)
b. Blended RS will use 70% of the 3 month and 30% of the 6 month %Price performance
c. Moving average filter: the 4 month MA of each top ranked ETF will be used to determine if the top performing ETFs can be purchased. If their current price is below the 4 month MA, then the cash ETF will be used instead for that month.
d. The MA filter will be applied to each ETF based on its own MA and price.
7. Entry rules:
a. Long Rules;
i. Rank the portolio 1-n; Buy the top 3 in equal dollar amounts if the price > 4 month MA. If Price < 4 month MA, place that portion of the portfolio in the Cash ETF.
ii. always invested, in either the top 3 rated ETFs or in the cash ETF
iii. Buy the top 3 in equal dollar amounts and hold for a month;
b. Short rules: none (research ongoing to develop a ruleset for this contingency)
c. Pyramiding entries?: None
d. recommended percent allocations: equal dollar amounts for your intermediate to long term portfolio
8. Exit rules:
a. Capital preservation stops: use a 10% trail from the highest high while in the trade as the stop; if the stop is hit put the money in the cash ETF until the recalculation at the next period.
b. Profit preservation stops: none
c. Scaled exits?: None
9. Position sizing rules:
a. equal allocation on a percent of portfolio basis
10. Optional decisions/rules: considerations for research
a. Perform the calculations 5 days before the end of the month in order to take advantage of the end-of-month seasonality advantage
b. Consider TLT as the cash alternative ETF replacing SHY; adds volatility but the possibility of additional return.
c. Continue to examine the universe of ETFs to find legitimate additions to the population: they should represent distinct sectors/subsets of the world market model.
d. Consider running a separate subset for commodities and expand the offerings
e. Consider allocations to the short side using tactical bear market strategies when cash positions are indicated.
11. Preferred brokers: low cost transaction per share is the most important decision criteria for this system. If you are an intraday trader trying for more, then consider slippage and speed of execution as well. Any large deep discount broker should satisfy these requirements.
12. How to start the portfolio from all cash: use money from your total portfolio that is allocated for equity exposure and begin.
Appendix 1: Rebalancing various single look-back periods with no MA filter, from a population of 13 ETFs and holding either 1 or 3 months:
331: Lookback 3 months, hold the best 3 performers, for 1 month
631: Lookback 6 months, hold the best 3 performers, for 1 month
333: Lookback 3 months, hold the best 3 performers, for 3 months
All are compared against a baseline of buy and hold SPY for the duration.
Conclusion: all are robust, 331 outperforms. Supports the position that timing models based on RS add value.
Appendix 2: effect of various MA filters on individual indices
Portfolio mix:
1. Browne 4: from Harry Browne’s Permanent Portfolio
2. PA5: 5 ETFs intended from a passive annual asset allocation portfolio allocation model that outperforms SPY Buy and Hold. Either equally weighted for weighted to equities
3. PA10: 10 global ETFs from a passive annual asset allocation model that outperforms SPY Buy and Hold. Either equally weighted for weighted to equities
4. Individual ETFs representing broad indices or well defined asset classes.
Conclusions: MA filters are robust across all indices: shorter term MA of 4 months and longer term MA of 10 satisfice. Robust for combined portfolios and single, broad indices. This is consistent with studies of long term Mas going back multiple decades and beyond. Adds credibility to the position that a combination of momentum and timing can add value, as is seen in Appendix 3.
Appendix 3: Monthly rebalancing, blended RS, various MA filters, cash equivalent SHY vs TLT, 13 or 22 ETF portfolios
Conclusions: all are robust; 4MA or 10MA satisfice; TLT adds volatility and return for cash position, holding 3 seems like a good tradeoff in either 22 or 13 ETF portfolios
Column 1: uses 13 ETF portfolio; Columns 2&3 use 22 ETF portfolios. SPY buy and hold returned 73.4% at 21% volatility in the test period
Appendix 4: portfolio components
Note: for Portfolio 22, I now recommend replacing semiconductors with FXF, the Swiss Franc ETF, as it provides diversification into non-equity, non-commodity, non-Treasury dimension
Other candidates to consider:
GUR: S&P SPDR for emerging Europe
EEB: BRIC countries
EMB: emerging markets bonds