Strategies for the Simple Timing Indicator (STI) and MACD to Improve Investment Performance AFTA October 18 & 22, 2011 Jim Gunn jimgunn@jgunnresearch.com
Disclaimer No Representations are made that the investment strategies presented herein or discussed by the presenter are complete, without error, or will produce desired investment results. This presentation is made for information sharing purposes only. Nor are any representations made that the investment strategies discussed have been employed to achieve successful investment results. Many of the results are based on back testing with historical data and simulated investing. Future investing results may significantly differ. You should perform your own independent research and Due Diligence prior to making any investment decisions, and invest accordingly. 10/22/2011 2
AAII Technical Indicator Quote AAII Computerized Investing, 4 th Qtr 2011, p1, Building AAII Stock Screens Article, by Wayne Thorp All of the stock screens AAII develops are based on our interpretation of an approach, often as it was outlined in a book, article or academic paper. For example, in this issue I finish my two-part discussion of Charles Carlson s formulas for selecting promising dividend-paying stocks. In developing this and other stock screens, I follow the author s approach as it is outlined in the source. However, many successful investors do not want to give away the farm when writing about their approach, so they are purposefully vague when outlining their search criteria. Therefore, there are times when we must use our investment expertise, and the wisdom of other market gurus, when deciphering what the author is trying to say. Lessons Learned Investments Strategies offered by Experts and others are often incomplete; sometimes purposefully or sometimes due to changing market conditions No Technical Indicator or combinations of Indicators or strategies are reliable over all time frames, selected investments, and market conditions Best investment successes will usually be achieved by developing and adapting our own information, strategies, due diligence, study, and self-training 10/22/2011 3
Presentation Topics Overview of Excel as Investment Tool Overview of use of STI and MACD Strategies Explanations Simple Timing Indicator (STI) Moving Average Convergence Divergence Indicator (MACD) Where indicators usually work and don t Human Chart Pattern Recognition Weekly Chart Examples Seasonality Analysis Interactive Charts of Results using Excel (Tuesday, as time permits) QQQ Powershares NASDAQ 100 Index ETF SPY S&P 100 Index ETF IJH ishares S&P 400 Mid Cap Index ETF IJR ishares S&P 600 Small Cap Index ETF AMZN Amazon EMA vs SMA Analysis Above and Audience Requests (Saturday at Baylor) Advanced Inputs Appreciated to load data base, if necessary (and possible) Limited others possible, but more time consuming to download data ~1200 Symbols in current data base 10/22/2011 4
References Appel, Gerald; Technical Analysis; MACD Chapter 8; 2005, FT Press Elder, Dr. Alexander; Come Into my Trading Room; 2002, Wiley Elder, Dr. Alexander, Presentation at Money Show Dallas, June 2011 Richards, Fred, STI discussions in various Presentations at Dallas Investors Forum, 2009-2010 Saddler, Rick; Swing Trade the T-Line Presentation, DFW AFTA, October 10, 2010 (T-Line is the 8 EMA of Close) 10/22/2011 5
Methodology and Tools Goals: Better understand Indicators and their utilization in investing Utilize Black Box Indicators, but better understand when not reliable, and adapt Study Historical Charts of Investments of Interest Carefully note when patterns are most reliable and least reliable and adapt Judiciously apply human chart pattern recognition abilities and when appropriate follow most reliable indicator(s) or combinations Use Excel for all charting, indicator calculations, financial analysis, back testing, step forward analysis, seasonality analysis, etc. Excel/Office 2010, released in June 2010 Can use other tools, such as stockcharts.com, yahoo finance, TC2000 etc. for daily strategy(s) execution, but these often have deficiencies for back testing, historical data (especially longer time frames), step forward analysis, etc. Many Stock Indicator How to papers often give brief discussions Excel examples often provided to demonstrate calculations Many indicators are old and based on pre-computer era computing and plotting resources Simplifications were often used to reduce calculations or plotting efforts Modern Computer and Communication technologies offer potential to enhance use of indicators 10/22/2011 6
PS QQQ (NASDAQ ETF) (last 2+ months) 60 58 8 EMA PowerShares QQQ Trust, Series 1 Price 20 EMA 50EMA 200EMA 8EMA 20EMA 50EMA 200EMA 56 54 52 50 48 7/31/11 8/7/11 8/14/11 8/21/11 8/28/11 9/4/11 9/11/11 9/18/11 9/25/11 10/2/11 10/9/11 10/16/11 10/23/11 QQQ Volume Vol Vol SMA 220 200 180 160 140 120 100 80 60 40 7/31/11 8/7/11 8/14/11 8/21/11 8/28/11 9/4/11 9/11/11 9/18/11 9/25/11 10/2/11 10/9/11 10/16/11 10/23/11 QQQ MACD (8,20,7) Histogram DIF Sig 1.5 1.0 0.5 0.0-0.5-1.0-1.5 MACD(8,20) = DIF = 8EMA 20EMA STI: Long when 8EMA >= 20EMA or equivalent when MACD >= 0 Histogram = DIF SIG -2.0 7/31/11 8/7/11 8/14/11 8/21/11 8/28/11 9/4/11 9/11/11 9/18/11 9/25/11 10/2/11 10/9/11 10/16/11 10/23/11 10/22/2011 7
Investing Rules Primary Indicators: MACD(8,20,7) of Close Long MACD > SIG; Short or Cash MACD < SIG Simple Timing Indicator Long 8EMA >= 20EMA or MACD >= 0 Short or Cash 8EMA < 20EMA or MACD < 0 Secondary Indicators T-Line: (Especially note support or resistance over time) Long: C i >= 8EMA (Support) Short or Cash C i < 8EMA (Resistance) Supporting/Confirming Indicators (Maintain Awareness) EMA trending Up (Bullish) or Down (Bearest) MACD trending up (Bullish) or Down (Bearest) MACD Histogram trending up (Bullish) or Down (Bearest) Money Management Limit loses: 6% - 8% 10/22/2011 8
Excel Block Diagram: Data, Storage, and Calculation ADO to Access Data Excel/2010/VBA Charting Indicator Calculations Control Panel(s) Active Symbol data, MA, MACD, & Financial Analysis Seasonality Analysis VBA Statement to download QQQ Workbooks.Open ("http://ichart.finance.yahoo.com/ table.csv?s=qqq") Access/2010 Over 1000 Symbols: Price: Open, High, Low, Close, Volume Data Storage Yahoo Web Stock Prices & Volume download 10/22/2011 9
Functionality Additions since AFTA Dec 2010 Presentations 1. Multipage Menus 2. Menu Setting STI & MACD parameters (e.g. 8 EMA or 10 SMA, etc.) 3. Support for Stock Splits 4. Enable Re-charting with select data update a. Sub-Time Frame Results in back testing b. Stock Splits 5. Weekly Charts 6. Menu Driven Daily Price Updates from Yahoo 7. Increased Symbols in Data Base a. S&P 500 b. NASDAQ (Market Capitalization > $900,000 c. Working Database for Indexes, ETFs, Mutual Funds, others, etc. 8. Step Forward Analysis 9. Seasonality Analysis 10/22/2011 10
Stock Price/Volume Data Yahoo Data Source www.finance.yahoo.com End of Day (EOD) Data EOD data typically fully available by ~9 10 pm on trading days Data Downloaded: Date, open, High, Low, Close, Volume, and Adjusted Close Only Adjusted Close accounts for splits and dividends Initial download of all historical data for a Symbol requires a special setup in automation Subsequent daily updates via automation Mutual Funds, indices, sectors, etc. ~1200 Symbols in current data base 8000+ Symbols can be accommodated in Access, but download time would be long 10/22/2011 11
Back Test Methodology Use End of Day (EOD, Close) Data Make next day s Long, Short, or Cash Decisions with EOD indicators Go Long, Cash, or Short with Open of next Trading Day Note: Can be more flexible with human intervention all discussions today will use next day open for buy/sell 10/22/2011 12
Exponential Moving Average of Close at time i EMA i EMA EMA C EMA of Close where C is the close on the latest date; C latest Trading Day minus 1 For EMA of other indicators, e.g. Volume, Use V i, instead of C i should be less than 1 and greater than 0 = (typical) or = where 20 is the length of a 20EMA Note: 20 SMA i = Observations EMAs typically take 2 3 times the period to stabilize For example, a 20EMA can require 40 to 60 historical calculations to converge Simple Moving Averages (SMA) may be employed, but EMAs respond faster and usually provide better performance. MA indicators generally are more reliable that MACD indicators, but catch market changes slower that MACDs. MA have whipsaws, but generally less frequent than MACD 10/22/2011 13
Moving Average Convergence Divergence MACD Indicator MACD(8,20) = EMA(8) EMA(20) often called DIFF or MACD(8,20) note when EMA(8) >= EMA(20) that MACD(8,20) >= 0, MACD is equivalent to STI as long as same EMAs are used. Sig(9) = EMA(MACD(8,20) ) often called SIG Histogram = MACD(8,20) SIG(9) often plotted with an MACD indicator plot Observations The MACD Timing Indicator is usually faster signal than the STI, but more whipsaws Appel clearly indicates that MACD is not always reliable and provides examples Apple in his book states that the MACD is most reliable when the DIF passes through 0 before signaling a change (e.g., long, cash, or short) Appel also indicates that the MACD will become unreliable due to whipsaws when price (usually close) trends in a narrow channel Elder states that divergence of price and the MACD is the best indicator in technical analysis 10/22/2011 14
QQQ (NASDAQ ETF) BH & STI Return (1/1/2000 Through 10/21/2011) 140 PowerShares QQQ Trust, Series 1 Price (Daily) 8EMA 20EMA 50EMA 200EMA 120 100 80 60 40 20 0 1/2/00 7/2/00 1/2/01 7/2/01 1/2/02 7/2/02 1/2/03 7/2/03 1/2/04 7/2/04 1/2/05 7/2/05 1/2/06 7/2/06 1/2/07 7/2/07 1/2/08 7/2/08 1/2/09 7/2/09 1/2/10 7/2/10 1/2/11 7/2/11 $220 $200 $180 $160 $140 $120 $100 $80 $60 $40 $20 $100K Long Shrt Cash QQQ: Simple TI $ Return BHRet: 40.43% ReT: 44.06% 1/2/00 7/2/00 1/2/01 7/2/01 1/2/02 7/2/02 1/2/03 7/2/03 1/2/04 7/2/04 1/2/05 7/2/05 1/2/06 7/2/06 1/2/07 7/2/07 1/2/08 7/2/08 1/2/09 7/2/09 1/2/10 7/2/10 1/2/11 7/2/11 QQQ MACD (8,20,7) Histogram DIF Sig 6 4 2 0-2 -4-6 -8 1/2/00 7/2/00 1/2/01 7/2/01 1/2/02 7/2/02 1/2/03 7/2/03 1/2/04 7/2/04 1/2/05 7/2/05 1/2/06 7/2/06 1/2/07 7/2/07 1/2/08 7/2/08 1/2/09 7/2/09 1/2/10 7/2/10 1/2/11 7/2/11 10/22/2011 15
QQQ (NASDAQ ETF) Back Test Results 2011 through 10/21 STI (8,20) MACH((8,20,7) YR BH LG/CH LG/SH LG/CH LG/SH 2000 39.31% 0.25% 42.23% 40.81% 47.11% 2001 33.56% 8.59% 11.86% 32.66% 42.01% 2002 38.41% 13.36% 15.46% 34.63% 33.89% 2003 47.49% 1.61% 31.23% 22.32% 0.80% 2004 8.89% 5.94% 2.65% 4.26% 1.17% 2005 0.80% 3.68% 8.39% 11.17% 21.96% 2006 6.17% 8.63% 8.04% 2.69% 11.63% 2007 17.86% 2.99% 20.69% 11.23% 3.86% 2008 41.99% 3.97% 38.85% 27.38% 15.26% 2009 53.78% 28.10% 5.86% 31.61% 11.46% 2010 17.55% 10.98% 4.56% 25.84% 32.97% 2011 3.91% 10.98% 16.62% 0.54% 3.84% Total 40.43% 14.08% 44.06% 50.40% 73.30% BH: Buy & Hold; STI: Simple Timing Indicator; LG: Long; CH: Cash; SH: Short Notes: Black Box (Mechanically) follow STI or MACD Indicators 10/22/2011 16
QQQ 2011 (Daily) Buy & Hold, Best Performance 60 PowerShares QQQ Trust, Series 1 Price 8EMA 20EMA 50EMA 200EMA 58 56 54 52 50 48 1/2/11 1/23/11 2/13/11 3/6/11 3/27/11 4/17/11 5/8/11 5/29/11 6/19/11 7/10/11 7/31/11 8/21/11 9/11/11 10/2/11 10/23/11 $100K Long Shrt Cash QQQ: MACD TI $ Return BHRet: 4.24% ReT: 0.85% $110 $108 $106 $104 $102 $100 $98 $96 $94 $92 $90 1/2/11 1/23/11 2/13/11 3/6/11 3/27/11 4/17/11 5/8/11 5/29/11 6/19/11 7/10/11 7/31/11 8/21/11 9/11/11 10/2/11 10/23/11 QQQ MACD (8,20,7) Histogram DIF Sig 1.5 1.0 0.5 0.0-0.5-1.0-1.5-2.0 1/2/11 1/23/11 2/13/11 3/6/11 3/27/11 4/17/11 5/8/11 5/29/11 6/19/11 7/10/11 7/31/11 8/21/11 9/11/11 10/2/11 10/23/11 Comparing MACD and STI, Long Short and Long Cash; and Buy and Hold 10/22/2011 17
QQQ 2010 (Daily) MACD Indicator Long Short, Best Performance 56 PowerShares QQQ Trust, Series 1 Price 8EMA 20EMA 50EMA 200EMA 54 52 50 48 46 44 42 40 12/27/09 1/17/10 2/7/10 2/28/10 3/21/10 4/11/10 5/2/10 5/23/10 6/13/10 7/4/10 7/25/10 8/15/10 9/5/10 9/26/10 10/17/10 11/7/10 11/28/10 12/19/10 $100K QQQ: MACD TI $ Return Long Shrt Cash BHRet: 17.55% ReT: 32.97% $140 $135 $130 $125 $120 $115 $110 $105 $100 $95 $90 12/27/09 1/17/10 2/7/10 2/28/10 3/21/10 4/11/10 5/2/10 5/23/10 6/13/10 7/4/10 7/25/10 8/15/10 9/5/10 9/26/10 10/17/10 11/7/10 11/28/10 12/19/10 QQQ MACD (8,20,7) Histogram DIF Sig 1.2 1.0 0.8 0.6 0.4 0.2 0.0-0.2-0.4-0.6-0.8-1.0-1.2-1.4 12/27/09 1/17/10 2/7/10 2/28/10 3/21/10 4/11/10 5/2/10 5/23/10 6/13/10 7/4/10 7/25/10 8/15/10 9/5/10 9/26/10 10/17/10 11/7/10 11/28/10 12/19/10 Comparing MACD and STI, Long Short and Long Cash; and Buy and Hold 10/22/2011 18
48 46 44 42 40 38 36 34 32 30 28 26 QQQ 2009 (Daily) Buy and Hold, Best Performance PowerShares QQQ Trust, Series 1 Price 8EMA 20EMA 50EMA 200EMA 24 12/28/08 1/18/09 2/8/09 3/1/09 3/22/09 4/12/09 5/3/09 5/24/09 6/14/09 7/5/09 7/26/09 8/16/09 9/6/09 9/27/09 10/18/09 11/8/09 11/29/09 12/20/09 $100K QQQ: MACD TI $ Return Long Shrt Cash $160 $150 $140 $130 $120 $110 $100 $90 $80 12/28/08 1/18/09 2/8/09 3/1/09 3/22/09 4/12/09 5/3/09 5/24/09 6/14/09 7/5/09 7/26/09 8/16/09 9/6/09 9/27/09 10/18/09 11/8/09 11/29/09 12/20/09 QQQ MACD (8,20,7) Histogram DIF Sig BHRet: 53.78% ReT: 31.61% 1.2 1.0 0.8 0.6 0.4 0.2 0.0-0.2-0.4-0.6-0.8-1.0-1.2 12/28/08 1/18/09 2/8/09 3/1/09 3/22/09 4/12/09 5/3/09 5/24/09 6/14/09 7/5/09 7/26/09 8/16/09 9/6/09 9/27/09 10/18/09 11/8/09 11/29/09 12/20/09 10/22/2011 Comparing MACD and STI, Long Short and Long Cash; and Buy and Hold 19
QQQ 2008 (Daily) STI, Long and Short, Best Performance 55 PowerShares QQQ Trust, Series 1 Price 8EMA 20EMA 50EMA 200EMA 50 45 40 35 30 25 12/30/07 1/20/08 2/10/08 3/2/08 3/23/08 4/13/08 5/4/08 5/25/08 6/15/08 7/6/08 7/27/08 8/17/08 9/7/08 9/28/08 10/19/08 11/9/08 11/30/08 12/21/08 $100K QQQ: Simple TI $ Return Long Shrt Cash BHRet: 41.99% ReT: 38.85% $180 $160 $140 $120 $100 $80 $60 $40 12/30/07 1/20/08 2/10/08 3/2/08 3/23/08 4/13/08 5/4/08 5/25/08 6/15/08 7/6/08 7/27/08 8/17/08 9/7/08 9/28/08 10/19/08 11/9/08 11/30/08 12/21/08 QQQ MACD (8,20,7) Histogram DIF Sig 1.5 1.0 0.5 0.0-0.5-1.0-1.5-2.0-2.5-3.0-3.5 12/30/07 1/20/08 2/10/08 3/2/08 3/23/08 4/13/08 5/4/08 5/25/08 6/15/08 7/6/08 7/27/08 8/17/08 9/7/08 9/28/08 10/19/08 11/9/08 11/30/08 12/21/08 10/22/2011 Comparing MACD and STI, Long Short and Long Cash; and Buy and Hold 20
Conclusions Simple Timing Indicator (STI), MACD, T-Line, and other Indictors Have varying Performances if Mechanically followed Time Range Market Conditions Specific Symbol (Company, ETF, Index, etc.) Human Chart Pattern Recognition abilities can improve performance Examples provided using MACD and T-Line Others are possible with Investor ingenuity and work Investment books Investment web searches Investment Seminars and Meetings Best Successes are undoubtedly achieved by: Studying Indicators and understanding strengths and weaknesses Devoting time to develop chart pattern recognition skills Books, presenters, etc. usually can only cover best and worst examples Many others patterns occur, including ones they have not yet observed Effective use of Money Management to minimize losses when strategies are unreliable and run with winners when reliable 10/22/2011 21
Excel VBA Book Walkenbach, John; Microsoft Excel 2010, Power Programing with VBA; Wiley, 2010 10/22/2011 22