![METHODOLOGYSelect target market [forex], vehicles [7 major currency pairs] and time frame [daily]Gather publicly available trading strategies [63 total]Common usageAmmermann, Conceicao (2010)CSULB Finance Professor and CSULB AlumBollinger (2002)Elder (1993)Katz (2000)Leip (2010)Systems and indicators I createdMurphy (1999)Pruitt, Hill (2003)](https://image.slidesharecdn.com/srcpresentation-technicaltradingsystemsinforex-110225183200-phpapp02/85/Src-presentation-technical-trading-systems-in-forex-21-320.jpg)
Src presentation technical trading systems in forex
- 1. Brian LeipProfitability of automated technical trading systems in the foreign exchange marketCollege of Business AdministrationFinance Major, CBA Honors ProgramSupervisor: Dr. Pamela Miles Homer
- 2. OUTLINEPurpose of the StudyIntroductionBackground and Literature ReviewMethodologyHypothesesResults
- 3. PURPOSE OF THE STUDYExamine the profitability of 63 publicly available Technical Trading Systems (TTS)Which systems work the best and why?Currency pairsTechnical indicatorsExit techniqueComplexity/sophistication
- 4. PURPOSE OF THE STUDYExpand the understanding of Technical AnalysisTechnical analysis - “the study of market action, primarily through the use of charts, for the purpose of forecasting future price trends” (Murphy, 1999)TA historically disregarded by academia due to its conflict with the efficient market hypothesis (EMH)Sparse coverage of TA at the college levelTA is widely used by practitioners creating a large gap between “the classroom” and “the street”
- 5. INTRODUCTION – Efficient Market Hypothesis (EMH)What is the Efficient Market Hypothesis (EMH)?Dominant paradigm in financial theory from the 60s to the 90sMarkets created by hyper-rational decision makersNo one can beat the market except by luck or by taking on riskTherefore TA and FA should not work
- 6. INTRODUCTION – Efficient Market Hypothesis (EMH)Prominent CriticsWarren Buffett - InvestorJohn Maynard Keynes – EconomistRobert Haugen – Professor at UC IrvinePaul McCulley – Managing director at PIMCOThe field of behavioral financeAll practitioners using fundamental and/or technical analysis Universities now expanding beyond EMH but still very few classes on Technical AnalysisSkepticism of TA slow to change
- 7. INTRODUCTION – Technical AnalysisPurpose of Technical AnalysisTo capitalize on market inefficiencies (e.g. trends)Types of TA MethodsQualitative – Charting (pattern finding)Quantitative – Technical indicators, trading systemsUsed in this study
- 8. INTRODUCTION – Technical AnalysisTenets of Technical AnalysisMarket action (price and volume data) efficiently summarizes all microeconomic, macroeconomic and behavioral informationPrices move in trendsHistory repeats itself
- 9. INTRODUCTION – Chart with Technical Indicators
- 10. INTRODUCTION – The Foreign Exchange MarketFloating Rate Foreign Exchange MarketAlso called forex or FX marketBegan in early 70sAll currencies are quoted in relation to anotherEURUSD = Price of the Euro in relation to the US DollarTo buy one currency you must also sell anotherDaily turnover = 2 trillion USDSeveral times greater than all stock exchanges in the world combinedTA used extensively in forex market
- 11. INTRODUCTION – The Foreign Exchange MarketTA use in the foreign exchange (forex) market
- 12. Taylor and Allen (1992)
- 13. 90% of UK forex respondents use some form of TA
- 14. Cheung & Chinn (2001)
- 15. 30% of US forex market practitioners would best describe themselves as TA tradersSource: Cheung & Chinn (2001)
- 16. BACKGROUND AND LITERATURE REVIEWEarly Studies – TA in the Stock Market (60s & 70s)Widely cited studies from the 60s find TA to be unprofitableFama and Blume (1966)Van Horne and Parker (1967, 1968)Jensen and Benington (1970)Fama declares TA to be a futile undertaking (1970)Note: Fama is the founder of the Efficient Market Hypothesis
- 17. BACKGROUND AND LITERATURE REVIEWEarly Studies – TA in the Forex Market (60s, 70s & 80s)In contrast, TA studies in FX market generally found sizable net profitsPoole (1967)Dooley and Shafer (1976)Logue and Sweeney (1977)Logue, Sweeney and Willett (1978)Cornell and Dietrich (1978)Dooley and Shafer (1983)Sweeney (1986)Schulmeister (1987)Shortcomings in study methodologies
- 18. BACKGROUND AND LITERATURE REVIEWModern StudiesAddress shortcomings found in early studiesMixed results on profitability56 of 95 (59%) - positive returns20 of 95 (21%) - negative returns19 of 95 (20%) – mixed resultsSource: Park, Irwin (2007)
- 19. BACKGROUND AND LITERATURE REVIEWSurvey of literature for TA in forex marketMenkhoff, Taylor (2007)Review of 44 academic studiesConclusionBeyond question that TA may be used to provide very high returnsTA is an intrinsic part of the forex marketFor researchers, this means TA must be understood and integrated into economic reasoningFor practitioners, TA strategies must be constantly evaluated as potentially important tools
- 20. METHODOLOGYWhat makes my study unique?Number of Technical Trading Systems – 63To my knowledge, 23 is the max in other studiesTime frame1975 – 2010 (35 years)Results geared towards finding the best technical trading systems rather than refuting EMH
- 21. METHODOLOGYSelect target market [forex], vehicles [7 major currency pairs] and time frame [daily]Gather publicly available trading strategies [63 total]Common usageAmmermann, Conceicao (2010)CSULB Finance Professor and CSULB AlumBollinger (2002)Elder (1993)Katz (2000)Leip (2010)Systems and indicators I createdMurphy (1999)Pruitt, Hill (2003)
- 22. METHODOLOGYProgram the 63 Strategies into TradeStation
- 23. METHODOLOGYBifurcate the available data and run optimizations on recent ½ (in-sample) to generate optimal inputs for the strategy63 strategies * 7 currency pairs = 441 optimizations69,030 average tests * 441 = 30,442,230 total tests
- 24. METHODOLOGYOrganize the results and apply a scoring metric to all testsSelect the top performing test from each of the 441 optimizations. Apply to the older ½ of the data (out-of-sample) and save resultsGather the 441 out-of-sample results and analyze
- 25. METHODOLOGYTradeStation Software PlatformGold standard for rule based tradingExpensive - $100/monthRecipient of numerous awards
- 26. HYPOTHESESH1 –Technical Trading Systems will have out-of-sample excess profits that cannot be accounted for by the bearing of riskH2 – The more complex Technical Trading Systems will outperform less complex onesExcess profit testSharpe Ratio – Calculates excess returns over the risk free rate
- 27. RESULTSAvg. Sharpe Ratio by Strategy
- 28. RESULTSAvg Sharpe Ratio by Complexity (1=simple, 5=complex)
- 29. HYPOTHESES – Expected ResultsH1 – Sharpe ratios indicate that Technical Trading Systems have out-of-sample excess profits that cannot be accounted for by the bearing of riskH2 – Sharpe ratios indicate a clear link between excess risk adjusted returns and complexity
- 30. APPENDIX - RESULTSAvg. Sharpe Ratio by Currency
- 31. APPENDIX - RESULTSAvg. Sharpe Ratio by Technical Indicator
- 32. APPENDIX - RESULTSAvg. Sharpe Ratio by Exit Type
- 33. APPENDIX - RESULTSAvg. Sharpe Ratio by Source
- 34. Question & AnswerThank you for your time
Editor's Notes
- #11: Pairs – cannot only buy USD. You must buy USD and also sell another currency like JPY
- #12: Prior = 5 years prior to the 2001 studyTA expanded from 19% (least of the 4) to 30% (greatest of the 4)
- #20: Optimizing on ½ of the data is done to prevent “curve fitting” or data snooping biasThe most time consuming part of the processOptimizations take between 1 hour to multiple daysTook over 6 months with 5 computers working simultaneously 24/7
- #24: 27 out of 63 have positive sharpe ratios (43%)Comparable to 7 out of 23 (30%) in Likac and Brorsen (1990)
- #27: 3 of 7 currency pairs have excess return relative to risk (43%)Interesting that GBPJPY is the only unprofitable one. It is regarded as the most volatile and therefore should have most trading opportunitiesHigher transaction costs may be a factor
- #28: Channels and bands performed the bestMean reversion – RSI and Stochastic performed worst