The document provides an overview and review of the Ultra Fast GP Evolved Finstructure trading strategy designed by the Trading System Lab (TSL). Key points:
- TSL uses genetic programming to automatically design trading strategies that are transparent, unique, robust, fast, portable, and diverse.
- Strategies are tested out-of-sample during the design process to avoid overfitting.
- TSL strategies have outperformed other tracked systems and can be implemented across different platforms and languages.
- The evolutionary approach designs strategies an order of magnitude faster than manual methods in minutes rather than days or months.
Today’s trading is complex and frequently involves little human intervention. Five years after the "Flash Crash," do you know how high frequency trading and dark pools work? Our new report separates fact from fiction.
In The Speed Traders, Edgar Perez, founder of the prestigious business networking community Golden Networking, opens the door to the secretive world of high-frequency trading (HFT). Inside, prominent figures of HFT drop their guard and speak with unprecedented candidness about their trade.
Op Risk High Frequency Trading June 14 Finaltestytre
Presentation on High Frequency Trading risks delivered during OpRisk conference in London in June 2012. Content includes an overview of key risks affecting high frequency trading.
1. Failure to meet regulatory and exchange requirements.
2. Removal of human decision making once the algorithms are finished.
3. Extreme market behaviour: Flash Crash (2010).
4. Theft or loss of Intellectual Property.
5. Errors or problems suffered by clients using Direct Market Access and Algo/HFT.
6. Business impact of latency (system errors may increase delays).
7. Limited security controls at the infrastructure level.
8. Failure of hedges. 9. Incorrect/untested strategies.
David Ramirez
IT Audit Director
HackerEarth is pleased to announce its next session to help you understand what it really takes to become a data scientist.
Agenda of this session will include answers to the following questions:
- Why is it the best time to take up Data Science as a career?
- How can you take the first step in Data Science? (After all, first step is always the hardest!)
- How can you become better and progress fast?
- How is life after becoming a Data Scientist?
Speaker:
Jesse Steinweg-Woods is soon-to-be a Senior Data Scientist at tronc, working on recommender systems for articles and understanding customer behavior. Previously, he worked at Argo Group Insurance on new pricing models that took advantage of machine learning techniques. He received his PhD in Atmospheric Science from Texas A&M University, and his research focused on numerical weather and climate prediction.
Today’s trading is complex and frequently involves little human intervention. Five years after the "Flash Crash," do you know how high frequency trading and dark pools work? Our new report separates fact from fiction.
In The Speed Traders, Edgar Perez, founder of the prestigious business networking community Golden Networking, opens the door to the secretive world of high-frequency trading (HFT). Inside, prominent figures of HFT drop their guard and speak with unprecedented candidness about their trade.
Op Risk High Frequency Trading June 14 Finaltestytre
Presentation on High Frequency Trading risks delivered during OpRisk conference in London in June 2012. Content includes an overview of key risks affecting high frequency trading.
1. Failure to meet regulatory and exchange requirements.
2. Removal of human decision making once the algorithms are finished.
3. Extreme market behaviour: Flash Crash (2010).
4. Theft or loss of Intellectual Property.
5. Errors or problems suffered by clients using Direct Market Access and Algo/HFT.
6. Business impact of latency (system errors may increase delays).
7. Limited security controls at the infrastructure level.
8. Failure of hedges. 9. Incorrect/untested strategies.
David Ramirez
IT Audit Director
HackerEarth is pleased to announce its next session to help you understand what it really takes to become a data scientist.
Agenda of this session will include answers to the following questions:
- Why is it the best time to take up Data Science as a career?
- How can you take the first step in Data Science? (After all, first step is always the hardest!)
- How can you become better and progress fast?
- How is life after becoming a Data Scientist?
Speaker:
Jesse Steinweg-Woods is soon-to-be a Senior Data Scientist at tronc, working on recommender systems for articles and understanding customer behavior. Previously, he worked at Argo Group Insurance on new pricing models that took advantage of machine learning techniques. He received his PhD in Atmospheric Science from Texas A&M University, and his research focused on numerical weather and climate prediction.
These are the slides from my 4Developers 2017 talk. You can find the recording here: https://www.youtube.com/watch?v=pU0VRMqM5vs. All my other talks can be found here: https://train-it.eu/resources.
---
That talk will present the C++ world seen from Low Latency domain. The world where no dynamic allocations are welcomed, C++ exceptions are nearly not used, where STL containers are often not enough, and where developers often need to go deep down to assembly level to verify if the code really does its best.
EXANTE's lecture at Stockholm School of Economics in Riga.
– Objectives of algorithmic trading
– Various types of algorithms
– The process of creating one
– Testing and evaluation
– Understanding the possible pitfalls (and solutions)
The presentation is part of a conference conducted by QuantInsti Quantitative Learning Pvt. Ltd. along with a multinational investment banking firm that engages in global investment banking, securities, investment management, and other financial services.
Indian Exchanges follow the Price-Time Priority principle in Limit Order Books. This allows for quantification of the costs that a trader is willing to pay or receive in order to Trade. In order to gain price priority, the cost is in terms of ticks paid to gain price precedence over others. In terms of Time priority, it is atleast one tick, that allows the trader's order to leapfrog others at the same price level.
You can find a detailed version of this presentation on our blog - http://www.quantinsti.com/blog/empirical-analysis-of-limit-order-books/
Connect with us:
Facebook - http://facebook.com/quantinsti
Twitter - http://twitter.com/quantinsti
Youtube - http://youtube.com/quantinsti
Order book dynamics in high frequency tradingQuantInsti
An important task of high-frequency trading is to successfully capture the dynamics in the Data. Empirical Data on Indian Exchanges show that 95% of all NEW orders are placed within 5 ticks of best-bid and best-ask. The Quantinsti® Replacement Matrix shows that most of the orders that are being replaced are among the top 3 levels and these replacements allow us to visualize and generalize about market behaviour. This matrix gives a visual representation of the cost metrics and replacement behaviour.
Execution Algorithms provide a price which is between Limit Order Execution and Market Order Execution. Market Orders guarantee execution within a certain time but the price that it may get the trader remains uncertain. Limit Order guarantees the price but it may remain un-executed if price moves away. Most Execution Algorithms balance between these two order types.
The speaker, Mr. Gaurav Raizada, discusses Quantinsti® Replacement Matrix in the webinar along with basics on order book management theory for high frequency traders.
Trading in financial markets today is dominated by automated trading across most asset classes, but current programs are implemented using structured programming approaches which are static and represent a snapshot of the authors ideas, biases, and shortcomings at the time of implementation. Building automated trading bots that can learn from experience and can adapt to changing market conditions is changing the landscape and will deeply change trading as we know it.
In this presentation we will explore the history of automated trading, the environment in which these programs operate, current state, and challenges of the current approach. We will explore how a machine learning approach can be applied to automated trading and the forces driving this transformation. Analysis, which used to take hours or days, can now be done in seconds, back-testing over a larger length of time with fuller data now possible, and more data sources are available that can be used to build richer more accurate models.
Speaker
Diego Baez, GM Financial Services, Hortonworks
Презентация Сергея Трошина и Антона Антонова из EXANTE об алгоритмической торговле, инфраструктуре брокера и автоматизации торговли через FIX-протокол.
The slides cover the topics of algorithmic trading, broker IT infrastructure and trading via FIX protocol. Prepared by Sergey Troshin and Anton Antonov, EXANTE Ltd.
Theory of High Frequency Trading Systems TestingIosif Itkin
20-th of October. Software Development & Analysis Technologiesin Auditorium Seminar in Lomonosov Moscow State University
High Frequency Trading Systems. How they influece the market. Exchange systems requirements and load modelling
Data Structures for High Resolution, Real-time Telemetry at ScaleScyllaDB
The challenge within telemetry in real-time systems is that you need as many sources of telemetry as possible (Throughput, latency, Errors, CPU, and many more... ) but you can't pay for extra overhead when our users are expecting sub-ms ops that scale to millions of transactions per second.
In this talk, we'll describe how we're using and improving several OSS data structures to incorporate telemetry features at scale, and showcase why they do matter on scenarios in which we have Performance/Security/Ops issues.
Slides from my presentation on the Augur decentralized prediction market for the Blockchain Smart Contracts - Seattle Working Group Meetup on 07/23. The slides provide an overview of prediction markets, the benefits and challenges of creating one which operates in a decentralized manner, Augur's different market stages and their functions, and the risks and incentives of the Augur system.
[Data Meetup] Data Science in Finance - Building a Quant ML pipelineData Science Society
Georgi Kirov shares a common market-neutral statistical arbitrage framework. It will help showcase the many different ways to structure a systematic research project. From data reconciliation and signal backtesting to optimization and execution, what are some principled ways to evaluate and compare ML ideas? This process inevitably depends on the characteristics of a specific strategy, for instance, if it is liquidity-taking or liquidity-making.
UNDERSTANDING WHAT GREEN WASHING IS!.pdfJulietMogola
Many companies today use green washing to lure the public into thinking they are conserving the environment but in real sense they are doing more harm. There have been such several cases from very big companies here in Kenya and also globally. This ranges from various sectors from manufacturing and goes to consumer products. Educating people on greenwashing will enable people to make better choices based on their analysis and not on what they see on marketing sites.
These are the slides from my 4Developers 2017 talk. You can find the recording here: https://www.youtube.com/watch?v=pU0VRMqM5vs. All my other talks can be found here: https://train-it.eu/resources.
---
That talk will present the C++ world seen from Low Latency domain. The world where no dynamic allocations are welcomed, C++ exceptions are nearly not used, where STL containers are often not enough, and where developers often need to go deep down to assembly level to verify if the code really does its best.
EXANTE's lecture at Stockholm School of Economics in Riga.
– Objectives of algorithmic trading
– Various types of algorithms
– The process of creating one
– Testing and evaluation
– Understanding the possible pitfalls (and solutions)
The presentation is part of a conference conducted by QuantInsti Quantitative Learning Pvt. Ltd. along with a multinational investment banking firm that engages in global investment banking, securities, investment management, and other financial services.
Indian Exchanges follow the Price-Time Priority principle in Limit Order Books. This allows for quantification of the costs that a trader is willing to pay or receive in order to Trade. In order to gain price priority, the cost is in terms of ticks paid to gain price precedence over others. In terms of Time priority, it is atleast one tick, that allows the trader's order to leapfrog others at the same price level.
You can find a detailed version of this presentation on our blog - http://www.quantinsti.com/blog/empirical-analysis-of-limit-order-books/
Connect with us:
Facebook - http://facebook.com/quantinsti
Twitter - http://twitter.com/quantinsti
Youtube - http://youtube.com/quantinsti
Order book dynamics in high frequency tradingQuantInsti
An important task of high-frequency trading is to successfully capture the dynamics in the Data. Empirical Data on Indian Exchanges show that 95% of all NEW orders are placed within 5 ticks of best-bid and best-ask. The Quantinsti® Replacement Matrix shows that most of the orders that are being replaced are among the top 3 levels and these replacements allow us to visualize and generalize about market behaviour. This matrix gives a visual representation of the cost metrics and replacement behaviour.
Execution Algorithms provide a price which is between Limit Order Execution and Market Order Execution. Market Orders guarantee execution within a certain time but the price that it may get the trader remains uncertain. Limit Order guarantees the price but it may remain un-executed if price moves away. Most Execution Algorithms balance between these two order types.
The speaker, Mr. Gaurav Raizada, discusses Quantinsti® Replacement Matrix in the webinar along with basics on order book management theory for high frequency traders.
Trading in financial markets today is dominated by automated trading across most asset classes, but current programs are implemented using structured programming approaches which are static and represent a snapshot of the authors ideas, biases, and shortcomings at the time of implementation. Building automated trading bots that can learn from experience and can adapt to changing market conditions is changing the landscape and will deeply change trading as we know it.
In this presentation we will explore the history of automated trading, the environment in which these programs operate, current state, and challenges of the current approach. We will explore how a machine learning approach can be applied to automated trading and the forces driving this transformation. Analysis, which used to take hours or days, can now be done in seconds, back-testing over a larger length of time with fuller data now possible, and more data sources are available that can be used to build richer more accurate models.
Speaker
Diego Baez, GM Financial Services, Hortonworks
Презентация Сергея Трошина и Антона Антонова из EXANTE об алгоритмической торговле, инфраструктуре брокера и автоматизации торговли через FIX-протокол.
The slides cover the topics of algorithmic trading, broker IT infrastructure and trading via FIX protocol. Prepared by Sergey Troshin and Anton Antonov, EXANTE Ltd.
Theory of High Frequency Trading Systems TestingIosif Itkin
20-th of October. Software Development & Analysis Technologiesin Auditorium Seminar in Lomonosov Moscow State University
High Frequency Trading Systems. How they influece the market. Exchange systems requirements and load modelling
Data Structures for High Resolution, Real-time Telemetry at ScaleScyllaDB
The challenge within telemetry in real-time systems is that you need as many sources of telemetry as possible (Throughput, latency, Errors, CPU, and many more... ) but you can't pay for extra overhead when our users are expecting sub-ms ops that scale to millions of transactions per second.
In this talk, we'll describe how we're using and improving several OSS data structures to incorporate telemetry features at scale, and showcase why they do matter on scenarios in which we have Performance/Security/Ops issues.
Slides from my presentation on the Augur decentralized prediction market for the Blockchain Smart Contracts - Seattle Working Group Meetup on 07/23. The slides provide an overview of prediction markets, the benefits and challenges of creating one which operates in a decentralized manner, Augur's different market stages and their functions, and the risks and incentives of the Augur system.
[Data Meetup] Data Science in Finance - Building a Quant ML pipelineData Science Society
Georgi Kirov shares a common market-neutral statistical arbitrage framework. It will help showcase the many different ways to structure a systematic research project. From data reconciliation and signal backtesting to optimization and execution, what are some principled ways to evaluate and compare ML ideas? This process inevitably depends on the characteristics of a specific strategy, for instance, if it is liquidity-taking or liquidity-making.
UNDERSTANDING WHAT GREEN WASHING IS!.pdfJulietMogola
Many companies today use green washing to lure the public into thinking they are conserving the environment but in real sense they are doing more harm. There have been such several cases from very big companies here in Kenya and also globally. This ranges from various sectors from manufacturing and goes to consumer products. Educating people on greenwashing will enable people to make better choices based on their analysis and not on what they see on marketing sites.
Willie Nelson Net Worth: A Journey Through Music, Movies, and Business Venturesgreendigital
Willie Nelson is a name that resonates within the world of music and entertainment. Known for his unique voice, and masterful guitar skills. and an extraordinary career spanning several decades. Nelson has become a legend in the country music scene. But, his influence extends far beyond the realm of music. with ventures in acting, writing, activism, and business. This comprehensive article delves into Willie Nelson net worth. exploring the various facets of his career that have contributed to his large fortune.
Follow us on: Pinterest
Introduction
Willie Nelson net worth is a testament to his enduring influence and success in many fields. Born on April 29, 1933, in Abbott, Texas. Nelson's journey from a humble beginning to becoming one of the most iconic figures in American music is nothing short of inspirational. His net worth, which estimated to be around $25 million as of 2024. reflects a career that is as diverse as it is prolific.
Early Life and Musical Beginnings
Humble Origins
Willie Hugh Nelson was born during the Great Depression. a time of significant economic hardship in the United States. Raised by his grandparents. Nelson found solace and inspiration in music from an early age. His grandmother taught him to play the guitar. setting the stage for what would become an illustrious career.
First Steps in Music
Nelson's initial foray into the music industry was fraught with challenges. He moved to Nashville, Tennessee, to pursue his dreams, but success did not come . Working as a songwriter, Nelson penned hits for other artists. which helped him gain a foothold in the competitive music scene. His songwriting skills contributed to his early earnings. laying the foundation for his net worth.
Rise to Stardom
Breakthrough Albums
The 1970s marked a turning point in Willie Nelson's career. His albums "Shotgun Willie" (1973), "Red Headed Stranger" (1975). and "Stardust" (1978) received critical acclaim and commercial success. These albums not only solidified his position in the country music genre. but also introduced his music to a broader audience. The success of these albums played a crucial role in boosting Willie Nelson net worth.
Iconic Songs
Willie Nelson net worth is also attributed to his extensive catalog of hit songs. Tracks like "Blue Eyes Crying in the Rain," "On the Road Again," and "Always on My Mind" have become timeless classics. These songs have not only earned Nelson large royalties but have also ensured his continued relevance in the music industry.
Acting and Film Career
Hollywood Ventures
In addition to his music career, Willie Nelson has also made a mark in Hollywood. His distinctive personality and on-screen presence have landed him roles in several films and television shows. Notable appearances include roles in "The Electric Horseman" (1979), "Honeysuckle Rose" (1980), and "Barbarosa" (1982). These acting gigs have added a significant amount to Willie Nelson net worth.
Television Appearances
Nelson's char
Natural farming @ Dr. Siddhartha S. Jena.pptxsidjena70
A brief about organic farming/ Natural farming/ Zero budget natural farming/ Subash Palekar Natural farming which keeps us and environment safe and healthy. Next gen Agricultural practices of chemical free farming.
WRI’s brand new “Food Service Playbook for Promoting Sustainable Food Choices” gives food service operators the very latest strategies for creating dining environments that empower consumers to choose sustainable, plant-rich dishes. This research builds off our first guide for food service, now with industry experience and insights from nearly 350 academic trials.
DRAFT NRW Recreation Strategy - People and Nature thriving together
Tsl version 1.1_review
1. ULTRA FAST GP EVOLVED FINSTRUCTUREULTRA FAST GP EVOLVED FINSTRUCTURE
FROM UNSTRUCTURED GOOFROM UNSTRUCTURED GOO
TSL Version 1.1 ReviewTSL Version 1.1 Review
Michael L. Barna, CTAMichael L. Barna, CTA
Trading System LabTrading System Lab
PLEASE PUT YOUR PHONE ON MUTEPLEASE PUT YOUR PHONE ON MUTE
03/24/1703/24/17
2. REQUIRED DISCLAIMERREQUIRED DISCLAIMER
HYPOTHETICAL PERFORMANCE RESULTS HAVE MANY INHERENT LIMITATIONS, SOME OF WHICH ARE DESCRIBED
BELOW. NO REPRESENTATION IS BEING MADE THAT ANY ACCOUNT WILL OR IS LIKELY TO ACHIEVE PROFITS OR
LOSSES SIMILAR TO THOSE SHOWN.
IN FACT, THERE ARE FREQUENTLY SHARP DIFFERENCES BETWEEN HYPOTHETICAL PERFORMANCE RESULTS AND
THE ACTUAL RESULTS ACHIEVED BY ANY PARTICULAR TRADING PROGRAM. ONE OF THE LIMITATIONS OF
HYPOTHETICAL PERFORMANCE RESULTS IS THAT THEY ARE GENERALLY PREPARED WITH THE BENEFIT OF
HINDSIGHT. IN ADDITION, HYPOTHETICAL TRADING DOES NOT INVOLVE FINANCIAL RISK, AND NO HYPOTHETICAL
TRADING RECORD CAN COMPLETELY ACCOUNT FOR THE IMPACT OF FINANCIAL RISK IN ACTUAL TRADING. FOR
EXAMPLE, THE ABILITY TO WITHSTAND LOSSES OR TO ADHERE TO A PARTICULAR TRADING PROGRAM IN SPITE OF
TRADING LOSSES ARE MATERIAL POINTS WHICH CAN ALSO ADVERSELY AFFECT ACTUAL TRADING RESULTS.
THERE ARE NUMEROUS OTHER FACTORS RELATED TO THE MARKETS IN GENERAL OR TO THE IMPLEMENTATION OF
ANY SPECIFIC TRADING PROGRAM WHICH CANNOT BE FULLY ACCOUNTED FOR IN THE PREPARATION OF
HYPOTHETICAL PERFORMANCE RESULTS AND ALL OF WHICH CAN ADVERSELY AFFECT ACTUAL TRADING RESULTS.
03/24/1703/24/17
3. What is TSL?
• TSL is a platform for the automatic design of Trading
Strategies
• Code is written for you
• Trading Strategies are designed for you
• Strategies are tested OOS during design
• Written by traders (and machine learning scientists, mathematicians, engineers, programmers)
for traders
• TSL is fast, very fast
• TSL’s strategies are #1 as tracked by Futures Truth
• TSL has clients in 7 countries
03/24/1703/24/17
4. What Does TSL Emphasize?
1. Transparency
2. Uniqueness
3. Robustness
4. Speed
5. Portability
6. Diversity
03/24/1703/24/17
5. Transparency
• TSL systems can be “White Boxs”
• TSL systems show all forward and backtested trades
• Systems may be “turned off” at any time
• Risk/Stop Loss is completely transparent
• Poor Blind Tested Systems are not deployed
03/24/1703/24/17
6. Uniqueness
• TSL discovers new mathematical relationships
• Uses traditional or non traditional information
• Stochastic variability in solutions
• Simple to Complex behavior may be studied
• Fundamentals may be merged with Technicals
• Single or Multiple Market-Models deployed
03/24/1703/24/17
7. Robustness
(Over Fit Avoidance)
• Blind Testing (walk forward)
• Run Path Logs (path intelligence)
• Unbiased Terminal Set (directionless inputs)
• Multi-Run, Randomized Criteria (global optimum)
• Zero Point Origin (no predefined initial point)
• Parsimony Pressure (Occam’s razor)
03/24/1703/24/17
8. SpeedSpeed
• Higher Level Languages are too slow
• Targeted Fitness is an ENTIRE Trading System
• 20,000-100,000 Trading Systems needed/run
• 2.3 Million System-Bars/Second*
• <1 minute (10 yrs EOD)
• ~2-4 minutes (20 yrs EOD)
• ~5-30 minutes for daytrading system design
• Large Terminal Sets are accessed
• Large Function Sets are accessed
Intel Core i7 975 15% O.C. Single Thread. N-2NX Core Speedup Possible
Note: this is ~2 orders of magnitude faster than manual evaluators
03/24/1703/24/17
9. Portability
• Evolves code in:
– C
– C#
– JAVA
– ASSEMBLER
– TRADE STATION
– WEALTH LAB
– Other Languages through Translators
• Easily to implement in standard trading platforms
• Simulator results may be used for further analysis
• QuantHouse, TradeStation, WealthLab, Trading Blox, Apollo, others
03/24/1703/24/17
10. Diversity
• TSL Allows Diversification:
• Over Markets
• Over Systems
• Over Time Frames
• GP may assemble disparate blocks
03/24/1703/24/17
11. What do I get with TSL?What do I get with TSL?
• Unlimited number of systems*
• Unlimited number of time frames
• Unlimited number of markets
• Large selection of languages
• Fastest available speed
• Huge research capability
*The user builds the systems
03/24/1703/24/17
12. PARADIGM SHIFT*
• Machines taking on increasing roles on Wall Street
• Machine designed strategies are outperforming human designs*
• Machines are critical for HF trading
• HF Machine Trading is the most profitable category on Wall Street
• Machines allow low Latency, Alpha research and generation
• Machine inefficient markets exist in emerging countries
• Machines will mine world markets for alpha in the future
• Machines that are faster and more efficient are in high demand
• Customers demand more analysis capabilities in finance and trading
*First time machine designed systems took top ranking in third party blind testing
03/24/1703/24/17
13. TRADING AND ALGORITHMSTRADING AND ALGORITHMS
• In 2009, High Frequency Algorithmic trading accounts for 73% of all US equity trading
volume, but is conducted by only 2% of the firms
• The High Frequency Hedge Fund category is now the most profitable on Wall Street
• Futures and options:
- Easy to integrate algos.
- 20% of options volume expected to be algo by 2010
• CME quote time is now ~ 6 milliseconds
Clearly, the Futures markets are moving firmly towards Algorithmic trading as we have
seen in the equities
References:
Advanced Trading, Sept/Oct. 2009
http://en.wikipedia.org/wiki/Algorithmic_trading
http://www.informationweek.com/news/hardware/data_centers/showArticle.jhtml?articleID=219700577&cid=RSSfeed_IWK_All#
03/24/1703/24/17
14. TSL CRITICAL FEATURES
• Far faster and better than GA, tree GP’s or other AI
• Simultaneous Design/ Walk Forward during design time
• 56 inputs, full custom inputs possible
• 5 preprocessed fact sets
• 34 instruction sets, fully configurable
• 33 Fitness Functions, includes multi fitness
• 18 Entry tactics, including multi entry tactics
• 7 exit tactics, including GP adaptive stops and exits
• Risk/Size, Constant dollars embedded and evolved
• Money Management and Optional GP “f”
• Pairs, full hedge, partial hedge with evolved Money Management
• Portfolios, with evolved Money Management
• 17 Option Tactics and combinations (CRR-BTREE, Bjerk-Stens)
• Daytrading-custom ID preprocessing, DT entry/exit types
03/24/1703/24/17
15. TSL CRITICAL FEATURES
• 2.3 Million System-Bars/sec rate (measured Core i7 975)
• All time frames, overnight or daytrade
• HF, IF or LF systems
• C, C#, JAVA, ASSEMBLER, EL, WL, Blox, Others
• Max 20 markets in portfolio (Version 1.x)
• 60,000 bar limit EOD, ID systems (Version 1.x)
• 10,000 bar limit portfolios (Version 1.x)
03/24/1703/24/17
16. The Old Way of WritingThe Old Way of Writing
Trading StrategiesTrading Strategies
• Develop or use existing theory
• Use TA books, indicators, patterns, etc.
• Hand code the system
• Test, Optimize, Test, Optimize (curve fit?)
• Try to select best parameters
• Forward walk (maybe?)
• Implement and hope for the best
• Time of work flow: days to months
03/24/1703/24/17
17. The New Way of WritingThe New Way of Writing
Trading StrategiesTrading Strategies
• Select Market
• Run Strategy generation algorithm
• Observe Out of Sample performance
• Decide if adequate
• Implement
• Adjust, research, study, tweak, learn, etc.
• Time of work flow: minutes to hours
03/24/1703/24/17
18. MACHINE DESIGNED TRADINGMACHINE DESIGNED TRADING
SYSTEM TECHNOLOGYSYSTEM TECHNOLOGY
A Blend Of:
• Technical (and/or Fundamental) Analysis
• Data Mining
• Evolutionary Algorithms
• Trading Simulators
• HPC
Note: This is not possible with standard backtesting platforms
03/24/1703/24/17
19. REGISTER PROGRAM
GENOME STRING
Inputs-DNA(56) Outputs(8)
Outputs are used for basic and higher level learning:
1.Basic trading signals
2.Complex trading signals
3.Money management
4.Adaptive risk
5.Targets and stops
6.Optimal GP
03/24/1703/24/17
Function Set
GP Operators
20. CPU CYCLES
How many CPU cycles does it take to compute the following?
X = Y + Z
High level languages: 20 Clock cycles
CGPS(LAIMGP): 1 Clock cycle
So, TSL should be at least 20 times
faster than higher level languages
Reference: Efficient Evolution of Machine Code for CISC Architectures using blocks and Homologous Crossover,
Peter Nordin, Wolfgang Banzhaf, Frank Francone
03/24/1703/24/17
21. PREPROCESSING AND THE GOOPREPROCESSING AND THE GOO
((Numeric or BooleanNumeric or Boolean))
• Example Preprocessing:
? Close >= Close[1]
Result: 0 or 1 (True or False, natural language of machines)
Categories:
News: Machine readable
Market Stack Data
Volatilities
Short term intraday patterns
Support and Resistance
Intermediate term patterns
Long term patterns
Oscillators-OBOS
Filters and Indicators
Regression and deviations
Transforms
Channels
Intermarket/Fundamentals
Domain Expertise-Systems/Ind.
56 Total Inputs. Only a few will be
Selected and used in the final design
by the machine
03/24/1703/24/17
22. THE GOO
Numeric or Boolean
56 inputs + Date, O, H, L, C, Time
(Boolean data shown below)
03/24/1703/24/17
26. THE ARCHITECTURETHE ARCHITECTURE
Development followed no specific model since noDevelopment followed no specific model since no
specific model was readily availablespecific model was readily available
•*GPU/G80, EP or Many Core Implementations
•**VC++.NET, VB, C#, EL
•***Assembler, C, C++
Real Time
or
Static Data
Machine
Readable
News
Market
Stack data Trading
System
Code
(C#, C++,
JAVA, EL)
03/24/1703/24/17
27. ASYNCHRONOUS “WINDOW”
MACHINE LEARNING
EVOLVED CODE IS UPDATED BASED ON RT FITNESSEVOLVED CODE IS UPDATED BASED ON RT FITNESS
DATA STREAM
DATA PP
WINDOW
GP
CODE
EXECUTED
PERFORMANCE
FITNESS
ADAPTIVE
ADAPTIVE
TSL
GP
03/24/1703/24/17
28. FITNESS CAN BE MULTI GOALFITNESS CAN BE MULTI GOAL
Machine Design Allows Us to Adjust Critical System MetricsMachine Design Allows Us to Adjust Critical System Metrics
as Targeted Fitness Functionas Targeted Fitness Function
Net Profit
Drawdown
Percent Accuracy
Profit Factor
Average Trade
PRODUCES
CODE
03/24/1703/24/17
29. EVOLUTIONARY TRADEREVOLUTIONARY TRADER
DESIGNDESIGN
• Traders must trade profitably or they are deleted
• Profitable traders compete with other profitable traders
• Profitable traders are allowed to reproduce
• Some will be subject to random mutations
• Offspring will be subject to crossover
• Traders will be tested on Out Of Sample* continually
* Blind sample testing during run may reflect “direction” of process
03/24/1703/24/17
31. Summary of Evolutionary Algorithms
YEAR INVENTOR TECHNIQUE INDIVIDUAL
1958 Freidberg Learning Machine Virtual Assembler
1959 Samuel Mathematics Polynomial
1965 Fogel, Owens and Walsh Evolutionary Programming Automation
1965 Rechenberg, Schwefel Evolutionary Strategies Real-Numbered Vector
1975 Holland Genetic Algorithms Fixed Size Bit String
1978 Holland and Reitmann Genetic Classifier Systems Rules
1980 Smith Early Genetic Programming Var-Size Bit String
1985 Cramer Early Genetic Programming Tree
1986 Hicklin Early Genetic Programming LISP
1987 Fujiki and Dickinson Early Genetic Programming LISP
1987 Dickmanns, Schmidhuber and Winklhofer Early Genetic Programming Assembler
1992 Koza Genetic Programming Tree
03/24/1703/24/17
32. TSL GP MARKET
COMPARISON
*Speed may be proportional to inputs
**Million Input-System-Bars per Second
TSL-LAIMGP OTHER (GA/GP)
SPEED 129 MISBS** Way Slower
# Inputs* 56 5-?
# Functions 34 5-10
# Entries 18 1-5
# Exits 4 1-3
# Stops 3 1-3
Automatic OOS? YES NO
Daytrade? YES MAYBE
Pairs? YES NO
Portfolio/MM? YES MAYBE
Options? YES NO
# Preprocessors 5 1
H, Non H Cross? YES NO
Third Party Tracked? YES NO
03/24/1703/24/17
33. ES THEORETICAL TRADEES THEORETICAL TRADE
PERFORMANCEPERFORMANCE
For a 1 minute bar, Theoretical EV ~ tick size
However, the period alpha is inverse to interval
Bar Size, Minutes 405 135 60 15 1
Expected Value, $ 463 297 188 47 15
Total Trades/Day 0.6 0.75 1.6 5.7 77
Drawdown 8300 5225 4837 3062 1375
Period Alpha 277.8 222.75 300.8 267.9 1155
Reward/Risk 0.03 0.04 0.06 0.09 0.84
03/24/1703/24/17
34. MARKETS HAVE DIFFERENTMARKETS HAVE DIFFERENT
DESCRIPTIVE STATISTICSDESCRIPTIVE STATISTICS
So Why Design Symmetrical Systems?So Why Design Symmetrical Systems?
CME:E-MINI S&P CBOT:WHEAT
Power Spectral Density
Indicator Serial Correlation
Random Trend
03/24/1703/24/17
42. OPTIONS COMBINATIONS GP EVOLVEDOPTIONS COMBINATIONS GP EVOLVED
STRUCTURESSTRUCTURES
03/24/1703/24/17
43. SUMMARY
• TSL is patented, trademarked, licensed, restricted and exclusive
• TSL produces Daytrading, Pairs, Portfolios, Options and
Single Market Directional Systems
• TSL technology is unavailable anywhere else
• TSL strategies are #1, per Futures Truth
• TSL speed grows with CPU speed
• TSL can design many different types of strategies
• TSL produces code in a variety of languages
• TSL terminal set is customizable
• TSL sets a paradigm shift in strategy design
• TSL is used in many different countries and on many different
markets
03/24/1703/24/17
44. FUTURE PLANS
• Enhanced front and back end
• Strategic partnerships
• Hyper threaded simulator
• Version 2.x development
• Market depth strategy evolution
• Asynchronous: “evolution on the fly”
03/24/1703/24/17