Quantitative trading - page 22
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
Trading Alpha: Developing a Micro-Alpha Generating System | Algo Trading Conference
Trading Alpha: Developing a Micro-Alpha Generating System | Algo Trading Conference
In this webinar, the hosts introduce Dr. Thomas Stark, an esteemed expert in artificial intelligence and quantum computing from Sydney, Australia. Dr. Stark holds a PhD in physics and currently serves as the CEO of Triple A Trading, a renowned crop trading firm in Australia. With a background that includes previous work at proprietary trading firms, Rolls-Royce, and co-founding a microchip design company, Dr. Stark brings a wealth of knowledge and experience to the discussion.
The hosts begin by clarifying the concept of Alpha, which refers to independent returns in trading that are not influenced by market movements. They highlight the term "microalpha," which focuses on small trading strategies that contribute incrementally to trading success rather than producing extraordinary returns. While both concepts share the idea of independent returns, microalpha specifically emphasizes the importance of small strategies in achieving trading success.
Dr. Stark delves into the evolution of gold mining as an analogy for trading Alpha. He explains how gold mining methods have evolved from traditional nugget panning to large-scale mining operations that extract small amounts of gold from rocks. Similarly, trading for Alpha has also evolved, with many traditional strategies becoming overused and less effective due to arbitrage opportunities. Dr. Stark introduces the concept of micro Alpha development, which involves identifying systematic anomalies in the market that can be exploited for trading success. While machine learning plays a limited role in this process, manual work is required to identify exploitable inconsistencies. Dr. Stark believes that automation and backtesting can accelerate and enhance this process.
The speaker emphasizes the utilization of market inefficiencies to develop micro-alpha generating systems. These inefficiencies encompass various trading strategies such as pair strategies, trends, mean reversion, cross-correlation, chart patterns, and even machine learning techniques. The goal is to exploit these inefficiencies or strategies to generate systematic and reliable results. However, it is crucial to optimize these strategies without overfitting and combine them into a comprehensive trading strategy to create a complex yet effective system. Dr. Stark emphasizes the importance of understanding these different aspects to build a high-performing system.
Dr. Stark discusses the concept of exploiting trading anomalies and the significance of combining multiple trading strategies. While some traders may adopt unconventional methods like astrology, Dr. Stark emphasizes the need for creativity in constructing successful trading systems. However, combining strategies requires meticulous attention to detail, including precise timestamps and efficient programming. Traders must also consider the correlations and behaviors of individual strategies to ensure they complement each other and determine the optimal weighting of these systems.
The speaker highlights the importance of metrics when backtesting a trading strategy. They explain that studying a tear sheet with various metrics is crucial for understanding the unique characteristics of each individual strategy. There is no single most important or ideal metric, as different metrics apply to different use cases. For example, the Sharpe ratio may not be suitable for a strategy that trades infrequently but has high confidence in each trade. Metrics such as profit factor or Sortino ratio may be more appropriate in such cases. Additionally, the speaker emphasizes the significance of evaluating alpha and beta when assessing a system, ensuring that the system's beta is relatively low.
Different metrics for measuring the success of a trading strategy are discussed, including compounded annual growth return and drawdown. Dr. Stark emphasizes the importance of understanding all these metrics and developing intuition through experience. While intuition plays a role, it must be supported by hard facts and mathematical analysis. The speaker also notes that the choice of alpha depends on the asset class and its return profile, with equities tending to exhibit trends and upward movement due to added value from companies. However, there is no specific alpha that universally applies to all scenarios, and it is essential to understand the unique fingerprint of each strategy through comprehensive analysis.
The speaker addresses how different asset classes affect the development of trading strategies. They note that equities are non-zero sum, while foreign exchange tends to be more symmetrical. Making these distinctions and selecting the appropriate strategies based on the asset class is crucial. Liquidity of the traded assets also poses constraints that influence the approach, especially for options, futures, or small stocks. The level of expertise required to develop a trading system varies based on the type of system and whether it is fully systematic or automated. Dr. Stark suggests that knowledge of programming languages such as Python, Java, and C++ is necessary for fully automated systems.
Dr. Stark discusses the expertise and time required to develop a trading system, emphasizing the importance of understanding statistics and programming fundamentals. While it may seem complex, one does not need to be a financial or programming expert to learn and progress in this field. Developing a trading system can take anywhere from a few hours to several months, depending on one's expertise, and ultimately condenses down to a few lines of code. The process is compared to solving mathematical problems, highlighting the analytical and problem-solving nature of building trading systems.
The speaker emphasizes the importance of both studying and practicing to develop a successful trading system. While inspiration and guidance from external sources can be valuable, it is also essential to read and learn from reputable works in mathematics and programming. The speaker recommends "Active Portfolio Management" by Grinold and Kahn as a prerequisite for those interested in the course, as it covers alpha ideas and portfolio management concepts. However, the course goes beyond theory and mathematics, providing practical case studies and examples that teach students how to translate their knowledge into computer code. Dr. Stark asserts that even complex strategies can often be expressed in just one or two lines of Python code, and understanding programming can lead to more efficient backtesting and exploration.
The speaker advises attendees to not only read books on quantitative analysis and programming systems for trading but also delve into the trading mindset by exploring books like "Trading Wizards" and "Following the Trend." They emphasize that trading is not merely a strict science but rather a creative process that requires a particular mindset and emotional intelligence, which can be learned from the experiences of successful traders. The speaker promotes their course on trading alphas and offers special discounts for webinar attendees. The video concludes by inviting the audience to ask questions through a survey and provide feedback for future webinars.
During the Q&A session, the speakers address various audience questions. They discuss the difference between trading Alpha and deep reinforcement learning courses, highlighting that the deep reinforcement learning course focuses on computer learning, while the micro-Alpha course centers on the hands-on process of mining. The lack of a generalized code for market connectivity in the micro-Alpha course is also addressed, attributed to the diverse brokers and protocols used worldwide. However, the micro-Alpha course covers transaction costs and the combination of Alphas for portfolio optimization.
The speaker emphasizes the importance of factoring transaction costs into trading strategies. They note that while the impact of transaction costs can vary depending on individual cases, understanding how to incorporate them is crucial to ensure the system remains viable. However, a comprehensive analysis of transaction costs would require a separate course dedicated to transaction cost analysis or modeling. The speaker also advises against switching from languages like C++ to Python solely because of Python's popularity, especially if the existing system is already profitable. The decision to switch should be based on the desire to explore new modeling approaches or learn new programming languages. The speaker mentions an overview of the trading adverse course that provides comprehensive answers to various questions raised during the session.
In the closing remarks, the host expresses gratitude to Dr. Stark for his valuable insights and expertise. The audience is encouraged to provide feedback through a survey, submit questions, and share their thoughts for future webinars. The host concludes by thanking the viewers for their participation and Dr. Stark for dedicating his time and expertise to the webinar.
Introduction To Price Action Trading
Introduction To Price Action Trading
The webinar introduces the concept of price action trading, where traders study the fundamental price behavior of an asset over time to make trading decisions without relying on technical indicators. The speaker explains supply and demand in trading, which creates the price behavior, and the tools used in price action trading such as support and resistance levels, chart patterns, and pivot points. The different types of chart patterns such as reversal and continuation patterns are explained, along with their significance and how to trade them. The webinar also covers the use of Fibonacci series and its ratios in price action trading to understand the price behavior and take part in the trend. The course covers different trading strategies and provides codes and conditions needed to analyze trades and backtested strategies.
In this webinar, Varun Kumar Portula, a quantitative analyst at QuantInsti, delivers an informative session on price action trading. He begins by introducing the concept of price action trading, which involves analyzing the fundamental price behavior of an asset over time to make trading decisions. Unlike relying on technical indicators like RSI or MSCD, price action trading focuses on studying the supply and demand forces in the market. The simplicity and success rate of price action trading strategies have made it popular among traders.
Portula highlights that price action trading is primarily used for short-term and medium-term trading rather than long-term investing. He uses the example of a stock's price behavior to demonstrate how traders can analyze supply and demand to predict future price movements. The imbalance between supply and demand creates various price behaviors, which can be analyzed by examining the number of sell orders versus buy orders at specific price levels. Additionally, traders utilize tools such as support and resistance levels, chart patterns, and pivot points in price action trading.
The speaker explains the concept of supply and demand in trading, where supply represents selling in the market and demand represents buying. When supply surpasses demand, it leads to a decline in prices, while when demand exceeds supply, it causes prices to rise. This supply and demand imbalance creates zones, such as supply zones and demand zones, where prices tend to fluctuate. Portula also delves into the significance of support and resistance levels, which indicate zones where sellers or buyers are in control of the market. Traders can use these concepts to develop trading strategies and make informed decisions about entering or exiting positions based on supply and demand analysis.
The webinar then explores two types of chart patterns in price action trading: reversal patterns and continuation patterns. Reversal patterns signal a change in trend, either from an uptrend to a downtrend or vice versa. Bearish reversal patterns indicate supply zones and suggest a bearish market sentiment, while bullish reversal patterns represent demand zones and imply a potential reversal towards an uptrend. The speaker provides examples of commonly used patterns for both bearish and bullish reversals, such as head and shoulders, double tops, inverse head and shoulders, and double bottoms.
Continuation patterns are discussed as patterns that occur within an existing trend and indicate the potential continuation of that trend. In an uptrend, consolidation creates patterns like flag patterns, pendant patterns, and ascending triangles. In a downtrend, patterns such as Bear Flag and descending triangles can be observed, indicating a likely continuation of the downtrend. The video emphasizes the importance of studying price behavior and identifying these patterns to predict future price movements accurately.
The instructor also emphasizes the significance of the neckline in the Head and Shoulders pattern, as it indicates weakness in the uptrend. Trading this pattern involves waiting for the price to trade below the neckline, then taking a short position with a stop loss above the right shoulder and a profit target at the Head length. However, manual trading of this pattern can be challenging, which is why the course utilizes Python programming to scan for the pattern efficiently, even with large amounts of historical data.
The video proceeds to discuss the use of Jupyter Notebook to scan for head and shoulder patterns in trading. The provided code allows traders to detect the pattern and scan for it, and it also guides them on determining entry and exit points for head and shoulder patterns. The course covers backtesting for this strategy to determine risk parameters effectively. Additionally, the section covers pivot points, which are leading indicators used to calculate potential support and resistance levels. Different types of pivot points, such as traditional pivots, Camarilla pivots, and Fibonacci pivots, are explained, each with its own formula for calculating support and resistance levels. Pivot points serve as useful tools for swing traders and intraday traders, assisting them in planning exits, stop losses, and take profits.
The concept of Fibonacci series and its ratios in price action trading is also discussed. Traders employ Fibonacci ratios, such as 23.6%, 38.2%, 50%, 61.8%, and 100%, to understand price behavior and participate in trends. During an uptrend, traders utilize retracement levels of 38.2%, 50%, and 61.8% to enter trades during pullbacks, avoiding buying at higher prices and minimizing losses. The video includes examples that illustrate how these ratios are calculated and used to take long positions effectively.
The speaker emphasizes that the course covers various trading strategies, including the use of Fibonacci retracement and trade level analytics to analyze trades and study factors such as the percentage of winners, losers, and profit factor. Detailed explanations and code examples are provided for backtested strategies. Additionally, a question regarding the suitability of Camarilla or technology levels for intraday trading is addressed.
In conclusion, the webinar concludes with a gratitude to the audience and the presenter for their participation and attention throughout the session. Varun Kumar Portula successfully introduces the topic of price action trading, covers its basics, explains its underlying philosophy, and provides insights into the tools, chart patterns, pivot points, and levels used in this trading approach.
How to Lose Money Trading Options | Algo Trading Conference
How to Lose Money Trading Options | Algo Trading Conference
During the Algo Trading Conference, Dr. Euan Sinclair delivered a comprehensive talk on common mistakes made by options traders and shared valuable insights into successful options trading strategies. He emphasized the need for traders to have an edge in the market in order to consistently make profits. Sinclair highlighted the importance of buying assets at lower prices and selling them at higher prices, but pointed out that many options traders struggle with this concept and often overpay for options.
Sinclair candidly admitted that he, too, has made mistakes in his trading career but urged fellow traders to actively work on correcting those mistakes. While some of his advice was tailored towards traders with a background in options, he stressed that many of the mistakes he discussed are relevant to traders at all levels of expertise.
The speaker placed a significant emphasis on the importance of having an edge in options trading, regardless of the trade's structure. He cautioned against designing option structures that create an illusion of risklessness, as this often blinds traders to the underlying risks. Sinclair asserted that having an edge is the most crucial aspect of trading, and it cannot be achieved merely through discipline, risk control, hard work, or intelligence. Traders need to offer a valuable service to the market and actively provide something that fulfills a need.
Sinclair delved into the complexity of options trading, specifically the necessity of accurately predicting and accounting for volatility. He emphasized that traders cannot rely solely on predicting the market's direction; they must also consider the price of the option and potential changes in volatility. Even if a trader's market prediction is correct, they can still lose money if they pay the wrong price for the option or fail to properly account for volatility changes. Therefore, options traders must primarily be volatility traders and continuously model and analyze volatility throughout their trades.
The speaker addressed the misconception around buying put and call options. While buying a put option can benefit from increased volatility when the market declines, the option's price is typically already adjusted to reflect this. On the other hand, call options tend to be overpriced during trades. Sinclair also discussed the notion of Black Swan events, which are highly unpredictable occurrences. While it may seem logical to protect against Black Swans by buying out-of-the-money options, this strategy often proves to be a costly mistake. Sinclair highlighted the example of low volatility funds that have lost substantial amounts of money and cautioned against relying solely on social media for trading information, as it often presents a skewed view of winners.
The speaker also tackled the issue of long volatility funds frequently losing money due to incorrect systematic bets. While these funds may garner attention during market turbulence, they often end up with losses in the long run. Sinclair further emphasized that options are usually overpriced, suggesting that selling options can help offset the asymmetric risks. However, it is crucial to assess whether volatility is mispriced in the specific trade context to determine if there is a viable edge in selling options.
Sinclair discussed several common mistakes made by options traders, such as the belief that trading Theta (the decay of option value over time) provides an edge and the misconception that selling far out-of-the-money options is always profitable. He cautioned that while traders may collect premiums most of the time by selling these options, the potential risks outweigh the rewards. He recommended thorough analysis of trades to understand both successful and failed outcomes, highlighting the value of actively examining results rather than relying solely on automated scripts. Additionally, he suggested selling straddles rather than strangles for better feedback and improved trade decisions.
The speaker stressed the importance of continuously reassessing one's position and considering all available information to determine the desired position. While trading costs should be taken into account, Sinclair advised traders to focus more on reducing costs rather than striving for perfection in every trade. Minimizing costs can enhance the Sharpe ratio, which mathematically has no variance. While it is essential to avoid crossing the bid-ask spread, the speaker emphasized the need to avoid restricting oneself to selling only on the offer or buying only on the bid. Instead, one should assume the role of selling on the bid and buying on the offer, devising a strategy that encompasses all associated costs. The speaker advocated for conducting more trades with a lower expected value, acknowledging that many small favorable outcomes can be more beneficial than relying on a single large win.
The concept of adverse selection was another topic addressed by the speaker. He warned that even if a trade appears promising, someone with more knowledge and insight may come along and take advantage of the trader's offer, resulting in unfavorable outcomes. Realistic expectations, avoiding excessive trading or large positions, and focusing on smaller sustainable edges were highlighted as prudent approaches to mitigate the risk of losing money over time. The speaker emphasized the value of accumulating multiple small edges that can be combined into a diversified portfolio of interests rather than relying on a single big win that may vanish quickly.
Dr. Sinclair shared his conclusion that beginning as an algo trader or options trader is not the ideal approach to achieving consistent profitability. He stressed the importance of identifying a problem or niche that involves trading options, rather than starting with the tools themselves. If the goal is to trade based on market direction, options trading alone is not sufficient, as it requires consistent accuracy in predicting volatility as well. He cautioned against the notion that buying options can guarantee consistent profits, emphasizing that accurately predicting volatility is the key to success in any options trading strategy. In conclusion, he discouraged traders from fixating on the tools and instead encouraged them to focus on understanding and predicting volatility while identifying a successful trading niche.
The speaker provided insights into the implied curve of options and its relationship with volatility. He explained that skew in the implied curve is primarily driven by the correlation between volatility and the underlying asset's movement rather than volatility itself. Consequently, the speaker suggested that the skew can largely be disregarded when considering the price of the option. Furthermore, the speaker noted that market makers often perform well during periods of market turbulence, such as the crisis experienced in 2020, as it allows them to execute more trades within the same timeframe. Additionally, the short borrow rate, which functions as a negative interest rate, is factored into the pricing of options by market makers, as it is considered analogous to a dividend.
The speaker also discussed options that exhibit characteristics akin to a negative interest rate and provided an example of a trade that was previously profitable but no longer holds true. He recommended seeking out uncertain situations with timed events to sell options. Moreover, the speaker highlighted that the classic variance premium on indices and stocks is typically overpriced. When asked about the possibility of individual traders finding edges, the speaker asserted that risk premiums are always present and available to be traded, drawing a parallel to buying stocks. The speaker expressed skepticism regarding trading volatility around earning events, stating that while it used to be a profitable strategy, it no longer holds the same level of profitability.
Sinclair addressed the evolving landscape of options trading in recent years and acknowledged that the market is not as favorable for this strategy as it once was. He responded to a question regarding the use of algorithmic tools for portfolio optimization, stating that such tools may not be necessary for those who only trade once a week. Regarding finding an edge, he advised starting with a clear observation and constructing ideas based on that observation. For example, selling options when volatility is overpriced or buying stocks when there is a tendency for upward movement. Finally, the speaker tackled the question of constructing a portfolio with negatively skewed short volume and positively skewed long volume strategies. He suggested beginning with a top-down mental model as the most effective approach.
In closing, the speaker revealed that he retired several years ago but continues to spend his time actively day trading options. He expressed his intention to persist in trading options and occasionally write papers on the subject, viewing it as both a job and a hobby. As the Algo Trading Conference came to an end, the speaker expressed gratitude to Dr. Sinclair for sharing valuable lessons and experiences in options trading. He conveyed anticipation for future sessions on options trading and extended thanks to the conference organizers for the invaluable opportunity to exchange knowledge and insights.
The audience applauded, acknowledging the wealth of information and expertise they had gained from Dr. Sinclair's presentation. Participants left the conference with a newfound appreciation for the complexities and nuances of options trading, as well as a greater understanding of the importance of having an edge in the market. Inspired by Dr. Sinclair's insights, they were determined to refine their trading strategies, avoid common pitfalls, and continuously strive for improvement.
Outside the conference hall, conversations buzzed with excitement as attendees engaged in lively discussions about the key takeaways from the presentation. Traders shared their reflections, promising to implement the lessons they had learned and adapt their approaches accordingly. Some contemplated exploring new niches within options trading, while others pledged to deepen their understanding of volatility and its impact on trading decisions.
In the days and weeks following the conference, traders eagerly applied Dr. Sinclair's advice and recommendations to their own trading endeavors. They carefully evaluated their positions, considering the available information and making informed decisions rather than being attached to previous positions. Traders focused on reducing costs, realizing that minimizing expenses could significantly enhance their trading performance. They took Dr. Sinclair's words to heart, actively analyzing their trades and seeking opportunities to refine their strategies and improve outcomes.
Dr. Sinclair's insights resonated far beyond the conference attendees. Traders across the globe, both novice and experienced, eagerly sought out recordings and transcripts of his presentation. His valuable lessons spread through online forums, trading communities, and social media platforms, sparking discussions and debates on the intricacies of options trading. As traders absorbed his wisdom, they gained a renewed perspective on their trading approaches, armed with a deeper understanding of volatility, risk management, and the pursuit of an edge.
Dr. Sinclair's contribution to the world of options trading continued to make an impact long after the conference. His writings and research papers became essential references for aspiring traders and seasoned professionals alike. Through his dedication to sharing knowledge and experiences, he inspired a new generation of options traders to approach the market with discipline, a critical mindset, and an unwavering commitment to honing their skills.
As time went on, Dr. Sinclair's legacy grew, cementing his position as a prominent figure in the options trading community. Traders looked back on his words of wisdom, recognizing the profound influence he had on their trading journeys. The lessons imparted by Dr. Sinclair served as guiding principles, steering traders away from common mistakes and towards the path of consistent profitability.
In the annals of options trading history, Dr. Euan Sinclair's name stood as a testament to expertise, wisdom, and a relentless pursuit of excellence. His contributions to the field and his unwavering dedication to helping others succeed became a lasting legacy that would continue to shape the future of options trading for generations to come.
What is Corrective AI and how it can improve your investment decisions
What is Corrective AI and how it can improve your investment decisions
Dr. Ernest Chan introduces the concept of Corrective AI, which corrects and improves human or quantitative decision-making and can be applied to asset management and trading. Corrective AI overcomes issues such as overfitting, reflexivity, and regime changes and uses big data to optimize allocations by maximizing the allocation to portfolio components. This technique, called Conditional Portfolio Optimization (CPO), employs advanced use of the Kelly formula and has shown significant improvement in the Sharpe ratio. Corrective AI can also switch to a defensive position during bear markets and optimize for other metrics. The speaker emphasizes the importance of risk management and avoiding losing trades and advises against using AI to generate trading signals. Dr. Chan suggests approaching hedge funds with a pitch deck to raise funds for new fintech startups and advises aspiring quantitative traders to read, take courses, backtest, and trade live to gain intuition about the market.
Dr. Ernest Chan, a renowned expert in quantitative trading, delivered a captivating presentation on the concept of Corrective AI and its application in improving human and quantitative decision-making. He emphasized that AI is more effective in correcting decisions rather than making them from scratch, making it a valuable tool in asset management and trading. Dr. Chan cautioned against using AI directly for trading or investment decisions, instead advocating for its use in correcting decisions made by other systems or algorithms.
During his talk, Dr. Chan delved into the financial AI winter, a period spanning from 2000 to 2018 characterized by limited advancements in AI and machine learning (ML) applications in trading. He discussed the reasons behind the failure of many machine learning-based hedge funds, such as overfitting, reflexivity, and regime changes. However, he introduced a game-changing technique called corrective AI that overcame these challenges. By learning from private trading strategies or portfolio returns, corrective AI predicted their future returns, making it an invaluable and practical tool for traders and asset managers. Notably, corrective AI's resilience to arbitrage made it more reliable than traditional AI approaches in the trading domain.
The speaker highlighted the significance of big data in predicting trading strategies using AI. Various predictors, including oil filters, bond market volatility, macroeconomic indicators, and sentiment on heavily traded stocks, were analyzed to make accurate predictions. However, the speaker acknowledged the difficulty for individuals to amass such vast amounts of data, as it entailed thousands of inputs. To address this challenge, the speaker's company had created hundreds of predictors specifically for individual traders to utilize. Furthermore, he introduced the concept of using the probability of profit to size bets and allocate capital, a departure from traditional risk management solely based on returns. The AI system implicitly defined the trading regime based on the features it monitored, enabling adaptive risk assessment of investment strategies.
The speaker delved into the notion of regimes, differentiating between explicit and hidden regimes. While explicit regimes such as bullish and bearish markets were easy to identify in hindsight but hard to predict accurately, hidden regimes, such as the behavior of Robinhood traders buying call options, were challenging to identify but predictable through telltale signs analysis. Machine learning's expanded dimensionality of input greatly enhanced the prediction of hidden regimes.
Dr. Chan introduced an advanced technique called conditional portfolio optimization, which surpassed traditional portfolio optimization methods like risk parity, minimum variance, and Markowitz mean-variance. By maximizing the allocation to portfolio components through big data injection, corrective AI achieved impressive results. This technique leveraged big data to identify context, account for regime changes, and analyze the impact of factors such as inflation, interest rates, and commodity prices.
The speaker emphasized that AI had the ability to capture information that traditional portfolio optimization techniques could not. By considering big data and external factors, not just past returns, the technique called Conditional Portfolio Optimization (CPO) demonstrated significant improvements in the Sharpe ratio across various portfolios. It even exhibited up to three times improvement in the case of an S&P 500 portfolio. CPO further enabled defensive positioning during bear markets and could optimize for other metrics, including ESG ratings. The technique underwent scrutiny from reputable machine learning researchers and was currently being tested by major financial services companies worldwide. The speaker acknowledged the collaborative efforts of their research, data science, quantitative analysis, and engineering teams in achieving this success.
Dr. Chan advised against using AI solely to generate trading signals, instead recommending its application as "corrective AI" to compute the probability of profit in one's current trading strategy. He emphasized the crucial role of risk management and the importance of avoiding losing trades. When questioned about employing machine learning to understand the macroeconomic environment, he asserted that the specific type of machine learning used was not critical, and the primary factor lay in its ability to improve investment decisions.
In the discussion, the speaker emphasized the significance of amassing a vast number of inputs for big data to effectively predict the return of various portfolio capital allocations. By predicting returns at the portfolio level using big data and portfolio composition, Corrective AI had the capability to identify the best portfolio under each regime. In response to a query about sentiment analysis as a part of ML inputs, the speaker confirmed that any data stream could be added to provide additional features, which could then be merged into the input features. Furthermore, the choice of machine learning algorithm was deemed less important compared to the quality and relevance of the inputs themselves. Additionally, the speaker asserted that Corrective AI had the capability to predict black swan events, and their indicators had been successfully utilized to forecast market crashes.
The benefits of utilizing AI for tail event prediction in investment decisions were discussed, and recommendations for data providers were provided based on the frequency of trading strategies. The speaker also addressed questions related to data, machine learning techniques for financial data, and the potential use of reinforcement learning for trading. While emphasizing that risk management and portfolio optimization were the most valuable use cases for AI and machine learning in trading, the speaker admitted not being an expert in reinforcement learning and lacking first-hand experience in its efficacy.
The speaker explained the concept of AutoML, which involved the automation of parameter optimization in AI to enhance efficiency. Furthermore, the speaker discussed how hidden regimes in finance could not be explicitly identified but could be predicted implicitly using AI to aid in return prediction. Regarding adding features to a model, the speaker advised collecting as much data as possible from various sources. Lastly, the speaker described their approach as being within a supervised learning context, with the target variable typically being future returns or the future Sharpe ratio of a strategy.
Dr. Ernest Chan provided valuable advice to an individual who had been testing algorithmic trading models for the past six months but was unsure about raising funds and attracting venture capitalists for their new fintech startup. He suggested approaching various hedge funds with a pitch deck that included a track record demonstrating success. However, he cautioned that venture capitalists typically showed limited interest in algorithmic trading models. Additionally, Dr. Chan advised aspiring quantitative traders to immerse themselves in extensive reading, take courses in the quantitative field, and engage in backtesting and live trading to gain intuition about the market. He emphasized that the transition from being an armchair trader to a real trader was best achieved through live trading experience.
Dr. Ernest Chan's presentation explored the concept of Corrective AI, its advantages in improving decision-making, and its application in asset management and trading. He highlighted the limitations of traditional approaches, such as overfitting and regime changes, and emphasized the effectiveness of Corrective AI in overcoming these challenges. The speaker also discussed the importance of big data, portfolio optimization, risk management, and the ability of AI to predict hidden regimes and enhance investment strategies. Overall, Dr. Chan provided valuable insights and guidance for individuals interested in leveraging AI and machine learning in the financial industry.
Education in financial markets: Structured approach & emerging trends - Algo Trading Conference 2022
Education in financial markets: Structured approach & emerging trends - Algo Trading Conference 2022
Nitesh Khandelwal, the co-founder and CEO of Quan Institute, took the stage at the Algo Trading Conference 2022 to introduce a panel discussion focused on education in financial markets and the emerging trends within the industry. The panel consisted of experts from India, Singapore, and Switzerland who held significant roles in educational initiatives at various institutions, brokerages, global exchanges, and the asset management industry. Khandelwal stressed the importance of structured learning avenues for individuals venturing into the financial markets, as the industry continues to experience substantial growth and attracts participants from diverse backgrounds. The objective of the panel was to delve into the fundamental elements of investment and trading theses and shed light on how to acquire knowledge in these areas. The discussion encompassed topics such as asset allocation, data-driven research, the rise of retail investors, and the impact of technology on financial education.
As the panelists took turns introducing themselves, they shared their backgrounds in the finance industry and their involvement in educational initiatives, along with their best-selling finance books. They emphasized the significance of education in financial markets and the potential consequences of investing without proper knowledge. They highlighted the prevalence of scams and Ponzi schemes that exploit individuals with limited financial literacy. The panelists stressed the need for ongoing education, as markets continue to evolve and expand.
The speakers engaged in a conversation about the importance of acquiring adequate knowledge before entering the financial markets. They cautioned against blindly jumping into trading or investing without a solid foundation, as many are enticed by the ease of entry and the allure of quick profits. They warned about the risks of falling prey to unscrupulous individuals who take advantage of those lacking financial knowledge. The speakers also addressed the unrealistic expectations held by many newcomers, particularly during the pandemic, and discussed the essential skills that individuals often overlook, such as technical analysis and trading strategies.
The panelists further explored the educational modules that generate the most queries and interest from users. They observed a consistent stream of queries for the module on personal finance, specifically covering mutual funds, while the section on ETFs received fewer inquiries. The speakers also shared their personal journeys in the field of algorithmic trading and how the need for financial education in India inspired them to focus on educating the masses. They recognized the growing internet penetration in India as an opportunity to reach a wider audience and enhance financial literacy. The popularity of video-based education was also highlighted during the discussion.
The panelists delved into the distinction between investing and trading, shedding light on the common misconceptions surrounding these activities. While investing is often perceived as straightforward, trading is considered complex and challenging to profit from. The panel emphasized the need for education on both trading and investing and the importance of setting realistic expectations. They then transitioned into a discussion on emerging trends in the financial markets, with a particular focus on automation and screening tools, as well as the increasing demand for live trading demonstrations. The panel noted a growing interest in trading skills and automation, especially among younger individuals, and highlighted the rising use of screening tools for short-term trading.
The speakers addressed the misconception about the returns generated by automated trading and stressed the importance of educating the public about the inherent risks associated with such investments. They also provided insights into the various roles within the financial industry, noting that traders often have job descriptions that differ from common assumptions. Andreas, one of the speakers, discussed the changing skill requirements in asset management over the years, citing the development of more complex models driven by larger players in the market and an increased presence of PhDs and quants.
The impact of machine learning and technology on financial markets education was another key topic of discussion. While machine learning is often limited to price prediction, the panelists highlighted its potential for significantly influencing portfolio management and risk assessment. They emphasized that while technology plays a crucial role in trading, it is crucial to start with a foundation of basic knowledge and common sense before delving into more advanced strategies. The panelists noted that technology has evolved over time, and even rudimentary forms of technology can provide traders with an edge in the market.
The panelists went on to discuss how technology and social media have transformed the financial markets in recent years, creating new opportunities for traders. While advancements in technology have brought significant benefits to the industry, the speakers stressed that human input and analysis are still essential for success. They warned against overreliance on technology without fully understanding how to use it effectively, reinforcing the importance of education.
Furthermore, the speakers emphasized the importance of education in financial markets and highlighted the significance of critical thinking when applying technical analysis tools. They cautioned against blindly following outdated advice from financial gurus and encouraged traders to take an experiential and interactive approach to learning. While having an expert by one's side for guidance is ideal, they acknowledged that it may not always be feasible. Therefore, traders need to be diligent in testing and questioning technical analysis tools that were developed for a different era.
Andreas Clenow and Vivek Vadoliya discussed the value of interactive online teaching and online learning in financial education. Clenow emphasized the importance of learning by doing and advised traders to avoid blindly implementing rules from trading books. He stated that there is no universally best trading system and emphasized the personal nature of each trading model, which depends on an individual's goals. On the other hand, Vadoliya suggested paper trading and simulated environments as valuable bridges between theory and practice. He acknowledged that paper trading can have its drawbacks but explained that it is an excellent way for traders with limited capital to gain confidence and prepare for real-world trading.
The limitations of paper trading were also addressed, and alternative methods to gain experience in real market environments were discussed. The speakers suggested purchasing one or two shares of a company to experience the intricacies of placing orders, managing margins, and navigating the trading platform. They also emphasized that paper trading serves as a useful introduction to the trading system, providing traders with a feel for the dynamics of the market. The complexity of simulation was acknowledged, and the need to create simulators that accurately mimic market performance, especially for strategies that make markets, was emphasized.
Looking toward the future of financial markets, the panelists shared their views on potential changes in the next five to seven years. One speaker predicted that the retail market would become even more significant due to the increasing accessibility of trading platforms and the abundance of information flowing through social media channels. Another speaker highlighted that younger generations are less familiar with traditional financial players and predicted that the average age of traders would decrease to 13 years old. The uncertainty surrounding the future of financial markets centered on how the younger generation would shape the industry.
The panelists also discussed the impact of retail traders with unrealistic expectations and the resulting tightening of regulations in India. They anticipated a future market environment with stricter regulations, which would ultimately benefit retail traders in the long run. While operating as a broker might become more challenging, regulatory tightening was viewed as a positive development for market participants. Additionally, they recommended resources for those interested in learning how markets have evolved over the past 20 years and understanding the impact of these changes on investment strategies. Suggestions included reviewing circulars from regulators and studying books on market microstructure. The session concluded with a question about Andreas's plans for a new book, to which he responded that he had already written a programming book and a novel, but he had no immediate plans for new trading books.
In closing, the speaker expressed gratitude to the panelists and attendees of the Algo Trading Conference 2022. They hoped that the session had provided a structured approach and valuable insights into emerging trends in financial markets. They offered further assistance to anyone in need of additional support. The speaker concluded by expressing gratitude to everyone involved and passed the conference over to their colleague, Afrin, signaling the end of the session.
The panel discussion at the Algo Trading Conference 2022 provided a comprehensive exploration of the importance of education in financial markets and the evolving trends within the industry. The speakers emphasized the need for structured learning and ongoing education to navigate the complexities of trading and investing successfully. They highlighted the risks associated with entering the market without sufficient knowledge, including falling victim to scams and unrealistic expectations. The panelists also emphasized the role of technology, machine learning, and social media in shaping the financial markets, while underscoring the importance of human analysis and critical thinking.
The session shed light on various topics, including the distinction between investing and trading, the significance of practical learning experiences, and the impact of automation and screening tools. The speakers also discussed the future of financial markets, with a focus on the influence of retail traders, regulatory tightening, and the need for continuous adaptation to market changes. They emphasized the importance of education in empowering individuals to make informed financial decisions and cautioned against blindly following outdated strategies or relying solely on technology.
The panel discussion provided valuable insights and guidance to the audience, equipping them with the necessary knowledge to navigate the dynamic landscape of financial markets effectively.
Regime definition: Triage between bulls and bears, why it simplifies the work
Regime definition: Triage between bulls and bears, why it simplifies the work
Lauren Burnett, one of the speakers at the Algo Trading Conference 2022, delivered an insightful presentation on the concept of regime analysis and its significance in simplifying the trading workflow. The primary focus of regime analysis is to determine the state of the market, whether it is bullish, bearish, or inconclusive, and base trading decisions on that assessment. Burnett drew a parallel between regime analysis and the triage process used in field hospitals during wartime, as both involve making quick decisions with limited resources and time constraints.
The essence of regime analysis lies in categorizing the market into two or three distinct buckets, which facilitates a simplified approach to trading. By analyzing market regimes, traders can easily identify when to take action and when to stay put. Additionally, Burnett introduced a proprietary tool for global screening of asset classes, which further simplifies the analysis process.
During the presentation, the speaker explained the concept of regime analysis in absolute terms, where the market moves either up, down, or remains stagnant, resulting in bullish, bearish, or inconclusive market conditions, respectively. While only a few asset classes can be traded in absolute terms, the majority are traded based on their relative series. Relative series refers to the performance of securities compared to a benchmark, adjusted for currency fluctuations. To illustrate this, Burnett provided an example using the S&P 500 index, highlighting how the number of outperforming securities oscillated around 50 in relative terms while showing a different pattern in absolute terms. Understanding regime and its different series can simplify the work of sector analysts and provide valuable insights into market behavior.
The impact of regime analysis on long-short equity portfolios was also discussed. The speaker emphasized that a long-short equity portfolio is the sum of the net result of the long and short positions, and its performance is determined by the delta between the two. Focusing on relative performance and sector rotation, rather than absolute movements of individual stocks, provides a smoother and more manageable approach to working with the market. The speaker explained that during a bull market, high beta stocks are on the long side, while low beta stocks are on the short side. Conversely, during a bear market, low beta defensive stocks are on the long side, while high beta, volatile stocks that quickly give up performance are on the short side.
The importance of incorporating regime analysis into market analysis and investment decisions was heavily emphasized. While generating excess returns is crucial for survival in the financial field, it is not sufficient to rely solely on fundamental or quantitative analyses. Neglecting regime analysis, which considers the market's prevailing conditions that can dictate a stock's performance, may lead to poor investment decisions based solely on valuations and trends without considering the broader market context. The speaker cautioned against shorting stocks without considering momentum and investing in value traps that lack compelling narratives to attract investors. By overlooking regime analysis, one exposes themselves to significant business risk and potential loss of investor confidence in the long run.
The speaker provided insights into how regime analysis can be used to determine why a stock has moved up or down. They explained that there are three types of answers: consolidation, sector rotation, and stock-specific reasons. By categorizing these reasons, investors can simplify their workflow and adopt a more objective approach to the market. The presentation also touched upon various technical analysis strategies, including breakouts, and acknowledged that while conceptually simple, they can suffer from inherent lag, requiring patience. Simplification was emphasized as the key to achieving perfection, and investors were advised to be servants to the market.
Two methodologies for trading, namely asymmetrical entries and moving averages, were discussed during the presentation. Moving averages were highlighted for their ability to provide market context, although there is ongoing debate regarding the ideal duration. It was noted that moving averages are not suitable for choppy markets. Interestingly, moving averages can also be used as an exit strategy. When moving averages flatten out, it indicates that the market is transitioning, and during this period, many traders experience slippage and transaction costs that can lead to a significant loss of performance. The speaker further explained the concept of higher highs and higher lows, which suggests an upward trend when a market achieves successive higher highs and higher lows. Additionally, the speaker shared their favorite methodology called "floor and ceiling," which involves identifying the right shoulder of a head and shoulders pattern to determine optimal entry and exit points for trades.
The speaker delved into the concept of regime definition using floor and ceiling marks as an example. They explained that these marks represent a higher low (floor) and a lower high (ceiling), respectively. Any price movement between these marks is considered bullish. The speaker emphasized that this concept applies across different asset classes and time frames. However, they acknowledged that defining regimes computationally is a time-consuming task. The speaker introduced the concept of a "score," which represents the average of all diverging definition methods. The score helps determine whether various methodologies agree or diverge, both in terms of relative and absolute prices. A score indicating agreement suggests a bullish sentiment, while a score of zero indicates divergence.
The power of using a scoring method to assess whether bull and bear signals align in the market was discussed. A score of zero indicates disagreement between the methods, while a score above zero indicates agreement between absolute and relative indicators. The speaker introduced the concept of gain expectancy, which involves calculating the win rate multiplied by the average gain minus the loss rate multiplied by the average loss. This gain expectancy analysis helps segregate the market into two categories, bulls and bears, enabling focused analysis on sectors that are performing well. However, it was emphasized that this analysis serves as a preliminary step to identify outperforming securities that should be considered for investment.
The question of whether regime analysis can be applied to individual stocks or is limited to sectors was raised. The speaker clarified that the regime analysis can be scored for every individual stock and applied at the market level. They cautioned against the common mistake of shorting overbought stocks and highlighted the tendency for oversold stocks to become depressed, often leading to a swift rebound. Furthermore, the speaker explained that overbought and oversold conditions are contextual and are averaged based on whether a stock is in bearish or bullish territory, observed empirically over time.
The presentation also discussed how regime analysis can help traders avoid false positives in technical analysis. By applying regime analysis to differentiate between bullish and bearish scenarios, traders can simplify their workflow and make more objective trading decisions. The speaker cautioned against the compounding risk that can arise from exclusively practicing trend following on the long side and mean reversion on the short side. They advised treating both sides similarly to mitigate poorly managed risks. When asked about hedging the right and left tails with options, the speaker advised against it and suggested enjoying the ride instead. Relative indicators, such as moving averages, were also explained and their use on a chart demonstrated.
During the presentation, the speaker introduced different colored dots on a chart to represent specific patterns and indications. Red and green dots represented Swing High and swing lows, respectively. The chart also featured blue and pink triangles representing the floor and ceiling marks, with blue indicating a bullish regime. Additionally, light salmon and light green triangles represented a trading range. The speaker clarified that their regime analysis methodology was not influenced by any specific book but expressed appreciation for Robert Carver's work on systematic trading. Regarding the impact of monetary policy on regime analysis, the speaker emphasized the critical role of the US Federal Reserve's policies, as the US dollar directly or indirectly influences global sentiment and market trends.
Towards the end of the presentation, the speaker addressed different scenarios that can impact the market, particularly focusing on the concept of "regime." They discussed three specific scenarios that can affect the market regime. The first scenario referred to the market being too "frosty," indicating a cautious and uncertain market environment. The second scenario involved the arrival of bond vigilantes, who play a role in regulating interest rates and influencing market behavior. Lastly, the speaker mentioned the impact of inflation, which can force the hand of the Federal Reserve to adjust monetary policy. These scenarios were presented as external factors that influence the market regime rather than being controlled by it.
To navigate these scenarios effectively, the speaker introduced a tool that provides information on the current market regime. This tool assists traders in positioning themselves appropriately and adapting to changing market conditions. By having a clear understanding of the regime, traders can make more informed decisions and adjust their strategies accordingly.
The presentation emphasized the significance of regime analysis in simplifying the trading workflow. By categorizing the market into distinct regimes and understanding their implications, traders can make better-informed trading decisions. The concept of regime analysis was applied not only to sectors but also to individual stocks, enabling a comprehensive assessment of market dynamics. The presentation also highlighted the importance of considering both absolute and relative indicators, such as moving averages, to gain a comprehensive view of the market.
The speaker's insights on regime analysis, methodologies for trading, and the application of scoring systems provided valuable guidance to traders seeking to streamline their trading approach and improve decision-making. The presentation concluded by underscoring the impact of monetary policies, global sentiment, and market trends in shaping market regimes, and the importance of staying adaptable and responsive to these dynamics.
Micro-Alphas: Financial Geology | Algo Trading Conference
Micro-Alphas: Financial Geology | Algo Trading Conference
During his presentation, Dr. Thomas Starke delved into the concept of "micro alphas," which he referred to as financial geology. He began by discussing how the trading landscape has evolved from traditional open-outcry financial markets to screen-based trading and, more recently, to algorithms. To illustrate this transformation, he drew an analogy to the gold rush days, where individuals would pan for gold nuggets in rivers in their quest for fortune.
Dr. Stark emphasized that trading has become increasingly complex with the advent of advanced tools such as data analytics, machine learning, and artificial intelligence. He explained that simple technical indicators like moving averages are no longer as effective, and professional trading has shifted towards the utilization of quantitative strategies. The conventional definition of alpha, which represents returns that are not correlated to the market, was presented, with benchmarking against the S&P 500 or Spy ETF.
The speaker highlighted the challenges faced by alpha strategies in today's markets. They noted that the proliferation of players, including high-frequency traders, has increased market efficiency and randomness, making it harder to extract profits and reducing the effectiveness of predictive indicators.
Next, the concept of microalphas was introduced, and the speaker demonstrated how machine learning can be used to generate these small, specialized alpha-generating strategies. By combining multiple weak predictors using ensemble methods like bagging or bootstrap aggregating, stronger predictors with reduced variance and a lower risk of overfitting can be created. The speaker illustrated this concept using the moving average crossover trading signal as a weak predictor within a microalpha strategy. Through backtesting and splitting results into train and test sets, more profitable trading strategies can be developed.
Dr. Stark emphasized the importance of testing and optimizing trading strategies to avoid overfitting. Rather than simply selecting the best set of parameters, the speaker suggested plotting available parameters and finding correlations between the chosen test and metric. Robustness to overfitting in microalpha strategies was discussed, and the use of aggregation through bagging was highlighted as a method to combine weak alphas. The speaker presented a client's strategy as an example of how combining alphas can enhance results.
Furthermore, the speaker introduced the concept of "financial geology" or "alpha mining," where microalphas are individually unremarkable but can be combined to create a more solid and effective trading strategy. They emphasized the importance of breadth, which refers to the number of assets or trading strategies used and their correlation. While scaling up skill is challenging, increasing breadth can lead to a higher information ratio and improved performance.
The discussion then shifted to the importance of portfolio weighting and hierarchy in optimizing performance. Different weighting schemes, such as equal weights, tangency portfolios for asset managers with significant client assets, and optimal f for risk-tolerant retail traders, were explained.
The production of signals and their normalization to create position changes over time were discussed, along with the need to understand and minimize transaction costs. The speaker highlighted how a long-only strategy can be transformed into a quasi-short strategy through scaling. They also mentioned the existence of a weekday effect in strategies, where position sizes vary across weekdays, potentially leading to the design of new strategies. Trading algorithms were emphasized as a means to minimize transaction costs, with the Arrival Price algorithm showcased as an example.
The speaker introduced the alumgram I'm going Chris model, an execution curve model that helps identify close-to-best execution for transactions. By achieving execution better than the mid-price, traders can reduce transaction costs and capitalize on smaller edges, adding more microalphas to their models. An ESG strategy was presented as an example, demonstrating its resilience in volatile market conditions.
Dr. Starke addressed a question about overfitting and explained that it is challenging to measure and entirely eliminate overfitting. He suggested adding more alphas and running tests for each addition, observing whether the shop ratio improves or not. However, he cautioned against the possibility of cherry-picking and emphasized the importance of minimizing overfitting as much as possible, even though it cannot be completely avoided. He encouraged the audience to ask any further questions they may have in the survey they would receive after the session.
Towards the end of the session, the speaker announced a 15-minute break before the next session on regime definition trial between bulls and bears, which aimed to simplify the work. They also mentioned that Lauren Burner from Tokyo, Japan would be joining the session. The speaker expressed gratitude to Thomas Paul for his participation in the first session and expressed hope to see him again soon.
Dr. Thomas Starke provided valuable insights into the concept of "micro alphas" and financial geology. He discussed the evolution of trading from traditional markets to algorithm-based strategies, the challenges faced by alpha strategies in today's market environment, and the potential of machine learning to generate microalphas. The importance of testing, optimizing strategies, and avoiding overfitting was emphasized, along with the significance of portfolio weighting, transaction cost management, and the use of trading algorithms. The speaker also introduced the alumgram I'm going Chris model for better execution and announced the release of a quantra course on micro alphas. The session ended with a call for further questions and a break before the next session.
Introduction To Systematic Options Trading | Free Webinar
Introduction To Systematic Options Trading | Free Webinar
Akshay Chaudhary, a quantitative analyst at Continuum, delivered an insightful presentation on the significance of systematic trading in options. He began by illustrating the pitfalls of trading based on intuition and emotion, recounting a trader's unfortunate experience of incurring significant losses. Akshay emphasized the need for a well-defined trading plan, a stringent logical framework, and the implementation of stop-loss measures to mitigate risk.
The speaker delved into the systematic approach to options trading, explaining its multi-step process. It begins with acquiring options data, which can be obtained from vendors or free sources such as Yahoo Finance or Google Finance. The data is then organized and stored in CSV files or databases depending on its size. The next step involves screening the data based on specific parameters, creating a subset of the entire dataset. Following this, an option strategy is defined, and entry and exit rules are established. The strategy undergoes backtesting, evaluating its performance based on metrics such as maximum drawdown, Sharpe ratio, and variance. Finally, the strategy is optimized by adjusting parameters to maximize profits or minimize risk, and it is forward tested or paper traded to validate its effectiveness in a live market setting.
The systematic options trading process was further explained, highlighting the importance of retrieving and cleaning data, creating screeners to identify suitable options, defining clear trading rules for entry and exit, conducting backtesting to assess performance, optimizing strategies if necessary, and forward testing them in real-time market conditions. The speaker introduced a back short butterfly strategy as an example, utilizing technical indicators for trade entries and exits. They demonstrated the code for importing data, calculating indicators, generating signals, and backtesting the strategy.
The video presentation showcased the backtesting results of a simple strategy. The strategy relied on specific entry and exit conditions, with the backtesting results illustrating net profit and cumulative P&L. The speaker mentioned more complex strategies like iron condors and emphasized the importance of forward testing strategies through paper trading scenarios before deploying them in the live market. Dos and don'ts of systematic options trading were also discussed, including obtaining data from credible sources, factoring in transaction costs and slippages, maintaining capital buffers, and implementing stop-loss measures to manage risk effectively.
Risk management in options trading was highlighted, with strategies such as stop-loss orders and hedging being emphasized. The four key dos of options trading were outlined: backtesting and optimizing strategies, utilizing appropriate position sizing and risk management techniques, maintaining simplicity in the trading system, and adhering to the established plan. Conversely, traders were advised to avoid complicating the system, interfering with the strategy, overexposing themselves to a single strategy, and trading illiquid options. The speaker also promoted a comprehensive course called "Systematic Options Trading," covering various aspects of systematic trading and trading strategies.
In the context of acquiring historical options chain data, alternatives to Yahoo Finance were explored. Broker platforms such as TD Ameritrade or E-Trade were recommended as they provide access to historical options chain data. Third-party data providers like OptionMetrics or IvyDB were also mentioned as sources of historical options data, albeit for a fee. It was emphasized that thorough research should be conducted to select a reliable data provider that suits individual needs.
The speaker stressed the importance of data vendors for real-time data in options trading, emphasizing the need for credible data sources. They addressed a question regarding the course content, assuring viewers that files for backtesting butterfly options would be provided. The course covered strategies such as the butterfly strategy, iron condor strategy, and spreads. It was clarified that the course spanned from basic to advanced levels, catering to individuals with a foundational understanding of options. Technical analysis was mentioned as an exit tool, helpful to have knowledge about but not a prerequisite.
Various questions from the audience regarding the overlap between the executive program in algorithmic trading and options trading, the availability of data for backtesting in Python, and criteria for considering options as illiquid were addressed by the speaker. Python was recommended as the preferred coding language for backtesting, with the use of libraries for technical indicators and machine learning. However, it was noted that other languages like Java could also be used. The speaker mentioned BlueShift as another option for backtesting, as it provides a Python interface.
The importance of forward testing strategies before scaling up was emphasized. It was advised to conduct forward testing for a few months to a year to ensure the strategy performs well in the live market before increasing capital or making any adjustments. Confidence in the system's effectiveness is crucial before deploying it on a larger scale. The duration of forward testing may vary based on the trading frequency and specific strategy employed. The speaker emphasized the need for thorough backtesting and paper trading before forward testing, gradually scaling up capital while monitoring the system's performance.
The speaker recommended testing systematic options trading strategies for a minimum of three to four months to capture different market scenarios and assess performance under various conditions. Several audience questions were addressed, including inquiries about automating the supply and demand strategy and whether the course covered strategies based on the IV (Implied Volatility) surface. The speaker also provided a brief explanation of calendar spreads and advised interested learners to connect with course counselors to determine the most suitable course for their objectives, such as becoming a quant trader.
The possibility of using an algorithm to identify swing or reversal candles was discussed. The speaker explained that the feasibility depends on the development of logical rules based on specific candle parameters or properties, such as candlestick patterns like the hammer pattern. Regarding the choice between C++ and Python for trading, it was suggested that Python suffices for longer timeframes, while C++ is more suitable for low-latency and high-frequency trading. For newcomers interested in algorithmic options trading, the speaker recommended exploring the quantitative approaches in futures and options trading track. They also emphasized the relevance of automated trading using Python and Interactive Brokers.
The speaker wrapped up the webinar by encouraging attendees to complete a survey to provide feedback and ensure all their questions were addressed. They reminded viewers of an exclusive discount available only to webinar attendees and suggested reviewing the course page and taking advantage of the free preview before enrolling. Viewers were invited to connect with course counselors for further information and a customized learning path. The speaker expressed gratitude for the support of the audience and encouraged them to provide feedback for future webinars.
Competitive Edges in Algorithmic Trading | Algorithmic Trading Course
Competitive Edges in Algorithmic Trading | Algorithmic Trading Course
During the webinar, Nitesh Khandelwal, the co-founder and CEO of Quantum City, delved into the significance of competitive edges in algorithmic trading. He began by defining what an edge is and provided examples of different trading strategies. Khandelwal emphasized that competitive edges are crucial for trading businesses to thrive as they become more successful. Throughout the session, viewers gained a comprehensive understanding of the broad edges that trading businesses can acquire and the specific edges relevant to different types of strategies.
Khandelwal introduced QuantInsti, his organization on a mission to create an ecosystem that enables and empowers systematic trading and investment worldwide. He highlighted several initiatives, including their leading certification program called Quantra, the research and trading platform Blue Shift, and corporate partnerships spanning across 20 countries. By sharing these initiatives, the speaker showcased the commitment of QuantInsti to their mission.
Moving on, the speaker discussed competitive edge from a business perspective, defining it as an advantage that a company holds over its competitors. To illustrate this concept, he mentioned renowned companies such as Apple, Google, Tesla, JP Morgan, and Goldman Sachs, inviting the audience to contemplate what their competitive edge might be.
Next, Khandelwal delved into competitive edges specifically in algorithmic trading. He outlined various sources of competitive edges, including proprietary technology, intellectual property rights, unique products or services, cutting-edge technology, strong company culture, and access to specific resources or ecosystems. In the context of algorithmic trading, he explained that it involves placing orders based on certain logic or conditions, which can be automated or manually managed. The use of algorithms in trading provides a competitive edge by enabling faster data processing, efficient search capabilities, and improved user interfaces or flows. The speaker cited RenTech as an example of a company that has acquired significant edges through their intellectual property and systems in the algorithmic trading domain.
The discussion then shifted to the classification of trading strategies. Khandelwal broadly categorized investment or trading styles as quantitative, technical, or fundamental. He further categorized the underlying trading view or factor as trending, mean reverting, or event-based. He went on to explain 15 key segregations and competitive edges in the world of trading, encompassing strategies such as momentum trading, statistical arbitrage, value investing, breakout trading, carry trading, and event-based systems. The speaker highlighted that some of these systems are highly automated, while others involve more discretionary decision-making.
Addressing the importance of speed as a competitive advantage in algorithmic trading, Khandelwal emphasized the need to reduce latency in all aspects of trading, including transmission or network latency. He explained that achieving lower latency involves colocating or placing systems near the exchange in proximity data centers to minimize the time it takes for data to travel. After optimizing transmission latency, further enhancements can be made to the hardware and software infrastructure of the algorithmic trading system to reduce the time it takes for data to reach the exchange. The speaker emphasized that the faster the trading system, the more significant the alpha, which is crucial for high-frequency trading firms.
The discussion expanded to other competitive edges in algorithmic trading, such as the quality of data and access to alternative data sources like satellite imagery for demand assessment. Khandelwal highlighted the importance of a strategy infrastructure that efficiently converts ideas into executable actions. He also mentioned the advantages of extensive research capabilities, advanced pricing models, and access to various markets through brokers or prime brokers. Throughout the presentation, the speaker emphasized the significance of having a unique competitive edge to succeed in algorithmic trading.
One topic touched upon was the concept of "last look" in forex trading, where the market maker has the final say on accepting a trade after a buyer and seller agree on a price. This preferential access serves as a significant edge in trading. Additionally, Khandelwal highlighted the importance of a smooth back office and proper risk management as computational edges, as they help traders avoid substantial losses. He also emphasized the advantage of having access to funds without immediate payment, which provides flexibility in trading.
Furthermore, the speaker discussed the competitive edges that financial institutions and traders can have in algorithmic trading. He identified the low cost of funding and on-tap access to trading desks as a major edge enjoyed by banks. Another edge is having a tax structure that effectively reduces the capital gains tax to zero. Access to information, news, and regulatory changes also serves as a significant edge. Finally, intellectual property, including unique strategies, hardware and software enhancements, and proprietary processes, provides traders with a substantial advantage over their competition.
Continuing the discussion, Khandelwal highlighted nine competitive edges that can contribute to the success and rapid growth of traders. These edges include process know-how, patents, skills, dedicated teams, and continuity. Possessing one or more of these edges can be a solid foundation for traders to thrive in the market. The speaker then outlined the relevant edges for specific strategies such as pair trading and high-frequency market making, including factors like speed, market data, strategy infrastructure, back-office risk management, funding cost, and intellectual property.
The speaker underscored the importance of identifying and acquiring specific edges that are relevant to one's own trading strategy. Understanding the types of edges that align with the chosen strategy is crucial, as it enables traders to focus on acquiring and leveraging the right advantages. Khandelwal also emphasized the significance of effective risk management and mentioned the utilization of their proprietary risk management tools.
To navigate regulatory challenges, the speaker suggested starting with the regulator's resources, such as their FAQs or frequently asked questions section, which can provide valuable insights. Lastly, Khandelwal encouraged viewers to consider the EPAT program for those interested in establishing their own algorithmic trading desk or pursuing a career in quantitative trading.
During the Q&A session, the speaker addressed various audience questions on topics ranging from regulations to specific trading strategies like short gamma strategy. He highlighted the importance of market microstructure and introduced Dr. Robert Kissel, a new faculty member with extensive experience in the field. Khandelwal also responded to a question about applying data science in trading, emphasizing that data science has multiple applications beyond just machine learning or data analysis. He recommended having a basic understanding of statistics and financial markets to fully leverage the potential of data science in trading.
In addition, the speaker discussed the use cases of machine learning in algorithmic trading, including predicting market trends, managing risk, and detecting regimes to determine suitable strategies. He acknowledged that automation can help overcome psychological aspects of trading to some extent, but ultimately, a systematic approach, with or without automation, is what leads to success. Khandelwal advised those who are not proficient in programming to begin with free resources to learn programming and gauge their interest level before fully committing to algorithmic trading.
In the final segment, Khandelwal focused on programming tools used in algorithmic trading. He highlighted that creating software to connect to the exchange and decode data is typically done in C++ or even directly on hardware. However, for strategy development, Python is often used unless high-frequency trading, which requires order processing in microseconds, is the focus. The speaker encouraged participants to email their unanswered questions due to time constraints.
Nitesh Khandelwal delivered an insightful presentation on the concept of competitive edges in algorithmic trading. He provided a comprehensive understanding of the different types of edges, trading strategies, and the importance of acquiring relevant advantages to succeed in the dynamic trading market.
traders to thrive and grow at a breakneck speed. He then outlines the relevant edges for a specific strategy, such as pair trading and high-frequency market making, including speed, market data, strategy infrastructure, back-office risk, funding cost, and intellectual property.
Ask Me Anything: Sentiment Analysis and Alternative Data in Trading
Ask Me Anything: Sentiment Analysis and Alternative Data in Trading
The webinar began with the host introducing three panelists who are part of the Certificate in Sentiment Analysis and Alternative Data for Finance (CSAF) faculty. The CSAF is a comprehensive course designed for professionals in the finance industry, covering various aspects of trading, investment decision-making, and news analytics. The panelists included Professor Christina Alvanoudi-Schorn, Professor Gautam Mitra, and Dr. Pete Black, each bringing remarkable backgrounds and expertise in finance. The session also provided information about CSAF and its benefits, along with brief introductions to Unicom, Opturisk Systems, and Contingency.
After the introductions, the presenters explained the format of the "ask me anything" (AMA) session. They mentioned that questions received from various countries had been combined and sorted into four categories: sentiment analysis, alternative data, career opportunities, and other questions. Although they aimed to answer all questions, they acknowledged that time constraints might prevent addressing everything.
The first set of questions focused on sentiment analysis and trading. The presenters referred to a 2007 paper by Professor Peter Tetlock that initiated the field. They discussed the concept of sentiment analysis in trading, explaining how sentiments can be assigned positive or negative values before affecting asset prices in the market. They referred to handbooks on news analytics and finance, as well as sentiment analysis in finance, as valuable resources for those interested in the topic. The importance of analyzing not just words but also the semantics of information presentation, as highlighted by Professor Stephen Pullman of Oxford, was also emphasized. Professor Christina Alvanoudi-Schorn took over to answer specific questions related to sentiment analysis implementation and its broad applications within the finance industry, such as asset allocation, portfolio optimization, and credit risk analysis.
The presenters then discussed the use of Python and machine learning techniques for sentiment analysis and predicting market movements. They mentioned that Python is commonly used due to its availability of well-known packages for sentiment analysis and financial market applications. They also touched on deriving sentiment from fixed and open interest data and how market sentiment impacts option pricing. They noted that the time delay between market announcements and data processing provides traders with an advantage to inform their trading strategies.
Transitioning to the topic of alternative data, the speakers explained how it can be used to predict company revenues in a much shorter time frame compared to traditional data sources. Alternative data encompasses various sources, including email and credit card data, as well as satellite and drone imagery and geo-location data from cell phones. They highlighted that sentiment analysis can also be applied to alternative data from social media, providing insights into positive or negative views among traders on individual stocks. The objective is to use alternative data to predict future earnings or revenues for making profitable investment decisions.
The speakers mentioned an upcoming use case study on using e-commerce receipts to predict the revenue of products and producers sold on Amazon in the Foundations of Alternative Data lecture. They referenced an interesting study conducted by a colleague, who used receipts from Walmart and a pizza company to predict changes in their sales. They also discussed other case studies, such as the one involving a terabyte of open-source news data from Google called GDELT. Various sources of alternative data were listed, highlighting the rapid growth of data brokering.
Moving forward, the presenters discussed compliance issues and data ethics related to acquiring and using alternative data in trading. They stressed the importance of being mindful of data privacy and ensuring personally identifiable information (PII) is not present in the acquired data. The ethical considerations of data harvesting strategies were also emphasized. Regarding sentiment analysis, they likened it to alchemy, where the goal is to find winning strategies using alternative data, while cautioning the need to assess the worthiness of the pursuit.
Career opportunities in the financial market were then explored, particularly for individuals with advanced programming and software technology skills. The speaker mentioned the challenges of transforming quantitative and AI machine learning models into applications with rewarding implementation. They suggested that professionals already in the financial industry with traditional qualifications like CFA or FRM should explore new areas in the evolving financial market, where big players such as information suppliers offer new opportunities. The speaker also advised against setting overly ambitious research goals to avoid ending up with no tangible outcomes.
The correlation between AI and machine learning talent in hedge funds and their returns was discussed. Referring to a research paper from Georgia State University, it was noted that hedge funds with senior or junior level AI and machine learning skills can earn approximately 2.8% annual alpha, making it a great career opportunity for individuals capable of generating extra returns. The speakers highlighted the various career opportunities available in alternative investments that utilize AI, such as stock selection or assisting banks in underwriting credit cards and mortgages. They mentioned programs like CAIA Charter and Financial Data Professional, which provide training on AI and machine learning techniques as well as data ethics for financial markets, and encouraged students to pursue data science positions opening up in the industry.
Professor Christina Alvanoudi-Schorn emphasized the importance of understanding the dataset and sentiment data, as well as how to interpret results from machine learning algorithms when pursuing finance careers. She noted that data science is not limited to finance but can be found in almost every company. However, she highlighted the abundance of positions open in finance, especially concerning sentiment analysis and alternative data. For those interested in algorithmic trading with knowledge of Python and forecasting skills, she mentioned the availability of books to help them get started. The course she discussed included nine foundation lectures, three of which covered alternative data, and 12 use case lectures presented by industry practitioners.
The speakers addressed the question of whether AFL or Python is better for trading. AFL, which stands for Amy Broker Formula Language, was developed by a former journalist and offers a language for rapidly implementing technical analysis. While acknowledging the usefulness of AFL, they recommended Python for a deeper level of analysis and strategy implementation. They also stressed the importance of using a variety of tools and techniques to make informed trades and manage risk. While no silver bullet guarantees trading success, even slight improvements in probability can lead to significant profits.
The professor and his colleagues discussed the significance of using both market data and sentiment data in constructing trading models. Market data reflects the reality of trade or investment portfolios, while sentiment data gathered from sources like microblogs and Google searches provides additional information for predicting market movements. They suggested using quant models or AI machine learning models for making predictions, but emphasized the importance of ensembles or voting systems to arrive at a consensus. The speakers expressed enthusiasm for working on sentiment analysis projects and providing education on the topic through webinars. They encouraged attendees to send in questions via email for future responses.
As the webinar concluded, the participants gained valuable insights into sentiment analysis, alternative data, career opportunities, and the interplay between AI, machine learning, and finance. The panelists' expertise and experiences provided a comprehensive overview of the field, leaving the audience with a deeper understanding of how sentiment analysis and alternative data can shape decision-making in the finance industry.