The QuantCon Keynote: "Counter Trend Trading – Threat or Complement to Trend Following?" by Andreas Clenow, Chief Investment Officer of ACIES Asset Management
Over the past 30 years, trend following has been a remarkably successful futures trading strategy. Once a fringe trading style barely known outside of Chicago, it has grown into a 300 billion dollar global industry. It would be very difficult indeed to claim that trend following doesn’t work in the face of decades of empirical evidence otherwise. But trend following isn’t completely without problems.
It is well known that classic trend following models tend to lose money on a majority of trades. This is not necessarily an issue, since trend following is all about accepting a large number of small losses in exchange for a small number of large gains. As long as the net is positive, all is fine. That is the underlying idea of the strategy and it has historically worked very well.
However, if you dissect trend following models you can find weaknesses which could be exploited. This is what counter trend trading models are about. These counter trend models usually operate on a shorter time frame and with nearly opposite logic.
As counter trend models are gaining popularity in the systematic trading hedge fund field, a few questions arise. Are these models a threat to trend following? Can they be a complement to trend following? Can trend following be adapted to be less susceptible to the counter trend issue?
"Machine Learning at Bloomberg" by Gary Kazantsev, Head of the Machine Learning group at Bloomberg
In this talk, we will discuss the evolution of the machine learning landscape from the perspective of the global financial industry. We will describe the development route of several Bloomberg machine learning projects, such as sentiment analysis, prediction of market impact, novelty detection, social media monitoring and question answering, illustrating the applications with recent results from strategy development using news analytics.
We will show that these interdisciplinary problems lie at the intersection of linguistics, finance, computer science and mathematics, requiring input from signal processing, machine vision and other fields. We will talk about the methods, problem formulation, and throughout, talk about practicalities of delivering machine learning solutions to problems of finance, emphasizing issues such as appropriate problem decomposition, validation and interpretability.
We will also summarize the current state of the art and discuss possible future directions for the applications of natural language processing and machine learning methods in finance. The talk will end with a Q&A session.
"Reinforcement Learning in Algorithmic Execution" by David Fellah, Head of the EMEA Linear Quant Research Group at J.P. Morgan
Institutional orders generally exceed the absorption capacity in the immediate order book and are frequently split horizontally over time and vertically over price. The task of splitting apart a meta-order is achieved through a sequence of market transactions performed by trading algorithms, causing market impact. Consequently a great deal of research is spent on understanding market impact and its role in algorithm design in order to reduce it.
In this presentation, we discuss an application of Deep Reinforcement Learning to minimise transaction costs across a diverse range of instruments. We first discuss high-frequency market impact and its role in optimal planning for single-position and portfolio trading. We then discuss examples of how machine learning is used in short-term forecasting to augment order placement decisions.
Finally, we discuss how the algorithm considers these effects jointly, how it optimises a dynamic policy, and how it improves performance against surrogate hand-tuned algorithms.
"Market Microstructure Evolution" by Kerr Hatrick, Executive Director at Morgan Stanley, Electronic Strategist Group in Asia
Transaction costs, particularly in Asia, can make or break the cleverest of algorithmic trading strategies. These costs are determined by market microstructure, the focus of this talk. We examine, for Asia’s biggest market, the changing face of microstructure. We examine which factors, if any, contribute to intraday volatility, and we search for the footprint of algorithms in the market.
Throughout, we present our results in a new series of scientific animations, which, we hope, will both challenge and inform.
"Quantitative Trading as a Mathematical Science" by Dr. Haksun Li, Founder and CEO, Numerical Method Inc.
Quantitative trading is distinguishable from other trading methodologies like technical analysis and analysts’ opinions because it uniquely provides justifications to trading strategies using mathematical reasoning. Put differently, quantitative trading is a science that trading strategies are proven statistically profitable or even optimal under certain assumptions. There are properties about strategies that we can deduce before betting the first $1, such as P&L distribution and risks. There are exact explanations to the success and failure of strategies, such as choice of parameters. There are ways to iteratively improve strategies based on experiences of live trading, such as making more realistic assumptions. These are all made possible only in quantitative trading because we have assumptions, models and rigorous mathematical analysis.
Quantitative trading has proved itself to be a significant driver of mathematical innovations, especially in the areas of stochastic analysis and PDE-theory. For instances, we can compute the optimal timings to follow the market by solving a pair of coupled Hamilton–Jacobi–Bellman equations; we can construct sparse mean reverting baskets by solving semi-definite optimization problems with cardinality constraints and can optimally trade these baskets by solving stochastic control problems; we can identify statistical arbitrage opportunities by analyzing the volatility process of a stochastic asset at different frequencies; we can compute the optimal placements of market and limit orders by solving combined singular and impulse control problems which leads to novel and difficult to solve quasi-variational inequalities.
"Opportunities and Pitfalls in Momentum Investing" by Gary Antonacci, Author of Dual Momentum Investing: An Innovative Approach for Higher Returns with Lower Risk
Gary will begin by explaining the origins and history of momentum investing. He will show why momentum is called “the premier anomaly.” He will describe the way momentum is most commonly used and why this may not be the best approach. He will discuss the hidden risks associated with momentum and other factor based investments.
Using easily understood examples and historical research findings, he will show how relative strength momentum can enhance investment returns, while trend-following absolute momentum can dramatically decrease risk exposure.
Gary will show which assets are best to use for momentum investing. Finally, he will describe the behavioral biases you must deal with and the mind set you need to become a successful momentum investor.
In this talk you will learn how to:
a) Spot the best momentum investment opportunities in any market environment.
b) Protect yourself from bear market risk exposure and behavioral biases.
c) Construct your own low-cost, rules-based dual momentum portfolio that is simple to understand and easy to maintain.
"The 6 Stages of a Quant Equity Workflow" by Dr. Jessica Stauth, Vice President of Quant Strategy at Quantopian
This talk will provide a deconstruction of the algorithm development process for a popular and deep area of the quantitative investment world: systematic cross-sectional equity investing, also known as statistical arbitrage or equity market neutral investing.
Dr. Jess Stauth will break this workflow into 6 distinct stages, each of which presents its own challenges and opportunities for differentiation to the algorithm developer. During this talk, she will give you an insider's look at how legions of quants at the biggest hedge funds in the world spend their days.
She will also briefly explore how innovations in the fintech space have the potential to reshape this workflow and throw open the doors wide open to a new global pool of talent.
"The Hunt For Alpha Among Alternative Data Sources" by Dr. Michael Halls-Moore, Founder of QuantStart.com
The lifeblood of many quantitative trading strategies is a mix of high-quality, high-frequency asset pricing data and detailed information on company fundamentals. Such data is now available quite readily at low cost from multiple vendors. In addition it is more straightforward than ever to "wrangle" the data into the necessary formats for rapid quant research.
Quantitative hedge funds, family offices, proprietary trading houses and even some retail quants are realising that many of the traditional sources of alpha are decaying. In essence, the search for alpha must be continued elsewhere.
So-called "alternative" data sources are a relatively recent solution to the problem of alpha decay. Satellite imagery, email receipts, social media, Internet-of-Things sensors, weather patterns and earnings calls can all provide insights that lead to novel trading ideas.
Along with these new sources of data are methods to quantify and analyse it, including statistical machine learning, computer vision, sentiment analysis and deep neural networks.
In this talk we will consider these new data sets and discuss how we can apply freely-available data science tools to help find new alpha among them.
"Build Effective Risk Management on Top of Your Trading Strategy" by Danielle Jiang, the Founder and CEO of Hedga Technology
Risk management is an essential but often overlooked prerequisite to success in trading. No one would like to see their substantial profits generated over his lifetime of trading just vanishing over a few bad trades.
In this talk, Danielle will discuss a quantitative understanding of risk. She will then share a few techniques in risk management, with a case study to show how a proper risk management system helps improve the overall performance of trading strategies.
"Robust Trading System Selection" by Dr. Thomas Starke, Quantitative Trader at Vivienne Court
Even with a wide range of statistical tools available, selection of algorithmic trading strategies can
leave the trader with significant out-of-sample variability. In most cases the final decision making
is still a manual process.
This presentation will show how a combination of statistical methods and machine learning can help to automate strategy selection and boost the robustness of automated trading systems.
“Real Time Machine Learning Architecture and Sentiment Analysis Applied to Finance” by Juan Cheng, Data Scientist at Infotrie
The vast proliferation of data related to the financial industry introduces both new opportunities and challenges to quantitative investors. These challenges are often due to the nature of big data and include: volume, variety, and velocity.
In this talk, Dr. Cheng will take the audience on a tour of the “big-data production line” in InfoTrie and show how the financial news collected from various and customizable sources are transformed into quantitative signals in a real-time manner. The talk will touch on various kind of topics like sentiment analysis, entity detection, topic classification, and big-data tools.
"Applying Deep Learning Techniques to Financial Time Series" by Scott Treloar, Founder of Noviscient
Deep learning is a subset of machine learning that draws on fields including applied mathematics, statistics, computer science and neuroscience. Currently it is experiencing tremendous growth due to the confluence of larger datasets, massive computational power and the development of new algorithms. While there is a lot of work on static data and some work on sequential data (such as text-based learning), less attention has been paid to (dynamic) time series data. In finance we are often interested in problems of prediction and classification, based on time series data.
In this talk, we introduce deep learning models and discuss their application to time series data. We do this in the context of using a trained model to make predictions from new data. After introducing the framework, we work through the application of deep learning to a number of areas in finance.
"Fireside Chat with Dr. Jessica Stauth and Dr. Cecilia Woo: Breaking into the Quant World"
Dr. Jessica Stauth and Dr. Cecilia Woo will discuss how and where they got started in their careers, their past roles and projects, lessons learned along the way, where they are now, and more importantly, where they are going. This is a great talk for students and people interested in hearing about how to secure a position as a quant.
"Intra-day De-Mark Plus Order-Flow Indicator" by Dr. Christopher Ting, Associate Professor of Quantitative Finance at the Singapore Management University (SMU)
Traders apply DeMark indicators on daily and weekly charts to indicate the area in which the market was considerably oversold or overbought that the opposite price move is deemed to be fairly probable. Lesser known is their applicability on intra-day (one-minute) charts, which present challenges and opportunities of a different kind.
In this talk, Dr. Ting will walk through the stages in designing and back-testing an Intra-day De-Mark Plus Order-Flow Indicator (Indempofi) as an algorithmic trading strategy for futures contract. Following standard practice, he will separate the intra-day data into three sets: one for ``training’’ the Indempofi algo, one for out-of-sample analysis, and another one for ``paper trading". This research study shows that you need order flow to enhance the algo performance on a variety of performance measures.
"Using Machine Learning to Build Predictive Trading Models on Futures" by Ajusal Sugathan, Co-Founder and Director of Uniquant Fund Management
Ajusal will discuss machine learning techniques like random forests and deep learning to build predictive models using statistical features as input. Various aspects like feature creation, feature selection, cross validation, and data normalization will be discussed in detail. He will also demonstrate a sample strategy using machine learning techniques on CME futures contracts. Viewers will get a deep understanding of how machine learning techniques can be effectively used to build practical trading strategies.
"Is Momentum Still Relevant for Today’s Markets?" by Anthony Ng, Senior Lecturer
Despite being ‘discovered’ over 20 years ago, there is still confusion on what a momentum strategy entails and people ‘invest in momentum’. There are two generally accepted definitions of momentum in academic literature. In the quantitative equity investment sphere, momentum is frequently referred to as across securities or assets (cross-sectional or relative) and typically traded in a long-short or hedged manner. In futures trading, momentum is often referred to the past return of the security (time-series) and normally traded in a directional fashion.
Following from the above, we conducted an analysis on the performance of a momentum strategy of different asset classes: equity, fixed income, futures, and currencies. The study showed that both types of momentum are prevalent and persistent across all asset classes. Furthermore, as the correlations between the two types of momentum strategies and amongst the asset classes are quite low, substantial diversification benefit can be derived by combining them.
"A Case Study on Deploying Learning Methods to Improve Predictive Performance of Trading Strategies" by Avirath Kakkar, Founder and Co-CEO of Limnah Capital and Arvin Sahni, Director, Limnah Capital
In our experience, it is extremely difficult to find a single predictor to have a linear bearing on a classification outcome when modelling financial markets. More so when they are considered in isolation by themselves. Which is why we are forced to delve in the multidimensional non linear space at the risk of getting too excited by the discovery of a pattern.
We will share our experience on how we fine tuned a strategy to improve our trading performance. The improvement in it self may or may not provide an edge in the market if considered in isolation, however coupled with a strategy with an edge, we found it to have an improvement.
"Using the Kalman Filter in Algorithmic Trading" by Dr. Aidan O'Mahony, The Algo Engineer
The financial markets are ever changing and every good trader knows they must adapt with the market to remain profitable and reduce risk. The same is true for algorithmic trading systems, they must have the ability to adapt or face increasing market exposure and risk. One such method to update trading systems or parameters as new information becomes available is the the famous Kalman filter.
In this talk, Dr. O'Mahony will walk through the underlying concepts of the Kalman filter using illustrative examples. He will show how the Kalman filter can be easily applied to financial time series and demonstrate its implementation in a simple arbitrage strategy.
"A Framework for Developing Trading Models Based on Machine Learning" by Kris Longmore, Founder of robotwealth
Machine learning is improving facets of our lives as diverse as health screening, transportation and even our entertainment choices. It stands to reason that machine learning can also improve trading performance, however the practical application is fraught with pitfalls and obstacles that nullify the benefits and present a high barrier to entry. Building on background information and introductory material, Kris will propose a framework for efficient and robust experimentation with machine learning methods for algorithmic trading. The framework's objective is to arrive at parsimonious models whose positive past performance is unlikely to be due to chance. The framework is demonstrated via practical examples of various machine learning models for algorithmic trading.
"Fitting GARCH Volatility Models with the Generalized Method of Moments" by Delaney Granizo-Mackenzie, Academic Director at Quantopian
Generalized AutoRegressive Conditionally Heteroskedastic models are part of a core set of modeling tools developed for economic and financial time series. The notion that recent volatility influences not only future price, but future volatility, is key. We’ll go over an in-practice implementation of a GARCH(1,1) model, plus show how to use maximum likelihood and generalized method of moments (using residual moments as an objective function) to fit parameters of the model.
"Artificial Intelligence Powered By Crowdsourcing - The Next Evolution in Quantitative Trading?" by Daniel Chia, Co-Founder of Call Levels
In recent years, many funds have moved towards machine learning, where artificially intelligent systems can analyze large amounts of data at speed and improve themselves through such analysis. At the same time, the introduction of cloud and mobile technology has meant that there are now more market participants than ever before in history. Combining these two powerful trends may lead to a breakthrough in quantitative trading that can consistently outperform human managers. Daniel Chia, ex-hedge fund and sovereign wealth systematic trading manager, and now, Co-Founder of fintech start-up Call Levels shares his thoughts on this.
"Genetic Algorithms and Evolutionary Computation" by Achin Agarwal, Co-founder and Director of Uniquant Fund Management
Genetic algorithms are inspired by nature and evolution. They help in solving hard computational problems in finite time. Based on a straightforward theoretical foundation of natural selection, they provide an easy to understand framework for solving different kinds of search and optimization problems. They are especially useful when the search space is too large or complex to handle using traditional search algorithms. The talk will provide a practical demonstration of how this framework can be applied to solve a common problem of portfolio construction, highlighting the key steps involved and examining the nuances of each step. The talk will also provide a bird’s-eye view of other common problems in quantitative finance that can be handled using genetic algorithms.