March 16 - 18, 2020
Metropolitan Pavilion NYC
Metropolitan Pavilion NYC
Conference Day One: March 19 2019
8:00 am - 8:50 am Welcoming tea, coffee and registration
KEYNOTES & OPENING PLENARY SESSIONS
8:50 am - 9:00 am Chair's opening remarks
9:00 am - 9:20 am Opening keynote: Outlook for the hedge fund industry: survival of the leanest or the most tech savvy?
9:20 am - 10:00 am Panel discussion: The asset allocator's view
- Evaluation: What does a buy side investor look for in quantitative strategies
- Analysis: Finding risk, uniqueness and maintaining a diverse yet unbiased holding: the challenges of asset allocation
- Risk: Meeting your mandate while incorporating technologically forward investment strategies
Afsheen Afshar
Former Chief Artificial Intelligence Officer and Senior Managing DirectorCerberus Capital Management
10:00 am - 10:30 am Interview: Learn from this industry icon as they share their experience and insights
10:30 am - 11:10 am Networking refreshment break in the exhibition area
MORNING SESSIONS
11:10 am - 11:30 am Keynote: A new breed of investor: The development and disruption of asset management through machine learning techniques
- The new model: What do you really need to launch a new age hedge fund?
- Practical recent examples and insights into what it takes to succeed and common pitfalls to avoid
- The best and worst attributes of AI: Where do funds go wrong?
11:30 am - 12:10 pm Panel discussion: The talent war: How to attract and retain the top quant and data science minds
Moderator:
Rob Sloan Research Director WSJ Pro
Speakers:
Sheedsa Ali Managing Director & Portfolio Manager PineBridge Investments
Dan Furstenberg Global Head of Hedge Fund Distribution & Data Strategy Jefferies
Nitish Maini General Manager, Virtual Research Centre & Vice President, Portfolio Manager WorldQuant
Richard Pook Partner Dore Partnership
Rob Sloan Research Director WSJ Pro
Speakers:
Sheedsa Ali Managing Director & Portfolio Manager PineBridge Investments
Dan Furstenberg Global Head of Hedge Fund Distribution & Data Strategy Jefferies
Nitish Maini General Manager, Virtual Research Centre & Vice President, Portfolio Manager WorldQuant
Richard Pook Partner Dore Partnership
- How to reach the best people and the top minds
- Understanding that all data scientists and quants are not created equal: what looks like a data scientist may not be!
- How to create a culture that will entice them to join and stay with you: understand your value proposition as an employer
- Motivating your workforce to share their best ideas, and building a culture of collaboration in a typically combative and siloed industry
- Considering the factors that will affect retaining and motivating top talent
Nitish Maini
General Manager, Virtual Research Centre & Vice President, Portfolio ManagerWorldQuant
12:10 pm - 12:30 pm Keynote: "Big Data" is the new currency
The way data is collected, anonymized and monetized largely without the owner’s permission is ready to be disrupted providing many benefits to hedge fund data buyers. This presentation provides a pathway for the individual to control and share in the value their data creates, and for data users to gain access to richer more specific data sets.
12:20 pm - 1:30 pm Networking lunch in the exhibition area
AFTERNOON STREAMS
STREAM A
1:35 pm - 2:15 pm Panel discussion: Understanding and navigating the ‘Wild Wild West’ of alt data vendors and suppliers- Usefulness of data – evaluating the landscape and the trends
- Filtering and working with vendors:
- Excessive demand for newer/more interesting data sources places the onus of data prep firmly on the purchaser whilst low quality/inconsistent data gives a low conversion rate to prospective sources. How can you work with your supplier to provide the product you need?
- Evaluating the pros and cons of searching for the foundations of specific trade ideas vs ‘playing’ with new and interesting concepts
- Best data sourcing practices
STREAM A
2:15 pm - 2:25 pm Innovation in Alt Data
STREAM A
2:50 pm - 3:30 pm Panel discussion: What are the next hot data sources/ regions? What are the different business models that are emerging?- We’ve had sentiment analysis on social media, car park image processing, mobile positioning, video, credit card transactions… Where or what is the next big alt data source?
- Evaluating the global demand for data
- Data, data everywhere. But how can you find the signals?
STREAM B
1:35 pm - 2:05 pm A specific framework and strategy for introducing quant methods and unique data into your fundamental approach to producing alpha- A step by step approach to building viable investment signals from unique data sources
- How to build a stock selection, risk management and portfolio construction practice using quantamental methods
- How fundamental firms can overcome institutional paralysis towards building better decision-making processes
- Why most investment firms don't do R&D, and why they should
STREAM B
2:05 pm - 2:50 pm Panel discussion: How best to integrate your portfolio management, research and data science teams to deliver alpha
Moderator:
Paul Rowady Director of Research Alphacution
Speakers:
Wesley Chan Executive Vice President & Portfolio Manager PIMCO
Ivailo Dimov Quant and Data Science Research Bloomberg
Tom Doris Chief Data Scientist Liquidnet
Poul Kristensen Managing Director, Economist, and Portfolio Manager New York Life Investment Management
Paul Rowady Director of Research Alphacution
Speakers:
Wesley Chan Executive Vice President & Portfolio Manager PIMCO
Ivailo Dimov Quant and Data Science Research Bloomberg
Tom Doris Chief Data Scientist Liquidnet
Poul Kristensen Managing Director, Economist, and Portfolio Manager New York Life Investment Management
- Embracing the culture: What is needed to make the changes
- Practical insights into effective integration
- Evaluating the success: How can you measure the benefits of a successful integration
- Understanding the risks of a market saturated with the systematic approach: when will the models fail?
Poul Kristensen
Managing Director, Economist, and Portfolio ManagerNew York Life Investment Management
STREAM B
2:50 pm - 3:30 pm Fireside Debate – When hedge funds ate their ownOne side believes a blended, “quantamental” approach will consistently outperform a purely systematic approach. The other side favors an increasing level of automation. Recent empirical evidence supports both arguments. Come and find out what happens when two heavyweights battle it out.
STREAM C
1:35 pm - 2:00 pm Utilizing AI/ML to lend predictive/adaptive capability in continuously changing market environments- Common problems around utilizing AI/ML in quant analysis
- Universality, Causality and Machine Learning
- Ad hoc demonstration: The walk forward test in randomly created market environment
- Understanding the market as an unstable dynamical system
STREAM C
2:00 pm - 2:20 pm The very latest AI developments machine learning for security selection and the dangers of overfitting- Data prep and feature engineering: Is the AI built over the data based on solid foundations?
- Issues of overfitting and maximizing the signal to noise ratio
- Evaluating your algorithm choice: what do you want to achieve?
- Understanding fake signals: when machine learning fails
Keywan Rasekhschaffe
Senior Quantitative Strategist and Portfolio ManagerGresham Investment Management
STREAM C
2:20 pm - 2:40 pm Practical applications of AI/ML- Once a qualitative “factor,” sentiment around individual stocks with scale levels of online discussion can now be measured
- Applications of ‘task-specific’ vs ‘general’ artificial intelligence
- Is sentiment predictive or contrarian?
- Implications for portfolio construction
STREAM C
2:40 pm - 3:05 pm Big Data's Dirty SecretAbstract:
"Let the data speak for themselves."
"We apply machine learning to the problem of..."
These are two commonly heard phrases these days. But what data exactly are we speaking about, and what do we intend to do with it? What is ignored all too often is the quality of the data being used and how it impacts the analyses being done. Are there holes in the data? Are there anomalies? Given how dirty data can be, a more apt phrase might be "Garbage in, garbage out".
In this talk we will discuss some of the data problems we've encountered in financial data, and approaches that can be used to address them. Our particular focus will be on techniques we've employed to address missing data and bad data in credit default swap (CDS) spread histories.
STREAM C
3:05 pm - 3:30 pm FED AI: Applying AI techniques and macro economics
3:30 pm - 3:55 pm Networking refreshment break in the exhibition area
STREAM A
3:55 pm - 4:00 pm NATURAL LANGUAGE PROCESSING
STREAM A
4:00 pm - 4:20 pm Hanging on every word: Natural Language Processing unlocks new frontiers in corporate earnings sentiment analysisHow can NLP be applied to corporate earnings call transcripts? Can you dissect the tone, complexity, and overall level of engagement with analysts as indicators of earnings sentiment?
STREAM A
4:20 pm - 4:40 pm Practical uses of cutting edge NLP: what is the current state of play, and how can you benefit from incorporating it into your investment strategy?- Defining the value you can add from overcoming language barriers and maximizing your translation services to processing unstructured data from around the world in a variety of formats
- Enhancing your systems abilities: overcoming the translation barriers: an insight into unsupervised learning of cyberslang and unusual language structure
- Other practical NLP hints and tricks to set you on the path to alpha generation
Mike Chen
Director and Lead ML Researcher, Dynamic Equity; Lead Portfolio Manager, ESG EquityPanAgora Asset Management
STREAM A
4:40 pm - 5:20 pm Panel discussion: Understanding the very latest developments in NLP: what is the current state of play, and how can you benefit from incorporating it into your investment strategy?- Rule based NLP vs deep learning, i.e. relying on human intervention and ingesting touches of augmented AI rules to better detect key elements of languages in a proper NLP system
- Using proprietary deep tech and being early adopters of new technologies as opposed to relying on open source
- The challenges of adapting an English NLP engine to different languages, public sources (such as Twitter that almost has its own language), or thousands of different types of file formats (proprietary textual content from clients)
STREAM B
4:00 pm - 4:20 pm Building a unique, fully quant-driven strategy on fundamental principles: Lessons learned, pitfalls to avoid- Great quant models are built based on a deep, refined and experience-driven understanding of some processes in asset pricing.
- There are many such in-depth fundamental investment philosophies that can provide robust frameworks for great quant models.
- Most important benefits of a fundamental framework are the insightful questions it poses. Quants with data and computing tools are great at answering questions, but not at articulating them.
- Example: Dynamic Contextual Alpha Model based on a Fundamental Investment Philosophy.
- Other potential examples: Industry Models, Stock Specific Models, Macro Sensitivities on Stocks.
- Things that don't work: Making fundamental analysts do quant things like rank stocks. Making quant models do fundamental things, like screen a small group of stocks for fundamental analysts to pick from; Quants using 'of the shelf' 'academically tested' 'fundamental' approaches. Fundamentals using 'textbook' security analysis. Alpha comes from innovation, uniqueness in style, refined and customized competitive edge in the investing process.
STREAM B
4:20 pm - 4:40 pm Explainable AI for asset managementCombining fundamental knowledge and quantitative techniques into an explainable AI framework for investment decisions:
- We use a fundamental macro framework to guide data selection and curation, in combination with AI
- Combining fundamental and quant, how to balance the two
How we approach the prediction problem:
- Definition of market regimes
- Accounting for investor behavior
- This includes human and machine, as well as machine-only modes of operation
Interrogation versus explanation:
- The former is used by our team, the latter to provide information to investors
- What is the best way of surfacing explanations to investment professionals?
- Investor mental models
Use of the system for:
- Investing
- Risk Management
- Indicators
- Dealing with new and uncertain regimes
STREAM B
4:40 pm - 5:00 pm Utilizing a blended man & machine approach to generating AlphaIntegrating data, machine learning and fundamental insights to drive alpha
- Finding patterns in data to aid your strategy
- Automatic feed: immediately processing news and live market updates to your portfolio to accurately maintain risk
STREAM B
5:00 pm - 5:20 pm Model risk management for alpha strategies created with deep learning- Understanding the challenges of using deep learning to build alpha generation strategies
- Model risk management to detect when machine learning strategies are not performing as intended.
- Concept drift: Can you model an undefinable and constantly moving market? When DL should (and should not) be used.
STREAM C
3:55 pm - 4:00 pm PRACTICAL APPLICATIONS OF AI/ML
STREAM C
4:00 pm - 4:20 pm Practical applications of machine learning in investment banking
STREAM C
4:20 pm - 4:40 pm Applications of deep learning in portfolio management- How deep learning can help CIOs with common portfolio management challenges
- Business drivers when considering an AI-driven approach to asset management
- An end-to-end deep learning model for portfolio allocation
- Deployment and scalability related practical challenges for a deep learning strategy
STREAM C
4:40 pm - 5:00 pm WALLACE: The self-taught fully functional investment systemWALLACE is a self-evolving AI based on the concept of genetic algorithms. Constantly learning, WALLACE analyses over 10,000 securities across 42 dimensions each day. WALLACE’s most powerful features is its ability to anticipate market events. Over a 48- month period, WALLACE successfully predicted 40 public company acquisitions. Similarly, WALLACE avoided major market selloffs. As a result, WALLACE outperformed benchmarks and designed numerous hedge fund strategies.
- Survival of the fittest: An insight into the process and journey of creating WALLACE
- The burden of unsupervised learning: enabling the machine to decide what inputs are more or less useful
- Becoming social: How WALLACE explains its decisions and exposes its thought process
Alex Bogdan
Chief Scientific Officer, Castle Ridge Asset ManagementCastle Ridge Asset Management
STREAM C
5:00 pm - 5:20 pm Merchant mapping, ticker tagging, and panel stabilization: Cracking the dirty jobs in alternative data- Application of natural language processing technology for ticker tagging.
- Using deep neural nets to clean credit card, email receipt, URLs and other alternative datasets.
- In a world where raw data validation and structuring are handled by AI, what would be the role of today's R&D teams across the data supply chain?
