A New Approach To Quantitative Investing
Korteks : An Innovation from The Mind of British Petroleum Chief Quant Cetin Karakus
Quantamental trading is an innovative trading process that complements traditional fundamental trading strategies with quantitative ones involving mathematical models and big data. Fundamental strategies involve assessing the success of a company based on information such as company financial statements, current news, macroeconomic dynamics, and industry specific microeconomics. Quantitative trading, on the other hand, involves the use of various mathematical models and real-time market analysis to profit from market anomalies. Global Head of Quantitative and Analytical Solutions at British Petroleum Cetin Karakus has conceptualized a quantamental trading platform called Korteks that allows more people access to the most advanced trading strategies. In this article, we’ll go through the intricacies of this trading platform and what makes it so special.
What makes this quantamental trading platform different from trading platforms already in the market?
Compared to other trading platforms on the market, Korteks has the advantage of being a centralized trading platform that contains all functions relevant to the trading process as a whole. Conceptualized with the intent of creating a one-stop shop for quantamental trading, Korteks implements a quantamental trading framework systematically and comprehensively so that the users of the platform have a low friction, seamless experience. In addition to the superior user experience, it also offers very advanced quantamental trading functionality across pricing, risk and data analytics domains.
What is Korteks Workbench?
The workbench is the suite of interactive applications on the Korteks platform that allows users to manage various workflows involved in commodities trading based on the quantamental processes. The workbench consists of several modules: Base system, Data Sourcing, Data Transformation, Analytics, Visualization and Collaboration. These modules work together to create a seamless and efficient environment for the users to perform various tasks and analyzes associated with the high pace commodity trading workflows.
Base System module:
How does it lay a foundation for this platform? What are the key components?
When creating a user interface for the trading platform, there are foundational elements that are necessary for the application to operate at the most basic level. These functions are classified under the base system module. The base module system involves a list of important functions: Access control, user profiles, logical filesystem, favorites, templates, search, library, versioning, audit trail system, archiving. The user profile contains a variety of settings and information that allow each individual trader to be distinguished from one another. Depending on different users, this may entail different levels of access to functions within the system.
Other components important to a smooth user experience include a logical filesystem, favorites, search, and library. It is important to be able to store data in an organized manner. In case they need to be referenced in future trading processes. All data involved in the operation of the trading platform must be archived with an organized audit trail system. This component is very important. As it allows trade details and financial data to be traced back to its source and any changes on the critical data. Moreover, to be transparently attributed to the actors who carried out those changes. On top of this feature being nice to have, it is oftentimes a regulatory requirement. All these components of the base system module are absolutely vital to the workbench. These components provide the backbone for Korteks’s cutting edge technology to build upon.
Data Sourcing Module:
What type of data is necessary for traders to make informed decisions? How is it organized?
The types of data involved in quantamental trading are market data, reference data, and deal data. The primary type of data sourced is market data, which involves prices of commodities, curves, surfaces, and various metadata associated with the market data. The most basic and fundamental data needed in trading are the prices of the commodities. When these prices are indexed with respect to time, a curve is created. These curves, called time-series curves, provide traders a good sense of commodity prices at any given time. To expand upon curves, surfaces can be created by adding another dimension to the structure. For instance, option strike becomes the second dimension when storing option prices and Black Scholes implied volatilities. The sourcing of these types of data are extremely useful to traders as they provide fundamental information involved in commodities trading.
The next type of data needed in trading is the reference data. This type of data is usually only defined once, and does not change frequently like market data. Examples of reference data are units of measurements, currency codes and names, various calendars such as exchange holiday calendars and contract expiry calendars, derivation rules and formulas describing the relationship betweens market data objects such as curves, mapping and geolocational data needed for physical commodity trading, etc.
The last type of data to source is the deal data. The deal data relates to all data involved in the transactions that take place when deals are booked. Deal data consists of deals, which could be structured as either single-leg and multi-leg structures, organized into hierarchies of books and portfolios that directly correspond to the trading activity of the various teams in a trading organization.
All these types of data are crucial in the process of trading and must be sourced in an organized manner. The data sourced here will be used in the quantamental framework to create viable trading strategies.
Data Transformation Module:
What is its role? How does it organize data? What can you learn when organizing data in this way?
The transformation module assists the user in altering the data sourced to be cleaner and more informational when making informed trading decisions. There are many different functions involved in the transformation framework. One of the most important features is the ability to filter data. The data being sourced is not always relevant to the analytics of taking a trading position. It is important to be able to filter out junk data in order for users to be able to utilize data in their strategy creation.
Other functions that are helpful to the trader are combining, splitting, blending, and smoothing datasets. These functions also allow datasets to be more directly helpful in creating trading strategy. For example, when combining and splitting curves, users are able to visualize data in ways specific to their needs. Smoothing removes noise from datasets to make trends clearer for the trader to see. Additional tools to transform data include, time-shift, extending, replicating, and shaping.
Other components necessary in the transformation module include persistence and sharing. Persistence refers to the importance of saving transformed data to be used in later stages of the analytics process. The data transformed must be stored in an organized database with a sophisticated file-system. This allows for traders to be able to locate and refer to the transformations when creating an analytical strategy. Sharing is another component important in the transformation module to encourage collaboration amongst peers. Once data is transformed in such a way that it is informational and relevant to creating trading strategies, it is effortless for traders to be able to share it amongst themselves and leverage the existing work gone into producing high quality datasets.
Analytics Module:
What is the purpose of the Analytics module? Moreover, what is the quantamental framework? What other factors become analyzed when creating a trading strategy?
The Analytics module is at the heart of the quantamental trading platform. This module uses the data structures that have been sourced and transformed to create trading strategies. The four components to the analytics module are risk analytics, quantamental framework, pricing framework, and time-series analytics.
The quantamental framework is the most significant component of the Analytics module. The quantamental framework merges the analysis of quantitative data and fundamental investment strategies to create a final trading strategy. Quantitative strategy includes the study of market data such as market prices and curves. In addition to quantitative data, the framework uses data relevant to fundamental investment strategies. And, in cases of equities, this includes financial statements, revenues, costs, balance sheet docs etc. In cases of commodities, this could be data on supply/demand activities, pipelines, storage tanks etc. These different types of data are quantified and formatted in the same structure. This allows the same framework to process the data and generate indicators and signals. The framework utilizes computer algorithms that contain formulas and mathematical models to output metrics that can be used to generate indicators. These indicators give traders a recommendation on optimal trading strategies.
The risk analytics involve examining many different risk aspects of trading a commodity. One of the most important elements to look at in a trading position is the profits and losses (PnL). The PnL reveals the daily change in the value of the trading position. Another element is called exposure, which is the quantified amount of positional capital that is exposed to the market forces and hence could incur losses due to the gyrations in the markets. In the process of assessing risk in a trade, it is important to foresee as many outcomes as possible. To do this, the framework creates different scenarios of what the future may look like and propose trading strategies that mitigate or eliminate the risk associated with the existing trading positions.
The pricing framework plays an important part in the quantamental framework as it helps different trading positions to be priced based on changing market data. When working with complex instruments such as options and derivatives, it requires a pricing model that determines the fair value of such contracts. Since these complex instruments depend on underliers and are driven by parameters such as volatility and interest rates, they cannot be priced directly. A quantitative derivatives analytics library must be used to determine the pricing of these instruments.
Time-series analytics is a fundamental toolkit used by the quantamental framework in order to create indicators. One of the most important aspects of the quantamental framework is to be able to sift through data and make sense of it all. The time-series analytics does just that by applying techniques such as moving averages, regression analysis, trend analysis, and seasonality analysis. These techniques create meaningful targeted information from a vast input of data sourced.
Visualization Module:
What is the most effective way to present data to users?
How users consume data is a very important aspect of how the workbench operates. The quickest way a user is able to intake vital information is through dashboards. A very useful feature of this module is the templates with which you can create customized dashboards. Using this feature, users can customize specific dashboards to display information such as where they are making money, losing money, various trends etc. Dashboards, though quick and efficient, do not always provide information in great detail. That is why the creation of reports is another feature of this interface. These reports may contain all the relevant information regarding a certain trading position, allowing the trader to make the most educated decisions moving forward.
One of the most efficient ways to organize and display various financial datasets on the platform is through the charts. The visualization module contains an entire charting framework dedicated to displaying information in versatile charts. Some important features that the charting framework includes are: advanced filtering, conditional formatting, dataset alignment, multi-series plotting, multi-plot graphs, multi-style view, and refresh policies. These features allow the data to be displayed in an organized manner even with different inputs. The various elements of the charting framework allow charts to be adaptable and clear.
Collaboration Module:
Why is collaboration important in this quantamental trading platform?
What are different features that can be used to maximize communication efficiency?
Quantamental trading includes many different moving parts. Traders will want to collaborate with others in order to create the best trading strategy. Oftentimes, it is very hard to share trading related data as they come in various forms. These templates/graphs have to be reconstructed when sent through other platforms. The easiest way for traders to collaborate is to be able to communicate and share information directly on the trading platform. There are four elements to this module:
- Market data sharing
- Analysis sharing
- Templates sharing
- Messaging
Market data is crucial information to be shared as it is the most important information to refer to when creating a trading strategy. Sharing constructed data curves, derivations, transformations can also be extremely helpful. Messaging is another very important feature of the collaboration module as it allows people to chat with their peers in real time. The ability to have access to the same information and work together to create the optimal trading strategies is the primary goal of this module. Having a collaboration module allows traders to stay on the same platform throughout the overall trading process.
Conclusion
Quantamental trading is a proven successful trading methodology that should be made accessible to more market participants. Cetin Karakus’ Korteks gives traders the opportunity to access the quantamental framework in their own trading endeavors. As explained above, the different modules of the Korteks Workbench provide an unmatched user experience where all the necessary functions can be found within a single platform. From the advanced quantamental framework to the convenient collaboration features, Korteks Workbench’s design and functional architecture is packed with a lot of ideas and features that could help the interested parties with what to expect from a quantamental trading platform.
Written by Steven Chen
