AI Asset Management Report (Part 3)
A three-part series that detail how AI can be a viable innovation driver for the asset management industry.
— — — — — — — — — — — — — — —
ETFs exhibited an outstanding performance over the past year since their inception.
Performance data quoted represents past performance and is no guarantee of future results. Investment returns and principal value of an investment will fluctuate so that an investor’s shares, when redeemed, may be worth more or less than the original cost. Current performance may be lower or higher than the original cost. Returns for periods of less than one year are not annualized. For most recent month end performance visit https://qraftaietf.com. Returns are determined based on the midpoint of the bid/ask spread at 4:00pm Eastern time, when the NAV is typically calculated. Market returns does not represent the returns you would receive if you traded shares at other times. Market Price: The current price at which shares are bought and sold. Market returns are based upon the midpoint or the last bid/ask spread at 4:00pm Eastern time. NAV: The dollar value of a single share, based on the value of the underlying assets of the fund minus its liabilities, divided by the number of shares outstanding. Calculated at the end of each business day.
What’s more important than the outsized returns are that these AI ETFs were created based on strategies automatically extracted by an AI system.
Through the automated data processing system and deep learning-based automatic factor/strategy extraction system, these AI ETFs could be produced and tested in less than a month. With this AI system, Qraft Technologies reached a productivity level where five engineers can produce at least five active index ETFs every month.
However, the investment strategy created with a DL (deep learning) model learns new data every day, and the weight produced by the neural network changes as per the new data set, thus extending the life of the investment strategy. (Even deep learning models are subject to new model engineering if and when new data sets are introduced).
AI-Enhanced ETF Case
In May 2019, a rather unexpected event occurred in the US financial market. Qraft Technologies was able to list two active index ETFs (NYSE: QRFT, NYSE: AMOM) operated by AI in the New York Stock Exchange.
QRFT is an active index ETF that aims to enhance the S&P500 Index, and AMOM aims to enhance the S&P500 momentum index. Both AI managers, it usually takes 6 to 12 months, with 5 to 10 staff to launch one Active Index ETF.
Automated Strategy Extraction by AI
The quant fund operation is largely composed of data processing, strategy research, and order execution.
In the data processing stage, complex financial data is processed to remove any biases or incorrect information for better simulation. At the strategy research stage, numerous research analysts are employed to find the alpha source and the optimal portfolio, using pre-processed data from the previous stage. Sophisticated order execution (trading) is essential to minimize the market impact (and the transaction cost) when processing large scale trading orders.
Qraft Technologies has built an AI-based system for all these three essential steps to mass-produce active index ETFs.
The overall system structure is as follows.
The most important element in creating this automated system is how simple you can construct an environment where you can test the complicated models as often as you can. Automation can be achieved only if this condition is met.
1. Data Pre-Processing System: Qraft Kirin API
It is very difficult to use raw financial data from vendors such as S&P Global and Refinitiv (formerly Thomson Reuters). Elimination of survivorship bias (e.g. treatment of delisted stocks), look-ahead bias (e.g. correct treatment of revised financial statements) and accurate processing of corporate events such as M&As, private and re-listing rights offerings, take a tremendous amount of time to pre-process. However, Qraft Technologies’ data processing systems can automate the pre-processing tasks through parallel computations accelerated by GPU.
Qraft data processing system does not just collect and store data, but it also enables to test the investment universe from various angles with pre-processed data sets. For example, you can define the investment universe as “companies that have not passed two months since the latest public disclosures of patents,” and test the pertinent investment strategy with just a few lines of python codes. (In other words, Qraft’s data processing system allows us to automatically find an investment strategy). This system is packaged as API and will soon be commercialized as a standalone business solution.
2. Strategy Extraction System: Qraft Alpha Factory
The automatic extraction of investment strategies is the core function of Qraft’s AI research system. The vast amount of search universe multiplied as [investment universe (u) x degrees of freedom] * [data (x) x degrees of freedom] * {function form (f) x degrees of freedom] is much more massive than the search universe of a Go game. In such vast search universe, a well-engineered deep learning model has exhibited consistent results in narrowing the probable candidates and automatically back/forward testing the candidates to extract an investment strategy. The automatic extraction of investment strategies through deep learning is composed of two modules:
A. Factor Factory
Factor Factory applies AutoML technology to automatically search basic patterns that have the potential to bring excess returns. Using one NVIDIA DGX server, Factor Factory can produce more than 10 patterns (factors) per day without human intervention. It consumes a lot of electricity, but at least you won’t have to hire high payroll managers and not worry about employees leaving the job.
B. Strategy Factory
Factors found through Factor Factory are not independent sources, so it is more appropriate to construct non-linear combination models as opposed to linear combinations. Strategy Factory extracts the investment strategy with a non-linear asset price model by non-linearly combining factors extracted automatically from the Factor Factory. The more factors accumulated through Factor Factory, the more sophisticated asset price models that Strategy Factory can generate.
C. The Case of Value
A lot of research has been conducted on the value factor about how the performance of value investing is falling noticeably than ever before. Several papers have been published to support that the era of value investing has not yet ended but instead, the value measurements designed for traditional structures might be wrong.
Traditionally, value investing compares book values to market capitalization. But under the asset debt accounting system, only tangible assets are accounted for in assets and intangible assets are often treated as costs due to uncertainty in measurement. The fact that they are not properly reflected has emerged the cause of the value factor failure. The study also analyzes that intangible assets have become a more important factor in the actual value of the company due to the development of information technology. (It’s easy to imagine which of the Amazon’s warehouses (tangibles) or the logistic systems (intangibles) is more important in measuring the firm’s value).
Qraft Technologies extracted more accurate intangible asset measures (i.e. function f) through the Factor Factory using R&D cost, marketing cost, SG&A related to restructuring, and patent issuance as input data. This way, we greatly improved the performance of value investing.
3. AI Execution System (AXE)
The history of creating order execution system through computer algorithms is quite extensive. However, most automated order execution systems use predefined rule-based algorithms such as VWAP, TWAP, and IS. JP Morgan Chase announced the world’s first AI order execution system, LOXM, which applies deep reinforcement learning technology to tick data of individual stocks. Shortly after, Goldman Sachs announced that it has also developed and tested AI order execution systems.
Qraft Technologies also developed an AI order execution system that applies reinforcement learning technology to stock tick data. It’s the world’s first case of commercialization as Shinhan Investment Corporation had implemented the system. (JP Morgan and Goldman Sachs use their AI execution system for internal use only). Unlike JP Morgan’s LOXM, Qraft Technologies was able to apply the latest agent-based reinforcement learning model to significantly increase the performance of order executions.
In 2018, NVIDIA, Shinhan Bank, KOSCOM, and PwC sponsored a competition called the AXE Challenge with a total grand prize of $100,000 dollars. The challenge was essentially for human traders from well-known securities firms to compete against AXE in buying a portfolio at a cheaper price. The portfolio was randomly generated by KOSCOM. The competition lasted for five days and was broadcasted live on TV. At the end, the AXE system had beat the human traders by a large margin.
AXE explores the optimal order execution strategy by learning patterns from tick data, which includes the price and transaction volume as well as its history. When AXE is applied, it’s possible to improve the return of active index funds by minimizing the transaction cost of all financial products. It especially works well with super-sized or small cap mid cap funds that could have a larger market impact.
Qraft Technologies recently announced that it will package the AI order execution system AXE in the form of a cloud-based API with Microsoft. This means that by the end of the year, once the cloud solution is finalized, anyone can apply Qraft’s AXE to their services through cloud subscriptions and API connections.
With the implementation of AI order execution technology, messenger platforms can simplify the complex UI system for stock trading and create a tremendously easy and intuitive service with just mere voice or words. To put simply, with messenger platforms, you can just say or type “Please buy me $1,000 dollars of Tesla stocks for the next week” and the process will be finished. No longer will investors have to worry about when to buy or sell while looking through MTS. If this type of UI/UX becomes widespread, complex MTS that requires order price, market price/specified price selection, etc. Will become a relic of the past.
NVIDIA, the world’s largest GPU company, recently selected Qraft Technologies among 30 AI startups worldwide for its NVIDIA Inception Premier program that supports marketing and sales. Qraft is the first startup in South Korea to be chosen and the only one in the world from the financial sector. (Other selected companies are mostly unicorn startups or famous AI firms that have been acquired by Apple and other blue-chip companies.)
NVIDIA’s selection of Qraft Technologies in the Inception Premier program may be a sign of anticipation that if AXE penetrates the global financial market, NVIDIA’s GPU will also be sold substantially. With AI order execution, the more customers order, the more real-time simultaneous processing capability is required. And AXE requires a lot of parallel GPU computing power in the learning process and inference process.
Is this the beginning of another 1990?
Back in 1990, a few pioneers had found Strategy C with the arrival of data and computers. In 2020, another group of pioneers have the potential to achieve new levels of innovation in the asset management industry. This is mainly through the “automated alpha” process with the help of AI technologies.
AI-driven asset management model
AI-driven asset management models only require a team of data engineers for discovering and dealing with new data sources as well as a team of AI engineers to build deep learning models that bring highly efficient strategies. The cost is just limited to accessing new tick data sources for order execution. No matter how many ETFs will be developed and operated, there will be no extra added costs other than the server expansion. And the best part is, this model is being used as we speak.
The team that can introduce efficient strategy extractions (alpha factory) that human researchers simply cannot achieve, would be the right candidate to innovate the investment industry. A new pattern of investment strategy (e.g. dynamic investment universe, just like surprising AlphaGo’s next move that was beyond human perception) will arise and the same team that makes active index funds will seize a significant portion of the entire ETF market. Perhaps this might be the very beginning of innovation to unlock inexpensive alpha with AI technology in the growing ETF markets.
About Qraft Technologies, Inc.
Qraft Technologies aims to innovate the inefficiencies of today’s asset management industry by leveraging AI technology. From data processing to alpha research and portfolio order execution, our goal is to provide you with high level of alpha at low cost.
We automate complex financial data preprocessing, accelerate it through parallel computing, and automatically search for the alpha factors through AutoML technology in a well-established simulation environment. Using the alpha factors found in this way, a deep learning based Deep Asset Pricing Model is created through the Strategy Factor as per the fund universe defined for a fund concept. The final portfolio made through this model is efficiently executed with the order execution engine AXE, which is based on reinforcement learning.
Qraft Technologies’ AI ETF lineup produced by the systems above are listed on the New York Stock Exchange from May 2019. AI ETFs minimize human intervention and is 100% managed with an artificial intelligence system. As of 09/30/2020, QRFT and AMOM have outperformed their benchmark indices (S&P500, S&P500 Momentum Index) by more than 10%p since inception over the past year. For the standardized performance and expenses of QRFT click here and AMOM click here.
— — — — — — — — — — — — — — — — — — — —
GPU — Stands for Graphics Processing Unit and refers to a programmable processor specialized for rending all images on the computer’s screen.
API — Stands for Application Programming Interface and refers to a set of programming code that enables data transmission between one software product and another.
AutoML — Stands for Automated Machine Learning and refers to the process of automating the use of machine learning to solve real-world problems.
SG&A — Stands for Selling, General, and Administrative and refers to the costs of running a business.
UI — Stands for User Interface and refers to the series of screens, pages, and visual elements that enable a person to interact with a product or service.
MTS — Stands for Make to Stock and refers to a traditional production strategy that is used by businesses to match the inventory with anticipated consumer demand.
Important Information
The performance data quoted represents past performance. Past performance does not guarantee future results. Current performance may be lower or higher than the performance data quoted. The investment return and principal value of an investment will fluctuate so that an investor’s shares, when sold or redeemed, may be worth more or less than their original cost. Returns less than one year are not annualized. Performance data current to the most recent month end may be obtained by visiting https://qraftaiet.com.
- For a complete list of holding for QRFT click here.
- For a complete list of holding for AMOM click here.
- This material was prepared for informational purposes and cannot be used for the purpose of soliciting the sale of financial investment products such as funds.
- This document contains the contents of the patent-pending or registered by Qraft Technologies, Inc.
- Short-term performance, in particular, is not a good indication of the fund’s future performance, and an investment should not be made based solely on returns. Because of ongoing market volatility, fund performance may be subject to substantial short-term changes.
Qraft AI ETFs aim to create excess returns over traditional market indices taking advantage of cutting-edge artificial intelligence (“AI”) technologies.
The ETFs are designed to enhance quantitative investment strategies with AI technologies and seize investment opportunities from nourishing data that human portfolio managers may not capture.
Investing involves risk including possible loss of principal.
Investors should consider the investment objectives, risks, charges, and expenses carefully before investing. For a prospectus or summary prospectus with this and other information about the Fund, please call (888) 123–4589 or visit our website at https://qraftaietf.com. Read the prospectus or summary prospectus carefully before investing. For a copy of the prospectus, click here: https://qraftaietf.com.
Distributed by Foreside Fund Services, LLC
Investing involves risk, including loss of principal. The Funds are subject to numerous risks including but not limited to: Equity Risk, Sector Risk, Large Cap Risk, Management Risk, and Trading Risk. The Funds rely heavily on a proprietary artificial intelligence selection model as well as data and information supplied by third parties that are utilized by such model. To the extent the model does not perform as designed or as intended, the Fund’s strategy may not be successfully implemented, and the Funds may lose value. Additionally, the funds are non-diversified, which means that they may invest more of their assets in the securities of a single issuer or a smaller number of issuers than if they were a diversified fund. As a result, each Fund may be more exposed to the risks associated with and developments affecting an individual issuer or a smaller number of issuers than a fund that invests more widely. A new or smaller fund’s performance may not represent how the fund is expected to or may perform in the long term if and when it becomes larger and has fully implemented its investment strategies. Read the prospectus for additional details regarding risks.
QRAFT AI-Enhanced U.S. Large Cap ETF: Companies in the health care sector are subject to extensive government regulation and their probability can be significantly affected by restrictions on government reimbursement for medical expenses, rising costs of medical products and services, pricing pressure (including price discounting), limited product lines and an increased emphasis on the delivery of health care through outpatient services.
QRAFT AI-Enhanced U.S. Large Cap Momentum ETF: The Fund is subject to the risk that market or economic factors impacting technology companies and companies that rely heavily on technology advances could have a major effect on the value of the Fund’s investments. The value of stocks of technology companies and companies that rely heavily on technology is particularly vulnerable to rapid changes in technology product cycles, rapid product obsolescence, the loss of patent, copyright and trademark protections, government regulation and competition, both domestically and internationally, including competition from foreign competitors with lower production costs. Technology companies and companies that rely heavily on technology, especially those of smaller, less seasoned companies, tend to be more volatile than the overall market.
QRAFT AI-Enhanced US High Dividend ETF: Securities that pay dividends, as a group, may be out of favor with the market and underperform the overall equity market or stocks of companies that do not pay dividends. In addition, changes in the dividend policies of the companies held by the Fund or the capital resources available for such company’s dividend payments may adversely affect the Fund. In the event a company reduces or eliminates its dividend, the Fund may not only lose the dividend payout, but the stock price of the company may also fall.
