Building a new fund company with artificial intell

  • Detail

Building a new fund company with artificial intelligence

Zheshang fund

general manager Nie Jinjin

China fund retribution is particularly good

recently, the industry's first artificial intelligence fund based on in-depth learning - Zheshang intelligent industry preferred fund was approved, and Zheshang fund is currently arranging its issuance plan

it is reported that this fund will be handled and competed by hundreds of artificial intelligence "machine fund managers", and ultimately the human fund manager will choose the winning artificial intelligence portfolio. This is just the beginning of the overall AI layout of Zheshang fund. Zheshang fund will continue to layout AI products and fully enable active equity and fixed income investment with AI technology. In the future, they also plan to extend AI technology to all aspects of the company

within Zheshang fund, fund managers are divided into artificial intelligence "fund managers" and "human" fund managers. In order to enable artificial intelligence to invest and compete with human beings, they have made efforts for four years. Recently, China fund daily interviewed the general manager of Zheshang fund and the artificial intelligence team of Zheshang fund to learn more about the company's characteristic development path

artificial intelligence system in the real sense

Zheshang fund is preparing to issue this investment research result based on hundreds of artificial intelligence "fund managers". In the model, hundreds of artificial intelligence (,) based on different investment logic and methodology form their own investment decisions according to their learning of human investment methods, and then compete with each other in hundreds of different investment decisions and configuration schemes, and finally produce the optimal scheme

"at present, we have more than 300 'fund managers' of artificial intelligence. The number will increase, and those who do not do well will be eliminated." About 30% of the non frozen soil in the world is acid soil. Dr. Xiang Wei, the manager of Zheshang intelligent industry, introduced that every time the team finds a new investment logic and method, it can form a new "fund manager" and let new ways participate in the learning of artificial intelligence

what the AI team mainly does is to let these AI "fund managers" learn the excellent investment methods of human beings. They use big data, methodology research, in-depth industry research and other data achievements to train these AI "fund managers" so that they can continue to learn and iterate quickly. "These AI 'fund managers' keep learning the excellent investment methods of human fund managers under their own factor settings, which is their advantage." Xiang Wei said

"from the current operation, it is possible that every AI 'fund manager' can learn 70% of human investment styles. However, since they can learn all (source: Chinese household appliances) styles and compete with each other, it is easier for them to maintain stable performance when the market switches between various investment styles." Xiang Wei said, "if a fund manager scores 70 points in all styles, the performance of the fund manager should be stable and upward for a long time, but it is difficult to do so."

Dr. chaxiaolei, another proposed fund manager in the intelligent industry of Zhejiang merchants, said that from the results of artificial intelligence operation in the past, the investment effect of artificial intelligence "fund managers" was relatively general in the early stage of operation, but with the increase of operation time, artificial intelligence "fund managers" showed the effect of rapid learning and rapid iteration, and their investment performance continued to improve

build fund companies around artificial intelligence

for a long time, many fund companies have been introducing artificial intelligence means. Some companies are used to improve the level of investment research, some companies focus on improving the level of intelligent investment advisory services, and some companies are used to improve the operation efficiency of the middle and back offices

for Zheshang fund, artificial intelligence is not an auxiliary tool, but the core of the enterprise. "Since 2015, Zheshang fund has started a series of intelligent investment infrastructure construction." Nietingjin said, "in the past two years, we have been building some low-level things. Based on big data and AI technology, we have realized data signalization, signal strategization, strategy modeling and model visualization in the investment field, and finally achieved human-computer interaction to improve the model operation effect."

in order to achieve this goal, the IT staff of Zheshang fund has increased three times in just two years. At present, the investment and research background of Zheshang fund is certainly very special in the industry. Most of the artificial intelligence teams are doctors with science and engineering and computer backgrounds. In addition, corresponding data officers are also configured in terms of active equity and fixed income to connect human beings with artificial intelligence

for artificial intelligence "fund managers", in order to improve their investment ability, they need to have enough excellent investment methods to learn. In the past 20 years of domestic investment experience, although there have been a large number of public data for learning, they also need a large number of internal data support from fund companies in terms of transaction and other details. To this end, Zheshang fund is preparing to establish an intelligent equity investment department, which will combine active investment and intelligent investment team in the future. On the one hand, it will strengthen the integration of human and machine intelligence, on the other hand, it will also promote the AI team to deeply empower the hi team (active equity investment), and improve the internal cooperation efficiency of the company

"To be honest, in a large fund company, it may be difficult to promote the transformation of artificial intelligence and integrate AI and Hi, but we are a start-up company and innovative company. Most of our fund managers and researchers are trained internally. Many investment researchers have started to cooperate with AI since they entered the investment research work. Their work direction and foothold are essentially different from those of traditional fund companies, which is also our advantage 。” Nietingjin said

Nie Jinjin wrote the progress of human-computer interaction into KPI assessment. Every year, investment research personnel need to quantitatively provide investment research support for AI, and the AI team also needs to quantitatively help the investment research team improve the efficiency and ability of active management every year

(down to version 06) (up to version 04)

within Zheshang fund, many details are set up for human-computer interaction, even seating arrangements. The seat of the AI team is in the middle of the investment research office, which facilitates the most direct communication between the team and the surrounding researchers and fund managers

according to nietingjin, since the investment research team is used to the "technological empowerment" of the AI team, they often take the initiative to communicate with the AI team about some problems encountered in investment research, and the AI team often helps them solve problems through various technical means. "Now they communicate with each other a lot. This has become a new work habit."

in particular, a complete set of artificial intelligence systems are built unevenly.

in the past three years, Zheshang fund has initially built three sets of artificial intelligence systems: ficc system, Lucy system and AI beta system

ficc system is equivalent to alphago (alpha go) of fixed income, which excavates the driving force of bond market from the perspective of globalization and cross market; From the credit factor model, the quantitative credit analysis model and the pricing method combining machine learning and manual work are displayed for investors, and a knowledge map is built to mine the negative

nietingjin introduced that on the road driven by science and technology, they also made a choice between alpha go and alpha zero at first, and finally found that the way of using alpha zero in investment was not ideal. In this regard, Nie Jingjin sighed: "artificial intelligence needs continuous investment, because some investments will lead to the process of trial and error. Only with continuous investment can we 'try to be right' after trial and error, and never make mistakes again."

lucy system is a core stock investment system based on human-computer interaction. Relying on digital and automated information tools, Lucy system proactively grasps investment opportunities, tracks the target situation in real time, scientifically monitors stock portfolio risks, greatly improves investment efficiency, and pursues higher excess returns

ai beta system is a set of smart beta system, which disassembles the stock market, tracks effective risk pricing factors, and examines smart beta from the perspective of AI, making the risk more transparent and the process more scientific

since April this year, Zheshang fund has begun to try to export some systems to various business systems of Zheshang, and even to external customers for trial use

"we want to let customers know how to better configure smartbeta and how to build a long-term investment perspective instead of pursuing short-term returns." Nietingjin said that now some core customers in the market have begun to pay attention to their achievements

"the products to be issued this time are planned to be issued in (), and CMB is at the forefront of the market in technology and finance. They also welcome our innovation." Nietingjin said

in Nie Qianjin's plan, the artificial intelligence strategy of Zheshang fund has just begun, and now it is only the first step. He believes that in the next three to five years, with the breaking of the rigid exchange, net worth public funds will usher in a golden development period

what Zheshang fund will do next can be summarized as four aspects: adhering to long-term principle, focusing on development strategy, realizing professional division of labor, and accelerating system integration. They will continue to improve the product line layout of artificial intelligence. In the future, they will add artificial intelligence investment products in the absolute income field and the fixed income field. At the same time, artificial intelligence means will be applied to the active management of fund managers. In the future, they will add AI to their investment advisory business to improve the level of investment advisers, and gradually extend AI from the front desk to the middle and back offices

sharpening a sword in four years

aiming at the pain points of the fund industry

before joining Zheshang fund, Nie Jinjin was managing investment research, where he met the then general manager of Boshi fund. Xiao Feng is a financier with a scientific and technological complex. He emphasized scientific and technological investment when he was in Boshi fund

2015, the two met again in Zheshang fund, and started the "second venture" for the common ideal of science and technology finance

"do it or not?" Nie Tingjin did not hesitate. At that time, the child was already studying in Shenzhen, and coming to Zheshang fund meant moving the family to Shanghai. Finally, Nie Jinjin decided to take the risk. "At that time, we decided to come to Zheshang fund, which can be said to be supported by idealism." He recalled

as early as 2013, Xiao Feng realized that with the development of big data and the emergence of AI technology, great changes might occur. They wanted to build a future fund company on the platform of Zheshang fund

"for a long time, the fund industry has been using personal investment ability to realize the value of the company and the industry, but the excess returns obtained by relying on personal ability are often difficult to trace and copy." Nietingjin said that compared with other industries, the pace of reform in the Chinese and foreign financial industries is relatively slow

from the very beginning, they decided not to follow a large and comprehensive fund development path, but to build a brand by seeking unique advantages. "What we need to do is to adhere to the development direction of high-quality products and differentiation, adhere to the long-term principle, and lay out strategic products." Nietingjin said that they hope to use scientific and technological means to better solve the pain points of the fund industry

Nie Jinjin introduced that first of all, they should pursue a better investor experience. "At that time, we proposed that individual heroism and smallCold Rolled Steel Sheet 201 304 430 2mm 3mm 5mm Thick Half Hard Stainless Steel Sheet for Wall and Ceiling Decoration with ISO: 9001 Certification
Inspur special computer services Expo
Current situation and development of China's optical fiber and cable and optical passive components industry
On December 22 the TDI commodity index was 6067

Copyright © 2011 JIN SHI