On January 17, Yueran Technology released an artificial intelligence trading system in Shanghai, the “Ai (AI) Trading†system. At the same time, it also announced the first private equity fund based on the Love Trading System, the Artificial Intelligence Trading Fund. Established.
According to the interface news reporter, Ziwu Yueran artificial intelligence No. 1 securities investment fund, this is called "China's first artificial intelligence system to manage operating funds" investment field involving stocks, options and commodity futures, and has been in China's securities investment fund industry The association completed the registration and will be put into trial operation at the end of this month. The fund has a scale of 30 million yuan.
Ye Chengyu, chief scientist of Yueran Technology, introduced that the core algorithm of the trading system uses the architecture of the most advanced artificial intelligence technology, AlphaGo, which consists of two parts: the valuation network and the decision network. The valuation network is used to predict the future changes and probabilities of the stock market, and the decision-making network seeks the best ordering time and the number of shares placed each time through deep reinforcement learning according to the customer's own position preference. Similar to AlphaGO, the valuation network of Love Trading also searches for trading signals from high-frequency data through convolutional neural networks without any human involvement.
Zhu Jiaqi, CEO of Yueran Technology Co., said at the press conference: "2018 will be the first year of AI trading. The fund is only a test product. In the future, they will share this AI model with partners in the industry. The investment is a game of capital giants with high thresholds, high risks and opaque information, so AI will provide us with diversified, customized and personalized services."
As the founder of Meridian Investment and Yueran Technology, Zhu Jiaqi has 10 years of quantitative trading experience in the US, Europe and India markets, including stocks, futures and options. In addition, the chief scientist of the main team member of Ziwu Investment, Ye Chengzhen, in the field of deep learning, CTO Zhu Xi has a senior resume in the US capital market in network mobile security.
Zhu Jiaqi does not shy away from the huge returns from quantitative trading. At the press conference of the day, Zhuo Qi, CEO of Yueran Technology recalled the incident called “Guangda Oolong Finger†five years ago. He said that “the mistakes of Everbright Securities Traders have caused huge transactions in more than 100 stocks, but I Fortunately, I seized a short trading opportunity through quantitative trading operations and made a lot of money on the day. I bought an Aston Martin to reward myself that day."
From quantitative trading to artificial intelligence, the expected profitability and team support have become the driving force for Zhu Jiaqi's team to move from quantitative trading to artificial intelligence. In his view, adding the characteristics of AI machine self-learning in quantitative transactions can help the quantitative model adapt to different market environments, dig deep into the new stock selection factors and the nonlinear relationship between factors and stock returns, and improve the model prediction ability. Sharply capture undiscovered market opportunities.
However, some quantitative hedge funds believe that machine learning is just one of many methods of observing market statistics. The so-called artificial intelligence is not unusual. The most deadly thing is that if a model proves to be effective and popular, it will quickly invalidate the model.
A recent example of the US AI fund field is also in sight.
On October 18, 2017, two American companies, EquBot LLC and ETF Managers Groupl, jointly launched the world's first ETF, the AI ​​Powered Equity ETF, which uses artificial intelligence and machine learning to invest. AIEQ performed very well when it first came out, but the performance of the fund was not satisfactory, and there was even a big deviation from the performance of the market index. For example, on October 25 and November 7, the market performed weakly. The S&P 500 index fell 0.47% and 0.02% respectively. The Nasdaq Composite Index fell 0.52% and 0.27% respectively. The day's decline reached 1.11% and 1.09%.
Even if there is progress in model construction and the machine is more involved in the formulation of strategies, artificial intelligence faces two major problems: First, the model prediction is too accurate (over-fitting) to be difficult to achieve in the short term; The decision of artificial intelligence is not transparent, and it is difficult for ordinary people to understand the operational logic of the machine. In fact, from all the public information that AI is deeply involved in the investment, there are no winners.
Adapter Ring,Ring Camera Adapter,Lens Adapter,Camera Adapter
Shaoxing Shangyu Kenuo Photographic Equipment Factory , https://www.kernelphoto.com