Jiukun Investment: A group of "thermostatic animals" that use formulas and algorithms to beat the market

Text | Yu Lili

Edit | Liu Jing

In the capital markets, there is such a hidden group: they are not from a financial background, do not blindly believe in subjective judgment, and their only weapon is formulas and algorithms. Since they are always tight-lipped about trading strategies, most people have a hard time even understanding exactly what they are doing. But it turns out that they tend to be in the minority that outperforms the market.

They are the "quantitativeists" among investors. Simply put, quantitative investing is a data-driven approach to investing. Different from subjective investment, quantitative investment is to find investment laws that can bring excess returns from massive historical data, and build quantitative models to guide investment according to these laws.

On Wall Street, quantitative fund practitioners are often referred to as "QUANT" – just as their job is to overcome emotional fluctuations in trading, their personalities are often unhurried, a group of emotional thermostatic animals.

After 2018, 8 of the world's top ten hedge funds have been quantified. In the United States, where the quantitative investment industry is mature, the largest quantitative fund has exceeded 100 billion US dollars, accounting for about 30%-40% of all securities private equity funds.

Let's take a look at a report card of Mons, a benchmark in quantitative investing: It is estimated that in the 30 years from 1988 to 2018, Simmons' Medal Fund, called Renaissance Tech, achieved an annualized compound rate of return of 39.1%. Even in 2020, when "black swans" are frequent, the annual after-fee income is as high as 76%. This achievement, even Buffett, Soros, and even Dalio, who was "canonized" in previous years, are beyond expectation.

When we pull our eyes back to China, we can see such a new trend: in 2020, huge amounts of funds poured into quantitative private placement, and quantitative private equity ushered in an explosive scale increase period, and the head effect became more and more obvious. According to the data of the private placement ranking network, as of January 21, 2021, the number of quantitative private placements of 10 billion yuan has increased to 12. From 2017 to 2020, the scale of quantitative fund management has increased by four times, exceeding 500 billion yuan. The share of all securities private equity funds rose from less than 5% to 15%.

So, what kind of Chinese evolution is an imported product like quantitative investment experiencing? Will the Simmonsian miracle be replicated in China? Compared with the United States in mature markets, is there a chance for China's quantitative investment to overtake the curve?

Jiukun Investment may be a representative sample to observe the evolution of quantitative investment in China.

As one of the earliest quantitative investment institutions established in China, Jiukun has witnessed several rounds of market cycles such as the pioneering, dormant and scale increase of China's quantitative community. At one time, together with Magic Square Quantification, Ruitian and Zhicheng Zhuoyuan, it was known as the "Four Heavenly Kings" in quantitative private placement. In 2019, Jiukun Investment ranked among the ranks of 10 billion quantitative private placements, and the current management scale has exceeded 40 billion.

<h3>The daily life of a wide range of guests </h3>

As the name suggests, everything in the quantitative fund can be quantified, and Jiukun also perfectly explains this - whether it is investment research, trading, or team building, incentive mechanism.

There is even a joke circulating inside Jiukun: once a ping-pong table in the lounge area is accidentally moved, it will be quickly "complained" because "the slope has changed, which will directly affect the winning rate.".

The mathematician-turned-Mons revolutionized the way financial markets are played by hiring a group of mathematicians and scientists who know nothing about Wall Street investment theory. And Jiukun's team also has almost no financial background, and their professions are strange: there are from mathematics, computers, statistics, as well as condensed matter physics, quantum mechanics and even astronomy, and their daily work is to write hundreds of thousands or even millions of lines of code based on massive data such as historical prices.

Just as their job is to use modern statistical and mathematical methods to overcome the erosion of market sentiment, they are also masters of controlling emotions on a daily basis. In Jiukun, you rarely see them get angry, and a somewhat exaggerated statement is that a standard wide customer should do "mood fluctuations up and down no more than 3sigma" (generally 3sigma outside of the known small probability event).

In addition to table tennis, another pastime of their daily life is to play Poker. Because of the familiarity probability and game, they are often good at playing cards. When they're at the table, you'll believe that the story of a gifted student of an MIT math professor winning a blackjack game at various Las Vegas casinos in the Hollywood movie "Blackjack" is not empty. In fact, the professor's prototype is Edward Thorpe, an earlier quantitative god than Mons, the head of the first quantitative hedge fund in history, the "Convertible Hedge Partnership Fund."

In addition, this group of wide customers of Jiukun are dressed in ordinary clothes and have a simple life. A staff member of Jiukun told 36Kr that colleagues are accustomed to carrying backpacks, wearing sports sweatshirts, and low consumption. Just as their companies are accustomed to defining themselves as tech or tech companies rather than investment firms, they themselves are closer to geeks than to the investment elite.

It's just that their hair is still very lush because they are still young enough. Jiukun's entire team, post-85s and post-90s accounted for the mainstream. Quantitative investing is always new: the market is dynamic and changing, all "old" strategies will be reduced in utility due to the gradual understanding of the market, and only newer, iterative low-correlation strategies will be valuable. Therefore, compared with the subjective investment field that is getting older and more popular, the quantitative team is also younger.

The same goes for the people who are in their palms. Unlike the stock subjective long private placement, which is more than 70, the head quantitative private placement is basically dominated by the post-80s manager with the background of returnees. Most of them have worked for some international quantitative giants such as World Quant (Shikun Investment), Two Sigma (Sigma), D.E. Shaw (Deshao Group), Citadel (Castle Investment), etc., and are the first quantitative investors in China to sinicize the investment methods and operating systems of Wall Street hedge funds.

The trajectory of Wang Chen's life trajectory invested in Jiukun is also roughly the same. Jiukun Investment is adjacent to the campus of Tsinghua University, where Wang Chen studied for a bachelor's degree in mathematical physics and a doctorate in theoretical computer science. The default life path should have been to go to IT companies such as Google and Microsoft, and live a clear and clear life. But occasionally, he found quantification to be a more exciting job: "Using statistical and technical methods to crack the laws of the market is a bit like hacking."

So, Wang Chen, who graduated with a doctorate in computer science in 2008, went to the international quantitative giant World Quant (Shikun Investment), where he met his later partner, Yao Sicong from Peking University.

<h3>Discipline of investment</h3>

Quantitative investing may seem mysterious, but it is not ancient. Because it requires specific soil.

Even gods like Simmons were the object of ridicule at first. Because no one believes that computers can be invested like humans, but with the development of computers, the impossible becomes possible.

Domestic markets are more restrictive: the short history of financial markets, for example, means a lack of financial instruments, derivatives, and the amount of valid data. More critical is the stubbornness of the market concept: even as early as 2004, China had the first public quantitative fund Everbright Prudential Quantitative Core Fund, before and after the 2007 and 2008 financial crises, some returnees began to bring quantitative investment strategies back to domestic public funds, but in the end it only stirred up a little splash, not enough to make the entire market have a clear understanding of quantitative investment.

Until April 6, 2010, the first stock index futures contract in China, the CSI 300 stock index futures, was officially listed. This means that quantitative funds finally have feasible hedging tools, and quantitative investment strategies such as alpha strategies and stock index futures arbitrage have begun to play out, so 2010 is known as the first year of China's quantitative investment. A group of overseas quantitative investors began to smell opportunities and returned to China to start a business. Wang Chen and Yao Cong were among them.

At that time, one of the more common choices after returning to China was to go to private equity or brokerages, and Wang Chen and Yao Cong decided to do their own team because they wanted more freedom.

In April 2012, Jiukun Investment was officially established. Within two years, Ruitian Investment, Mingtun Investment and Lingjun Investment were also established.

At the beginning of the business, because commodity futures, including stock index futures, came with leverage, the efficiency of self-operated funds would be higher, so Jiukun was mainly based on the self-operation of futures CTAs that could be traded with leverage. In 2014, Jiukun Investment applied for the record through the private fund manager and began to engage in private fund asset management business.

In Wang Chen's view, from 2010 to 2015, it was the 1.0 era of quantitative private equity in China: the quantitative investment model was relatively crude at that time, although many investment methods were crowned as quantitative, but there were very few companies that really had the correct quantitative investment system and framework. In 2015, stock index futures, which were strictly controlled because of the stock market crash, made the market liquidity tend to dry up, which brought the entire industry into the dormant period of the 2.0 era. Many quantitative private placements have collapsed in the sand during this period, and the quantitative teams that have gathered a certain strength have begun to transform and gradually become a climate. In this period, Jiukun relied on rich strategic factor reserves, cashed in reasonable neutral strategic returns through higher alpha, and initially formed its own brand.

In recent years, with the increasing number of market participants, the tendency of strategy homogenization is obvious, and the competition for investment opportunities to obtain alpha income has become increasingly fierce, and the related benefits have been diluted – which is also an important reason why the entire industry is accustomed to keeping its mouth shut. In order to avoid the failure of the strategy, quantitative private placement has also upgraded the hardware, improved the transaction speed and stability, and through the use of new methods such as machine learning, iterates on new factors, trying to build more diversified and advanced strategies, and grasping the initiative in the competition.

In this industry where every second counts, speed is king, and subversive, Wang Chen boils down Jiukun's investment philosophy to a "scientific, period" that cannot be more plain. He believes that one of the most basic investment logics of Jiukun is to rely on some statistical methods of history to find more investment opportunities in the market, and then gather their armpits to obtain investment returns. He trusts the discipline of trading and the determination to invest, and regards it as "basic".

The meaning of the name of Jiukun Investment is "the number of days of extreme competition is nine, and the load of virtue is Kun", which means that the company is as peaceful as possible for numbers and calculations, and for investment, it is peaceful. This is also a true portrayal of Jiukun's playing style: in terms of technology, Jiukun is always innovative, changeable, and the pursuit of the ultimate, while in terms of investment philosophy, it tends to be stable and tends to be diversified.

This year, the entire A-share market has seen a high and low market, especially after February, the A-share market fell by more than 20%, while the performance of Jiukun Investment was relatively stable. Wang Chen told 36Kr that this performance is the result of the diversification of quantitative investment itself and Jiukun's insistence on avoiding the style of timing strategy.

The reliance on decentralization is also reflected in its emphasis on diversity. Jiukun is the earliest quantitative private equity institution in the market to do three-year investment, not only has many product lines such as quantitative hedging, stock long and short, index enhancement, stock selection, commodity CTA, etc., but also belongs to one of the few institutions in the industry that has laid out all quantitative strategies.

Regarding discipline and adherence to investment determination, an example comes from Jiukun's active closure: at the end of May 2018, quantitative institutions continued to be blinded by funds, but Jiukun took the initiative to choose the fundraising of closed stock strategies. Because of the depth of the market at the time, they judged that the quantitative agency may face a capacity bottleneck. It was not until 2019, after vigorously investing and expanding in the direction of multi-cycle strategy and machine learning, and the strategy capacity was further released, Jiukun carried out a scale upgrade and ranked among the tens of billions of quantitative private placements in that year.

During this period, quantitative private equity has also begun to enter the 3.0 era of specialization, scale and team, and the competition has gradually become white-hot.

<h3>The Firth-like dilemma</h3>

Generally speaking, the investment research system of quantitative investment is roughly divided into four parts: data processing, α (excess return), risk (⻛risk), and execution (execution).

Data processing refers to the collection of various kinds of data, such as market research quotes, consumption records, weather, etc. α is to collect various kinds of data later and extract predictions for the future from it. Risk is the risk that you need to measure and control after all your investment decisions. Execution, when you decide to trade, buy or sell, you need to complete the transaction through a variety of modern ways, and then you need a strong IT platform to ensure that each part is executed in place.

As mentioned earlier, quantitative industries facing strategic homogeneous competition are always in the "Firth-like dilemma": in a gradually open market, with the rapid use of many strategies and data, the higher returns of the dividend period will quickly decline and return to the normal level of insurance returns, which requires the capture of more new data, the improvement and application of new technologies and even more refined transactions, to push the rock up the mountain again and again.

How did Jiukun cope?

From an organizational point of view, the core methodology is the division of labor + intelligence superposition of the team. Wang Chen emphasized that Jiukun is a quantitative fund built in a team model, which is different from those funds that rely on the research ability of the group, and realizes a quantitative analysis system similar to the Internet industrialization assembly line. It is relatively decentralized, which means that in each direction there are more experienced people who can independently do a good job of direction and can continue to push forward.

This investment and research model does not particularly emphasize individuals, it emphasizes the scalability of models and talents and strategies. For example, all the factors of Jiukun will be evaluated by a set of methods, and then submitted to the system, after which all the factors and models are completely detached from people and run automatically in an assembly line.

This model will also require a constant organizational division of labor to try. Usually, each strategy line needs to be laid out 1-2 years in advance, and a single research direction also needs to be explored for several months to more than a year. Each strategy and each direction must have someone to continue to focus on deep ploughing, and on this basis, try to achieve superposition.

From the data and technology level, in addition to the traditional market conditions and fundamental data sources, quantitative investment has widely used new data, including Internet data, for strategic development, and some new technologies, including AI, are also having a disruptive impact on quantitative investment.

Jiukun spends more than 10 million dollars on data every year, and also continues to study the application of AI in quantification. Last year, Jiukun set up an AI laboratory and spent more than 100 million yuan to build an AI supercomputing cluster. 70% of the technicians hired in the past two years also have research or industry backgrounds in AI.

From the perspective of talent reserves, Jiukun is probably one of the most "hungry" and hardest working in the quantitative private equity industry. Their belief is: "Strategies will fail over time, and only talent and technology are truly competitive."

As early as 2014, Jiukun began to recruit fresh graduates from Tsinghua, Peking University, Harvard, Stanford, Columbia and other well-known universities in the United States. In 2020, for the global undergraduate and doctoral graduates, Jiukun offered a million annual salary, and for overseas top quantitative fund talents, it also offered a million US dollars in annual salary.

For recruitment, Jiukun has its own training model. The internal association will organize experienced people to write teaching materials and conduct induction training for new people. At present, the backbone of Jiukun Investment Research Institute and a large number of intermediate forces are basically fresh graduates who enter the company as soon as they graduate.

In a speech to the outside world, Yao Cong, a partner at Jiukun Investment, compared the quantitative industry six or seven years ago to the Spring and Autumn Period: As long as there is a skill, many teams and individuals can make money. Now it has entered the Warring States era, there are mergers, there are eliminations, the head institutions have begun to gather more talents, began to control more resources, and a new era with new rules has begun.

On whether China's quantitative investment has the opportunity to overtake in curves, Wang Chen's answer is: In China, there is still a lot of room for quantitative investment, with the increase in the size of China's quantitative market and the entire securities market, China is likely to appear like Renaissance Technology, Two Sigma, Deshao, etc., or even higher level, larger quantitative funds.

For him, no matter what, it is wonderful to live a life of leniency. When he graduated with his Ph.D., he stumbled upon a book called "My Life as a Quant," a book about the high-energy particle physicist Deman's career on Wall Street. It gives Wang Chen's original life to an IT company an extra possibility. In his view, "quantitative investment, first, attaches great importance to technology, and second, it is very fair." It measures the ability to process data information and transactions in a market environment."

For a science and engineering student who does not want to be too disturbed by human factors, such a life is simply ideal.