"Post a show-off post that I dare not post in the circle of friends, and my income has gone up every two years since I started working. After five years of work, I have earned figures that I couldn't think of when I just graduated (thanks to the company's platform)."

Qingdao University chief Wu may not have imagined that two years ago, his own "low-fashioned rich tasting" could bring himself a prison disaster.

Image for Wu Ships photo / Wu Ships LinkedIn account

In 2023, a netizen located in the United States posted anonymously on Xiaohongshu, posting photos of his income: in 2022, his income will be as high as 23.05 million US dollars, or about 167 million yuan. He admits that people in the same company are likely to recognize it, but "don't need to care so much". But he didn't expect that this post quickly detonated the network and attracted the attention of local media.

Anonymous Anonymous Anonymous Anonymous.

The outside world quickly speculated that his company was Two Sigma, the top quantitative hedge fund in the United States, and the company launched an internal investigation. In October 2023, according to US media reports, the person showing off his wealth was Wu Jian, vice president of quantitative research at Two Sigma and a Chinese citizen. He had been suspended by the company by this time and was fired by the company in 2024.

On September 11, 2025, local time, the U.S. District Court of the South District of New York and the U.S. Securities and Exchange Commission (SEC) announced civil and criminal lawsuits against Wu Shi, accusing him of telephone fraud, securities fraud and money laundering, violating the securities law, etc. Once the crime is established, Wu will face up to 60 years in prison.

Who is WOW?

At the end of 2023, Wu submitted a lawsuit to the New York State Court, dismissing the company's allegations, and detailed his professional background.

According to comprehensive information, Wu Jian was born in 1991 and grew up in Hefei, Anhui Province. Because of his outstanding performance in the national physics and mathematics competition, Wu Jian entered Tsinghua University in 2007 and received a bachelor's degree in automation engineering. In 2011, he entered the University of Southern California to study for a doctorate, transferred to Cornell University in 2013, and received a doctorate in operations research and information engineering in 2018.

Wu Jian joined Two Sigma in 2018. In just three years, he was promoted from quantitative researcher to vice president, and in early 2022 to senior vice president.

Two Sigma, a headquartered quantified hedge fund in Wall Street, was founded in 2001 and manages assets of more than $110 billion, with approximately 1,500 employees worldwide, most of whom come from well-known academies in mathematics, computing and engineering.

Unlike traditional fund managers who rely on experience and research, quantitative funds do not believe in "subjective feelings", but firmly believe in the power of data and algorithms. They use mathematical models and computer programs to find patterns from massive data, predict securities prices and automatically place orders. Help customers achieve their investment goals. For a long time, Two Sigma has emphasized that it is different from traditional investment companies and claims to be a "technology-driven hedge fund company."

This is why the accuracy and originality of the model are crucial. Wu Shift¡¯s responsibility is to study the development and monitoring of such ¡°money-making models¡± that can generate trading signals, mainly involving U.S. stocks listed on exchanges such as the New York Stock Exchange.

Nowadays, he is accused of email fraud, securities fraud and money laundering, precisely because he moved his hands on these models.

More than 14 models.

The SEC indictment disclosed that from November 2021 to August 2023, Wu Jian privately tampered with the key parameters of at least 14 models without authorization.

According to the regulations of Two Sigma, the newly developed quantitative model must maintain a low correlation with the existing model, ensuring that it can provide unique forecasts rather than duplicate existing forecasts, and at the same time measure the benefits it brings. The "decorrelation parameter" is a key indicator, and the lower the parameter value, the higher the expected correlation of the model.

The investigation found that without the company's knowledge, Wu Jian filled in the parameter values that could pass the review when submitting them for approval, and secretly reduced the values or approached zero in the system after the model was approved, which made the model generate new benefits on the surface, but actually only copied the existing model results.

The SEC¡¯s lawsuit, for instance, cites the addition of two zeroes to two minus points in three models, reducing the parameter to 1% of the original value.

Picture of prosecution.

On September 11, local time, the SEC released an indictment against Wu Ships, showing Wu Ships manipulation of the model to the relevant parameters: after the approval of the model, he secretly added two zeros after the original parameter value minus points in the newly built system.

This practice directly affects the model performance. In October 2023, Two Sigma admitted to clients that some funds made a profit of $450 million and others lost $170 million as a result of these "unauthorized modifications." The company will compensate the relevant customers.

The complaint states that Wu¡¯s motivation is directly linked to the remuneration mechanisms. Within Two Sigma, annual incentive remuneration, including cash bonuses and performance rewards, is linked to the performance model developed by employees.

At the end of 2022, Wu Jian negotiated with his supervisor to request a substantial increase in salary based on the performance of his model. The company ultimately paid him more than $23 million in salary, part of which was based on the performance of these "new models". According to reports, Wu Jian then used the proceeds to buy a multi-million-dollar apartment in Manhattan, new york.

The fuse of this incident being exposed is the above-mentioned "showing off wealth post". Wu Jian's high-profile speech on social media attracted the attention of Two Sigma, and the company immediately launched a review of Wu Jian's model. In order to cover up his previous behavior, he not only directed the survey to other employees, but also revised the decorrelation parameters of the model again. In August 2023, Wu Jian was fired from the company after admitting to changing parameters several times without approval. According to the U.S. Attorney's Office for the Southern District of new york, Wu Jian is currently absconding.

Rules can't keep up with algorithms

In October 2023, Wu¡¯s identity was exposed by U.S. media. At the end of the same year, Wu filed a lawsuit with the New York State Court, accusing Two Sigma of believing the company¡¯s claims failed. He claimed that although he was involved in model development, he did not decide on the actual deployment of the model, the losses should be blamed on the investment portfolio management and the trading team¡¯s ¡°irresponsible decision.¡±

Wu Jian also argued that researchers need to frequently modify parameters or update data after the model goes online. "Due to the lack of effective control, Two Sigma researchers and engineers often modify the model without formal approval, which is a common practice."

The newly disclosed criminal indictment in the Federal District Court for the Southern District of New York shows that Wu Jian was charged with wire fraud, securities fraud and money laundering, and the SEC filed a civil lawsuit against him according to the anti-fraud provisions of the securities law. The new apartment and other property he purchased using the proceeds of alleged fraud could face forfeiture.

Tian Lihui, dean of the Institute of Financial Development of Nankai University, pointed out to China News Weekly that the Wu Jian case is not an isolated case, but a systemic risk of the failure of algorithm ethics. At the technical level, the logic of the quantitative model lacks transparency, supervision is difficult to penetrate, and the "black box" feature makes it easy for parameter tampering to avoid auditing.

In fact, although Two Sigma was once known for its reliable technology, it has long had internal risk warnings. According to SEC disclosures, in March 2019, an employee reminded the company that there were loopholes in internal control, and researchers could change model parameters without review, but the management did not rectify them for a long time. In January of this year, Sanjay Wadhwa, acting director of the SEC's law enforcement department, bluntly stated in a statement: As investment advisors become increasingly dependent on models and advanced technologies, a sound compliance system is crucial, and Two Sigma has long-term inaction on known issues.

R. Tamara de Silva, a senior financial attorney in Chicago, also recently wrote that while Two Sigma has a model approval process and a security code base, it is far from sufficient for funds that rely heavily on the model, continuous monitoring, independent verification and rigorous management of parameter changes are key.

Tadeusz suggests that the case is revealed by the fact that in quantitative transactions, algorithm transparency is required, such as using blockchain technology to store evidence of model code and parameter changes, so that each line of code can be traced back.

Two Sigma links performance bonuses directly to model earnings, to a certain extent prompting Wu to take risks more easily under the temptation of high bonuses. Thalife believes that the incentive mechanism to say goodbye to "only the theory of earnings" also needs to include risk adjustments such as Sharp ratio, maximum withdrawal in the core assessment. In addition, it should also promote the establishment of a global regulatory alliance, unifying algorithmic trading standards and preventing cross-border regulatory suits.

The journalist: Yang Chige

Edited by: MJ