How does stronger securities enforcement affect stock prices? The conventional wisdom is that it deters fraud, improves disclosure, and lowers the cost of raising outside capital, so shareholders should welcome it. The most relevant evidence supports that view. When the United Kingdom created a new private right of action for misstatements to secondary-market investors, the stock prices of affected firms rose.

In a new paper, I study a Chinese reform that made private securities enforcement much more credible, and I find that stock prices fell. Firms that came under the jurisdiction of China’s new specialized financial courts lost about 3 percent of their market value in the 10 trading days after the announcement. Measured against comparable firms that the reform left untouched, the loss was about 2.3 percent. The decline was largest for firms with a high risk of fraud and for firms facing claims that investors could still bring for past misconduct. I argue that the market was pricing liability for fraud that had already happened, and that much of this liability fell on shareholders who never benefited from the fraud.

The Reform

Chinese investors have long had a legal right to sue listed companies for false disclosure. In practice the right had little force. Each investor had to sue separately, individual losses were small, and many ordinary courts were reluctant to accept securities cases, which are complex, slow, and poorly suited to the performance metrics by which Chinese judges are evaluated.

Two developments changed this. The 2020 Securities Law introduced a special representative action, in which a state-backed investor-protection institution can sue on behalf of all affected investors unless they opt out. Over the same period, China established specialized financial courts with judges experienced in securities matters, and online platforms that can process thousands of claimants. The Shanghai Financial Court was established in 2018, the Beijing Financial Court in 2021, and the Chengyu Financial Court, serving Chengdu and Chongqing, in 2022. In 2021, jurisdiction over companies listed on Shanghai’s STAR Market was also transferred to the Shanghai Financial Court.

The courts changed how often investors sue. Of the 1,138 first-instance representative securities lawsuits in the PKULAW database of Chinese court decisions as of August 2026, 82 percent were decided by financial courts. Coming under a financial court’s jurisdiction raises the annual probability that a firm is sued by investors by about 0.3 percentage points compared with a base rate of about 0.1 percent. Administrative sanctions by the China Securities Regulatory Commission (CSRC) show no lasting change, which indicates that the courts affected private litigation, while the detection of fraud stayed about the same.

Because jurisdiction depends on where a firm is located or on which board it is listed, firms came under the new courts’ jurisdiction on different dates, and most never did. This allows affected firms to be compared with unaffected firms at the same moment.

What Happened to Stock Prices

I study the three events that followed the 2020 Securities Law: the creation of the Beijing court, the expansion of the jurisdiction of the Shanghai court to firms listed on the STAR Market, and the creation of the Chengyu court. Across 698 affected firms, stock returns over the 10 trading days starting on the announcement date were about 3.25 percent below what a standard asset-pricing model predicts. The decline was just as large among the 644 firms that had never been found to have violated disclosure rules, so it reflects a broad reassessment of enforcement exposure across the affected firms.

Because all the news arrives on three dates, a market-wide shock on one of those dates could be mistaken for an effect of the courts’ existence. I address this concern in two ways. First, I compare the affected firms with roughly 2,900 unaffected firms on the same dates. The gap between their abnormal returns is 2.3 percent over 10 days, and it is negative for each of the three events. Second, I repeat the entire analysis on 2,000 randomly chosen dates while keeping the same firms. Only five of the 2,000 random dates produce a gap as negative as the one observed on the actual announcement dates.

The STAR Market transfer coincided with a sharp fall in growth stocks in March 2021. The stock of similar growth firms listed on the ChiNext board, which the reform did not affect, fell by an average of 3.2 percent in that window. The stock of STAR Market firms fell by about four percentage points more.

Why Would Better Enforcement Lower Prices?

It helps to think of the price change on the announcement date as the sum of four parts. The first is the gain from better deterrence and disclosure, which raises firm value. The second is the cost of enforcement, such as litigation expenses, compliance, and management distraction, which reduces it. The third is expected liability for future fraud, which also reduces firm value, but only in part. As University of Chicago law professors Vincent Buccola and Adriana Robertson have pointed out, an investor who buys shares after the reform knows that, if a fraud is under way when the investor buys, the investor will have a claim for damages once the fraud comes to light. The price investors are willing to pay reflects the expected value of that claim, and this offsets part of the firm’s expected liability.

The fourth part is liability for fraud that has already occurred. Before the reform, claims for past misconduct were expected to go nowhere, so the prices at which those shares traded included no allowance for compensation. When the courts made such claims credible, the expected payout became a new liability of the firm, with nothing to offset it. I call this latent liability. Its size depends on how much past misconduct a firm carries that can still be the subject of a lawsuit.

Two tests support this interpretation. The first uses a statistical model that predicts, from information available at the time, which firms will be penalized in the coming year. The stock of firms with a higher predicted risk of fraud fell more when the courts were announced: a one-standard-deviation increase in fraud risk is associated with a 0.8 percentage point larger decline over the first four trading days. The same relationship is absent in a window of a month before the announcement, so the pattern appears only when the news arrives.

The second test uses the statute of limitations. I take firms that had already received a CSRC finding of misconduct but had not yet been sued and divide them according to how old the finding was when the court acquired jurisdiction. Where the finding was less than three years old, investors could still bring claims, and the stock of these firms fell by 2.3 to 3.5 percent. Where the finding was three to five years old, the limitations window had largely closed, and the stock of these firms rose. The groups are small, but the gap between them is large and statistically significant. The rise for time-barred firms fits the argument: These firms had passed the full limitations period without being sued, which left them better placed than firms still exposed to claims.

Litigation fees are too small to explain the result. Court and lawyers’ fees together come to roughly 2 to 3 percent of the damages claimed. The price effect was also a one-time adjustment. Over the following six to 24 months, affected firms performed about as well as unaffected firms.

Stranded Shareholders

Who bears latent liability? A securities-fraud claim belongs to the investor who bought during the fraud, and it stays with that investor after the shares are sold. When the courts make such claims credible, the stock price falls for everyone who holds the shares at that moment. Investors who bought during the fraud and have since sold will be compensated and bear none of the price drop. Investors who bought during the fraud and still hold bear the drop, but they hold a claim that offsets part of it. Investors who bought before the fraud began or after it ended bear the full drop and hold no claim at all.

I call this last group stranded shareholders. Their position is a transitional form of the circularity problem that professors John Coffee, Merritt Fox, and Jill Fisch have described, in which compensation paid to defrauded investors comes partly from other investors in the same company.

Implications

The comparison with the United Kingdom is instructive. The UK created a new right that applied only to future conduct, so there was no latent liability, and stock prices rose. China made an existing right credible, which revived claims for past fraud, and prices fell. Other differences between the two settings may contribute to the impact, but the contrast suggests that markets respond differently to creating a right and to making an existing right enforceable.

The purposes of enforcement point in the same direction. Deterrence requires only prospective reach, because fraud that has already happened cannot be deterred. The reason to reach back to punish past misconduct is compensation, and reaching back creates the transition cost. China had good reasons to accept that cost. Retail investors had suffered from a series of widely publicized frauds, and turning them away from the new courts would have been politically difficult. The right to sue also already covered past fraud, so a prospective-only design would have reduced claims that the legal system recognized.

For other jurisdictions where the right to sue for securities fraud exists mainly on paper, the lesson is that making it effective produces a one-time loss for current shareholders, and that the timing and design of the reform determine which shareholders bear the loss.

James Si Zeng is an associate professor at the Faculty of Law, The University of Hong Kong. This post is based on his paper,“When Stronger Private Enforcement of Securities Law Lowers Stock Prices: Evidence from China’s Financial Courts,” available here.