Print
E-MailZhiguo He is the James Irvin Miller Professor of Finance at Stanford University; Wenxi Jiang is a Professor of Finance at the Chinese University of Hong Kong Business School; and Wei Xiong is the John H. Scully ’66 Professor in Finance and Professor of Economics at Princeton University. This post is based on their recent paper.
China’s stock market appears to predict future corporate earnings, but this predictability does not necessarily mean prices are a clean “crystal ball” for fundamentals. High valuations also appear to pressure firms to cater to investor expectations by inflating reported earnings, especially through non-recurring gains and losses. The result is a distinctive pattern: short-run earnings predictability followed by longer-run reversal, suggesting that price informativeness and earnings management can reinforce each other.
Can stock markets discipline firms and allocate capital efficiently when the accounting system is still developing? This question is central to current debates about China’s capital markets. Policymakers have sought to make the A-share market a more important venue for financing innovation, disciplining listed firms, and giving households a larger stake in corporate growth. At the same time, China’s equity market remains known for speculative trading, retail investor dominance, weak delisting discipline in earlier years, and recurring concerns about the quality of listed firms’ financial reports.
A growing literature has offered a more optimistic interpretation of China’s stock market. Bai, Philippon, and Savov (2016) propose measuring price informativeness by asking whether today’s market valuations predict firms’ future earnings. Using this approach, Carpenter, Lu, and Whitelaw (2021) find that Chinese stock prices appear surprisingly informative: despite the market’s volatility and speculative reputation, high-valued A-share firms subsequently report higher earnings. This finding has important implications. If prices in China forecast fundamentals almost as well as prices in mature markets, then the A-share market may already be performing a core economic function.
Our recent working paper, He, Jiang and Xiong (2026), revisits this interpretation. We do not deny that stock prices may contain useful information about future firm performance. Instead, we ask whether part of the documented predictability reflects a different mechanism. In markets with weaker governance and more flexible accounting practices, stock prices may not only forecast future earnings; they may also pressure managers to produce those earnings. When a firm enjoys a high valuation, managers face stronger incentives to justify that valuation. The result may be a form of feedback from prices to accounting reports, which we call “manipulate to cater.” This concern is especially relevant in China, where Piotroski and Wong (2012) emphasize the still-developing information environment of listed firms, and Allen, Qian, Shan, and Zhu (2024) document broader governance challenges in the Chinese stock market.
The basic idea is simple. Under the standard “crystal ball” view, a high stock price reflects investors’ information about strong future fundamentals. Under the manipulate-to-cater view, causality can also run in the opposite direction. A high stock price raises expectations, and managers respond by producing the reported earnings that investors appear to expect. This mechanism builds on two familiar ideas: Stein’s (1989) insight that stock-market pressure can induce managerial myopia, and Hirshleifer and Teoh’s (2003) argument that investor inattention can make financial reporting distortions matter for prices.
This distinction matters because reported earnings are not the same as cash flows or payouts. If prices truly reveal stronger fundamentals, high-valued firms should eventually generate more cash and distribute more value to shareholders. If instead managers inflate reported earnings, the effect should be strongest in accounting components that are easier to manage, should weaken or reverse over time, and should not translate into comparable increases in dividends or repurchases.
We examine all non-financial Chinese A-share firms from 1995 to 2024 and compare the results with U.S. S&P 500 firms. Consistent with earlier work, firms with higher market-to-asset ratios in China subsequently report higher earnings. This cross-sectional fact survives in the extended sample, though the magnitude is smaller in recent years than in the earlier sample ending in 2016. On the surface, this finding again suggests price informativeness.
But several additional patterns call for a more cautious interpretation. Figure 1 highlights the first pattern: in the United States, the predictive power of valuation for future earnings rises steadily with the forecast horizon. In China, by contrast, it rises only in the short run and then flattens or reverses at longer horizons. This divergence is difficult to reconcile with a pure crystal-ball interpretation. A second pattern comes from within-firm earnings dynamics. When a Chinese firm’s valuation is unusually high relative to its own history, its reported earnings tend to rise in the short run but weaken over three- to five-year horizons. This reversal is absent in the U.S. S&P 500 sample. A third pattern reinforces the first two: unlike in the United States, where valuation predicts both future earnings and future payouts, high valuations in China have much weaker predictive power for cash distributions to shareholders.
Figure 1. Stock price informativeness about future earnings. The figure plots the predicted variation in future earnings associated with current market valuation in Chinese A-share firms and U.S. S&P 500 firms. Error bars show 95% confidence intervals. Source: He, Jiang, and Xiong (2026).
The source of the difference becomes clearer when earnings are decomposed into operating cash flows, accruals, and non-recurring gains and losses. Operating cash flows are harder to manipulate and better reflect the firm’s ongoing business. Accruals involve more managerial discretion. Non-recurring gains and losses, or NRGLs, are one-off items such as asset sales, subsidies, restructurings, and other non-operating gains or losses. In China, these items have been especially important because, before recent reforms, they could help firms meet earnings thresholds and avoid delisting.
In the U.S. sample, market valuation primarily predicts future operating cash flows. This is what one would expect if prices mainly aggregate information about fundamentals. In China, however, the predictive power of valuation is much weaker for operating cash flows and much stronger for the more manipulable components of earnings. Accruals account for some short-run predictability, but NRGLs are the most persistent channel. High-valued firms are more likely to report future earnings boosted by non-recurring items.
Investors also do not appear to fully discount these managed earnings. If the market understood that NRGL-driven earnings are transitory, firms with high NRGLs should not earn predictably lower future returns. Yet they do. A two-standard-deviation increase in NRGLs is associated with about a 3 percent lower stock return over the subsequent year. This finding suggests that at least some investors take reported earnings too much at face value, allowing earnings management to support overvaluation temporarily before prices adjust.
The 2020 reform of China’s delisting rules provides further evidence. Historically, delisting from the A-share market was rare, in part because shell values were high and firms could use accounting tools to avoid regulatory thresholds. The 2020 reform excluded NRGLs from the earnings measure used for delisting purposes, effectively raising the cost of managing earnings through one-off items. After the reform, firms that had previously relied more heavily on NRGLs reduced their use. More importantly, the relation between market valuation and future reported earnings weakened. No similar structural break appears in the U.S. data. This policy change supports the interpretation that part of China’s earlier price-earnings predictability came from earnings management rather than pure information aggregation.
Evidence from A-H dual-listed firms points in the same direction. These firms are listed both in mainland China and Hong Kong, and thus face a different disclosure and investor environment. Their A-share valuations are less predictive of future earnings than valuations for firms listed only on the mainland. Moreover, for dual-listed firms, H-share valuations have stronger predictive power for future earnings than A-share valuations. This pattern is difficult to reconcile with the view that A-share prices are simply more informative. It is more consistent with the idea that institutional environment and investor sophistication affect whether prices reflect fundamentals or help sustain accounting-based expectations.
The policy lesson is not that China’s stock market is uninformative. Rather, it is that price informativeness cannot be assessed by looking only at whether prices predict reported earnings. In a market where accounting discretion is substantial and investors may be inattentive to the quality of earnings, reported earnings are themselves an outcome influenced by market pressure. Prices can be informative and distortive at the same time.
This perspective has broader implications for capital-market development in China. Strengthening price discovery requires more than expanding market size, increasing trading activity, or encouraging more firms to list. It also requires institutions that make reported earnings harder to manage and easier to interpret. Delisting reforms, better disclosure of non-recurring items, stronger enforcement, and investor education all matter because they reduce the scope for prices and earnings to reinforce each other in misleading ways.
A well-functioning stock market should not merely reward firms for meeting short-run earnings expectations. It should help allocate capital toward firms with durable cash-flow potential. China’s experience shows that the appearance of price informativeness can mask an important feedback loop: high prices create pressure for high earnings, managed earnings validate high prices, and only later do fundamentals reassert themselves. Breaking this loop is essential if the A-share market is to become not only large and liquid, but also a reliable guide to corporate value.
Link to the full paper: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5327469
