The Real Value of China s Stock Market

Transcription

1 The Real Value of China s Stock Market 中国股票市场的实体价值 Jennifer N. Carpenter New York University Fangzhou Lu MIT Robert F. Whitelaw New York University JOIM Conference Series Legacy of Jack Treynor, Future of Active Management, China Investing March 14, 2017

2 China s Stock Market Casino? Or Crystal Ball? Opened in 1991 as Deng Xiaoping s privatization experiment. Now the world s 2 nd largest stock market. First labeled a casino in 2001 after period of speculation and fraud. High volatility erodes B&H returns, fuels perception of poor performance. Rollercoaster ride of 2015 cements market s casino reputation (China's stock market A crazy casino The Economist 5/26/2015). Our research suggests that these perceptions are wrong. Stock prices in China are strongly linked to firm fundamentals.

3 Number of Listed Firms Chinext SME Shenzhen Main Board Shanghai Main Board Shanghai and Shenzhen main boards list larger, mature firms. SME and ChiNext boards on Shenzhen Stock Exchange list smaller, more entrepreneurial firms. Many distinctive features 80% of volume is retail, market segmented by capital controls, large non-tradable component. Now has over 3100 stocks listed, plus 600 firms seeking IPO approval.

4 Market Capitalization in Trillion RMB Other non-tradable Non-tradable executive or employee Non-tradable legal person Non-tradable state owned Tradable mutual fund Other tradable China s stock market now has $8T capitalization. Split-Share Reform of 2005 enabled the unlock of non-tradable shares. Now over 75% of stocks are tradable.

5 Stock Price Informativeness Good legal and market institutions Stock price informativeness about profits Efficient capital allocation and investment Long literature in economics, finance and accounting going back to Hayek (1945) & Fama (1970) links good institutions to stock price informativeness, and further to investment efficiency: Benefits of good listing, disclosure, and audit policy (Amihud & Mendelson (1988), Diamond & Verrechia (1991), Hail & Leuz (2009). Aggregation of diffuse information across individuals, incentives to generate information, and its inference from prices (Grossman & Stiglitz (1980), Kyle (1985), Glosten & Milgrom (1985). Managerial use of price signals in investment decisions (Wurgler (2000), Durnev, Morck, and Yeung (2004), Chari & Henry (2004)).

6 Measuring Stock Price Informativeness Bai, Philippon, Savov (2016 JFE) develop a theoretical model in which stock price informativeness promotes efficient corporate investment and economic growth. They define price informativeness as the extent to which market valuations differentiate firms that will have high earnings from those that will not. They analyze the coefficient, predicted variation, and marginal R 2 of equity value in the following cross-sectional regression using US data: Net Profit Assets i, t i, t+ k = + Mkt Equity i, t i, t c j, t at log + bt + ei, t+ k Assetsi, t Assetsi, t Net Profit Industry fixed effects control for differences in discount rates.

7 Stock Price Informativeness in China Coefficient for k= Coefficient for k= Coefficient for k= Predicted variation for k= Predicted variation for k= Predicted variation for k= Marginal R 2 for k= Marginal R 2 for k= Marginal R 2 for k=

8 Price Informativeness by Forecasting Horizon k Times series average coefficient The price informativeness coefficient increases with the forecasting horizon (similar to US). Perhaps because more distant future earnings are better proxies for the complete earnings stream capitalized into the current price, especially for China with its high growth rate.

9 Predicted variation Stock Price Informativeness in China

10 Stock Price Informativeness: China vs. US Coefficients for k= Coefficients for k=5 China US Top of 10% Rejection Region US results from Alexi Savov Stock price informativeness in China is essentially indistinguishable from that in the US. In most years, we cannot reject that China stock price informativeness is the same as in the US at any conventional significance level.

11 Cross-Sectional Variation in Stock Price Informativeness Coefficient estimates and t-stats for increase in price informativeness with dual H-share listing, state ownership, or both. H Shares Dual-Listed State Ownership k=1 k=2 k=3 k=4 k=5 k=1 k=2 k=3 k=4 k= The presence of dual-listed H shares appears to reduce A-share price informativeness noise from discount rate shocks in Hong Kong leaking into A-share prices? Higher SO associates with lower PI political risk in state subsidies makes earnings harder to predict?

12 Corporate Investment Efficiency We define the efficiency of corporate investment as the unexpected change in existing equity value associated with a unit of unexpected investment. I.e., the coefficient β t in the following cross-sectional regression: with industry fixed effects to control for differences in expected growth and depreciation rates of capital stock. A modification of the definition from Durnev, Morck, and Yeung (2004) for China s setting. The coefficient β t is cross-sectional average NPV of a unit innovation in investment in year t.

13 Stock Price Informativeness and Corporate Investment Efficiency Predicted variation Coefficient k=3 k=4 k=5 Average Beta t-stat Beta t-stat Marginal R 2 Beta t-stat Results from regressions of corporate investment efficiency on 1-year lagged price informativeness (and a time trend). Stock price informativeness and corporate investment efficiency are significantly positively related. Suggests the stock market is generating useful signals for managers.

14 Equity Pricing in China We ve just established a strong link between stock prices and future profits. Further questions: o How do investors discount these profits? o What are the implications for firms cost of capital? o What are the investment opportunities for USD global investors? QFII/RQFII Shanghai-Hong Kong Stock Connect Shenzhen-Hong Kong Stock Connect

15 Chinese Investors Price Stocks Like Investors in Other Large Economies Monthly excess returns for stock characteristics in the US (from Bali, Cakici, Whitelaw, 2011). Beta Size BM Mom. Illiq. MAX Rev. Coeff t-stat Monthly excess returns for stock characteristics in China Beta Size BM Mom. Illiq. MAX Rev. NTS SOE Coeff t-stat Coeff t-stat Coeff t-stat

16 Opportunities in China for Global Equity Investors In contrast to perception, China s market has performed very well using traditional performance measures R M -R F China US Europe Japan Ann. mean (%) Ann. vol. (%) China has a low correlation with other large economies. China US Europe US 0.16 Europe Japan

17 High Alphas in China for Global Investors China US Global 1-factor 4-factor 1-factor 4-factor R M -R F Alpha t-stat SMB Alpha t-stat HML Alpha t-stat WML Alpha t-stat High potential alphas available to global investors mean high cost of capital for China s firms. Research by Henry suggests stock market liberalization and integration would reduce the cost of capital for China s firms and boost investment, wages, and economic growth.

18 Are There Alternative Ways to Access These Investment Opportunities? Both theory and our empirical evidence suggest A shares offer alpha potential in a market where cross-sectional pricing is broadly consistent with that in developed markets. However, there are risks (e.g., repatriation) and constraints (e.g., QFII licensing) that create barriers to direct investment in China s stock market. Can these potentially attractive returns be accessed by investing in Chinese firms that list in other markets? For example, o N shares listed in the U.S. (e.g., Alibaba). o Red chips and P chips listed in Hong Kong (e.g., China Mobile). Problem: Are Chinese firms incorporated outside of China that choose to list internationally fundamentally different from China-listed firms? Solution: A significant sample of firms are dual-listed in China (A shares) and Hong Kong (H shares).

19 The Dual-Listed Universe The first dual-listed company was Tsingtao Brewery (1993). By the end of 2016 there were 92 dual listings, including Air China, PetroChina, and Bank of China. 100 Company Count There is significant time-series and cross-sectional variation in the price premiums between A and H shares.

20 H Shares Are Far From a Perfect Substitute for A Shares Correlations between equal-weighted A and H share returns for duallisted firms are relatively low but increasing The discount rate effect from the market in which the stock trades dominates the cash flow effect from the dual-listed counterpart. EW A EW H Global Mkt Size Value Mom. EW H China EW A HK Coeff t-stat Coeff t-stat

21 Conclusions 结论 1. Stock price informativeness Contrary to the Casino Theory, stock prices in China are as informative about future profits as they are in the US. Stock prices appear to be generating useful signals for managers. 2. Pricing Though it is a segmented market, China s stocks exhibit pricing patterns similar to those in developed markets investors pay up for size, growth, and long shots, and discount for systematic risk. The market has performed well and its low correlation creates opportunities (high alpha) for global investors. This high alpha is a high cost of capital for China s firms and a drag on economic growth. Foreign-listed Chinese companies are not perfect substitutes for A shares, therefore to access these returns investors must take on risks repatriation risk, exchange rate risk, legal, political, and taxation risk. Our research suggests that both China and global investors stand to gain from China s opening up its capital markets.