Frontiers in Business, Economics and Management ISSN: 2766-824X | Vol. 18, No. 2, 2025 208 Analyst coverage and Stock Price Crash Risk: Based on the Perspective of Social Trust Dexing Deng1, *, Xingyu Deng2, a 1School of Economics, Guangxi University, Nanning 530004, China 2School of Accounting, Wuchang Institute of Technology, Wuhan 430065, China *Corresponding author email: 1756753015@qq.com, a3268466143@qq.com Abstract: The "14th Five-Year Plan" points out the need to establish and improve a comprehensive system for risk prevention, early warning, response, and accountability, while also strengthening integrity construction and building an efficient, standardized, and fair competitive market. This article is based on data from Chinese A-share listed companies from 2003 to 2018 and employs a multiple regression model to explore the relationship between analyst attention and stock price crash risk from the perspective of social trust. Research findings indicate that analyst attention increases the risk of stock price crashes for companies. Further research results suggest that in regions with higher social trust, the positive impact of media attention on the risk of stock price crashes is significantly reduced. The conclusion of this study remains valid after a series of robustness tests. Keywords: Analyst coverage; Crash Risk; Social Trust. 1. Introduction In 2008, the Chinese A-share market experienced significant fluctuations, with the Shanghai Composite Index falling by a total of 3,445 points throughout the year, representing a decline of 65.39%, marking the largest drop in the 18-year history of A-shares. The Shanghai Composite Index fell from its historical high of 6124 points in October 2007 to a low of 1664.93 points in 2008, resulting in a cumulative decline of 72% over the course of one year, with both the market capitalization and circulating market value of the Shanghai and Shenzhen stock markets shrinking by more than 60%.In 2015, the Chinese A-share market experienced a severe stock price crash, reaching a peak of 5178 points in June 2015, after which the market began to plummet dramatically. Within just 17 trading days, the Shanghai Composite Index fell by more than 30%, dropping to 2850 points (Tian, 2020). The two "plummeting" stock prices in China's A-share market have seriously impacted the healthy development of China's capital market and investors' investment confidence and wealth (Meng et al., 2017). Therefore, it is of great theoretical and practical significance to explore the formation factors and governance methods of stock price crash risk to reduce the financial risk of China's capital market and promote the healthy and stable development of the market (Yang et al., 2018). As the "information bridge" between the capital market and investors (Tian, 2020), analysts can interpret the information of the financial market and then transmit it to investors. Analysts' investment income forecasts for listed companies and corporate investment income ratings provide an important reference for investors' investment decisions, which can effectively promote the effective allocation of market resources for the entire financial market. In addition, analysts can also be used as a substitute for corporate governance, which can effectively improve the authenticity of corporate information disclosure, thereby reducing the risk of corporate stock price crash (Pan, 2011). On the contrary, analysts' optimism bias may increase the risk of stock price crash (Xu, 2012), and the increase in analyst attention will increase market sentiment, which will lead to irrational factors in stock prices and increase the risk of stock price crash; In addition, analysts' concerns increase the pressure on managers to achieve short-term earnings targets, prompting them to engage in upward earnings management information manipulation, thereby increasing the risk of future stock price crashes (Han et al., 2021). As an informal system, regional social trust can inhibit the incentive of firms to conceal bad news, thereby reducing the risk of stock price crash (Liu et al., 2016). At the same time, firms in areas with high social trust tend to be associated with higher accounting soundness and fewer financial restatements, thereby reducing the risk of a firm's stock price crash (Li et al. 2017). Social trust will also have a positive impact on corporate information disclosure and investment and financing behavior, thereby reducing the risk of stock price crash (Nie & Ran, 2020). In summary, it is of great significance to explore the impact of analysts' attention on stock price crash risk from the perspective of social trust for the stable development of China's capital market and society. Compared with the previous literature, this paper introduces the moderating effect of social trust into the impact of analysts' attention on stock price crash risk, and enriches the articles on stock price crash risk; it also provides a reference scheme of practical significance for the formulation of national policies and the maintenance of the long-term and stable development of China's financial market. 2. Theoretical Analysis and Research Hypotheses Stock price crash risk refers to a situation in which a company's stock falls significantly in a short period of time (Jin & Myers, 2006). Previous cases have shown that there are two triggers for stock price crash risk, one is agency problems, which can lead to opportunistic behavior in corporate management, which can lead to the risk of stock price crash (Liang & Zeng, 2016). The "hollowing out" of the management and the opportunistic behavior of obtaining more self-interest will cause the company to hoard more 209 "negative news", and when the "negative news" hoards to a certain extent, it will be released in a large amount in a short period of time, thus causing the company's stock price to crash. The second is the opacity of information disclosure of listed companies, and companies will selectively disclose some good information and hoard "bad news", This information asymmetry prevents external investors from making the right investment behavior, and when the "bad news" can no longer be hoarded, it is released externally, at which point the market reacts, causing the company's stock price to plummet (Hutton et al. 2009). 2.1. Analyst focus and stock price crash risk As an information intermediary between investors and listed companies, analysts screen and transmit investment signals to investors in the financial market, which will increase the value of the company's stock price information and reduce the synchronization of the company's stock price to a certain extent (Li et al., 2016). However, the role of analysts as information intermediaries is not always positive, and it can be seen from the relationship between media attention and analyst attention that analysts will enhance the media attention of enterprises, thereby increasing the risk of stock price crash of enterprises (Cu&San, 2022); Moreover, the impact of analysts' concerns on corporate earnings management varies depending on the development of capital markets (Degeorge et al. 2013); In addition, the positive effect of cash flow risk on stock price crash risk will be more significant in places where analysts pay more attention (Pei , 2021). Therefore, the increase in analyst attention will not only not improve the information transparency of enterprises, but will also increase information asymmetry, thereby increasing the risk of stock price crash. Therefore, this article proposes Hypothesis 1: H1: The higher the analyst focus, the higher the risk of a company's stock price crash. 2.2. Social trust, Analyst focus and stock price crash risk Regional social trust is an informal system that inhibits management's opportunistic behavior of concealing "bad news" (Liu et al., 2016), while firms in regions with higher levels of social trust tend to be associated with higher accounting robustness and less financial restatement (Li et al. 2017), Social trust also reduces the agency costs and tax evasion levels of enterprises (Shen , 2019), indicating that in regions with higher levels of social trust, corporate decision- makers are more likely to internalize the values of honesty and integrity as personal qualities, thereby reducing agency problems and the motivation to conceal negative information. Therefore, this article proposes Hypothesis 2: H2: Regions with higher social trust mitigate the positive relationship between analyst attention and stock price crash risk. 3. Research Design 3.1. Sample selection and processing The research sample selected in this article consists of all A-share listed companies from 2004 to 2019.In this paper, the number of analyst trackers of each company is selected as the indicator of analyst attention, and the stock price crash risk data comes from the CSMAR database, and the rest of the data comes from the WIND database.The text filters the data as follows: (1) samples with annual transaction data of less than 26 weeks are excluded, (2) samples from PT&ST companies are excluded, (3) samples with missing data are excluded, (4) all continuous control variables undergo a winsorization process at the 1% upper and lower tails, resulting in a final total of 26,768 annual samples from companies. 3.2. Variable selection and measurement 3.2.1. Stock price crash risk This paper refers to the practice of existing literature (Liang and Zeng , 2016, Hutton et al., 2009) to exclude the impact of market systemic risk through the following model: RET, =𝛼 +𝛽 𝑀𝐴𝑅𝐾𝐸𝑇 +𝛽 𝑀𝐴𝑅𝐾𝐸𝑇 + 𝛽 𝑀𝐴𝑅𝐾𝐸𝑇 +𝛽 𝑀𝐴𝑅𝐾𝐸𝑇 +𝛽 𝑀𝐴𝑅𝐾𝐸𝑇 +πœ€ (1) In the formula, the subscript i and t represent the company and the week, respectively, and RET is the weekly return of individual stocks; MARKET is the adjusted market rate of return weighted by the circulating market value of all listed companies (excluding the ith company); πœ€ is the random error term. In this paper, we add data from the leading and lagging periods to model (1) to mitigate the bias caused by the asynchronous nature of transactions (French et al., 1987). The company's characteristic yield is: Ο‰ log（1 Ξ΅οΌ‰ (2) where πœ€ is the residual value in model (1), with reference to Hutton et al. (2009), Kim et al. (2011a,b) and Chen et al. (2001), this paper uses three indicators to measure the risk of stock price crash. Referring to the practice of Liang and Zeng (2016), the negative value of the skewness of the company's characteristic return (NSKEW) is used to measure the risk of stock price crash: the larger the NSKEW, the greater the likelihood of a stock price crash, The NSKEW of the company in the year i is: π‘π‘†πΎπΈπ‘Š = n n 1 / βˆ‘Ο‰ / n 1 n 2 βˆ‘Ο‰ / (3) Referring to the practice of Xu et al. (2012), the second indicator to measure the risk of stock price crash is the ratio of falling volatility to rising volatility, and the larger the DUVOL, the higher the risk of stock price crash: π·π‘ˆπ‘‰π‘‚πΏ =log n 1 βˆ‘ Ο‰ / n 1 βˆ‘ Ο‰ (4) Where nu and nd represent the number of weeks of decline and the number of weeks of growth, respectively. Referring to Wu et al. (2019), the third indicator in this paper is COUNT, which indicates the difference between the frequency of downward and upward stock returns in a year, and the larger the difference, the higher the risk of stock price crash: COUNT Crash Crash (5) Which Crashdown indicates the frequency of upward movement of stock returns, and Crashup indicates the 210 frequency of downward movement of stock returns. 3.2.2. Analyst coverage Referring to the practices of Zhou et al. (2016) and Cui et al. (2022), this paper selects the sum of all analysts who have reported on the company in one year, and takes the logarithm of 1 + number of analysts to obtain the analyst attention index ANANUM. 3.2.3. Regional social trust In this paper, the indicators composed of the five options of CGSS and CGSS were selected, and the indicators were distinguished by the sample mean, if the sample mean was exceeded, 1 was taken, otherwise 0 was taken, so as to construct the TRUSTDUMM1 and TRUSTDUMM2 of the regional social trust dummy variables. 3.2.4. Control variables The article selects the following control variables based on the research results of previous scholars: leverage ratio (LEV), market-to-book ratio (MTOB), company size (SIZE), return on assets (ROA), stock turnover rate (DTURN), information transparency index (ABACC), earnings per share (EPS), degree of separation of rights (SEPRT), average weekly stock return (RET), standard deviation of average weekly stock return (SIGMA), and the current period's stock price crash risk indicator (NSKEW). To reduce variable bias, all control variables in this article are lagged by one period. 3.2.5. Model Design This paper explores the relationship between analyst attention and stock price crash using the following model: 𝐢𝑅𝐴𝑆𝐻 𝛼 𝛽 π΄π‘π΄π‘π‘ˆπ‘€ 𝛽 𝐢𝑂𝑁𝑇𝑅𝑂𝐿𝑉𝐴𝑅𝐼𝐴𝐡𝐿𝐸 πœƒ 𝛿 πœ€ (6) The term CRASH represents the risk of stock price collapse, ANANUM indicates analyst attention, and this article incorporates regional and year fixed effects in the regression, where πœ€ denotes the residuals, with the primary focus being on the Ξ²1 coefficient. Table 1 lists the definitions of each variable: Table 1. Variable definitions Variable name Variable symbol Variable Definition Stock price crash risk NSKEW Negative value of the skewness of the company's trait returns DUVOL The ratio of the falling volatility of a company's stock price to the rising volatility COUNT Formula: The difference between how often a stock price goes up and down Analyst attention ANANUM ln(1 + Analyst number) ANANUM2 The sum of the number of analysts Social trust TRUSTDUMM1 Variables consisting of the dumb CGSS variable, exceeding the sample mean by 1, otherwise 0. TRUSTDUMM2 The mean values of the samples calculated by the five indicators of CGSS are used as the boundary, and the value greater than the mean is 1, otherwise 0 is taken. Debt-to-asset ratio LEV Total Liabilities/Total Assets The size of the company SIZE ln（1+Total assetsοΌ‰ Return on assets ROA Net Profit/Total Assets Market capitalization-to-book ratio MTOB (Market value of outstanding shares + book value of non- tradable shares) / Book value of equity Stock liquidity DTURN Turnover rate in year t - turnover rate in year t-1 Information transparency ABACC Operational accrual earnings management calculated by the modified Jones model Earnings per share EPS Total Revenue/Total Number of Shares Separation of powers SEPRT The control of the actual controller - the right to cash flow Standard deviation of returns SIGMA The standard deviation of the company's earnings in year t. Average returns RET Average weekly holding yield 4. Empirical Analysis Results 4.1. Descriptive statistics The descriptive statistics of each variable are shown in Table 2, and the mean values of the stock price crash risk indicators (NSKEW, DUVOL, COUNT) are -0.485, 0.823 and -0.149, respectively, and the standard deviations are 0.753, 0.315 and 0.570, respectively. It shows that the risk of stock price crash varies greatly among different companies; The average analyst attention is 1.384, which indicates that the analyst coverage of listed companies is wide, and the variance is 1.150, which indicates that the analyst attention of different companies varies greatly. 4.2. Multiple Regression Analysis Firstly, the relationship between analyst attention and stock price crash risk is verified, as shown in Table 3. Regressions (1), (3) and (5) were the results without the addition of control variables, and the coefficients were 0.064, 0.018 and 0.033, respectively, which were significant at the level of 1%. After adding the control variables, the coefficients increased to 0.069, 0.021 and 0.040, respectively, and were significantly below the 1% level, indicating that the higher the analyst attention, the higher the risk of a company's stock price crash, validating hypothesis H1. 211 Table 2. Descriptive statistics OBS Mean SD Min p25 p50 p75 Max NSKEWt+1 26768 -0.485 0.753 -5.555 -0.850 -0.412 -0.034 4.181 DUVOLt+1 26768 0.823 0.315 0.130 0.610 0.775 0.974 4.640 COUNTt+1 26768 -0.149 0.570 -2.000 0.000 0.000 0.000 2.000 ANANUMt 26768 1.384 1.150 0.000 0.000 1.386 2.303 4.331 TRUSTDUMM1t 21844 0.607 0.081 0.317 0.538 0.628 0.647 0.865 TRUSTDUMM2t 21844 0.417 0.161 -0.270 0.323 0.440 0.490 1.040 LEVt 26768 0.451 0.203 0.051 0.295 0.456 0.608 0.886 SIZEt 26768 22.291 1.039 20.041 21.568 22.227 22.924 25.186 ROAt 26768 0.040 0.058 -0.197 0.013 0.036 0.067 0.213 MTOBt 26768 1.915 1.199 0.842 1.174 1.511 2.183 7.997 DTURNt 26768 -0.100 0.461 -1.879 -0.277 -0.038 0.136 0.953 ABACCt 26768 0.076 0.078 0.001 0.024 0.052 0.100 0.452 EPSt 26768 0.332 0.494 -1.182 0.081 0.249 0.508 2.409 SEPRTt 26768 5.071 7.925 -0.012 0.000 0.000 9.045 49.398 NSKEWt 26768 -0.447 0.703 -2.704 -0.814 -0.383 -0.008 1.283 SIGMAt 26768 0.048 0.019 0.017 0.035 0.045 0.058 0.113 RETt 26768 -0.134 0.114 -0.635 -0.168 -0.099 -0.059 -0.015 Table 3. Main regression results (1) (2) (3) (4) (5) (6) NSKEWt+1 NSKEWt+1 DUVOLt+1 DUVOLt+1 COUNTt+1 COUNTt+1 ANANUMt 0.064*** 0.069*** 0.018*** 0.021*** 0.033*** 0.040*** (14.55) (12.80) (9.58) (9.19) (10.30) (9.81) LEVt 0.014 -0.010 0.028 (0.49) (-0.82) (1.18) SIZEt -0.075*** -0.024*** -0.038*** (-12.55) (-9.30) (-8.38) ROAt -0.799*** -0.249*** -0.370*** (-5.66) (-4.08) (-3.46) MTOBt 0.031*** 0.012*** 0.018*** (5.99) (5.16) (4.54) DTURNt -0.011 -0.013*** -0.012 (-0.95) (-2.63) (-1.35) ABACCt 0.274*** 0.103*** 0.139*** (4.79) (4.17) (2.96) EPSt 0.079*** 0.022*** 0.046*** (5.01) (3.29) (3.82) SEPRTt -0.000 -0.000 -0.000 (-0.79) (-0.19) (-0.82) NSKEWt 0.034*** 0.015*** 0.018*** (5.09) (5.27) (3.41) SIGMAt 0.268 0.856* 3.787*** (0.23) (1.74) (4.53) RETt 2.231*** 0.769*** 1.303*** (11.07) (9.61) (9.40) Constant -0.573*** 1.288*** 0.798*** 1.360*** -0.195*** 0.565*** (-69.19) (9.23) (241.90) (22.58) (-32.79) (5.44) Observations 26,768 26,768 26,768 26,768 26,768 26,768 AdjR2 0.055 0.129 0.057 0.097 0.030 0.048 Industry FE CONTROL CONTROL CONTROL CONTROL CONTROL CONTROL Year FE CONTROL CONTROL CONTROL CONTROL CONTROL CONTROL *, **, and *** represent significant levels of 10%, 5%, and 1%, respectively, and the T value is in parentheses, the same below. 5. Empirical Analysis Results 5.1. Replace explanatory variables As shown in Table 4, ANANUM2t is the replacement indicator, and it can be seen from Table 4 that the stock price crash risk indicator is still significant at the level of 1% after the explanatory variable substitution. 212 Table 4. Robustness test1 (1) (2) (3) (4) (5) (6) NSKEWt+1 NSKEWt+1 DUVOLt+1 DUVOLt+1 COUNTt+1 COUNTt+1 ANANUM2t 0.009*** 0.009*** 0.003*** 0.003*** 0.005*** 0.005*** (16.33) (13.99) (11.13) (10.47) (11.84) (10.79) LEVt 0.011 -0.010 0.026 (0.39) (-0.81) (1.13) SIZEt -0.074*** -0.025*** -0.038*** (-12.57) (-9.64) (-8.41) ROAt -0.666*** -0.212*** -0.294*** (-4.79) (-3.52) (-2.78) MTOBt 0.026*** 0.010*** 0.015*** (4.99) (4.29) (3.71) DTURNt -0.016 -0.015*** -0.015* (-1.40) (-2.98) (-1.69) ABACCt 0.271*** 0.102*** 0.137*** (4.74) (4.12) (2.92) EPSt 0.063*** 0.016** 0.036*** (3.96) (2.34) (2.95) SEPRTt -0.000 -0.000 -0.000 (-0.64) (-0.07) (-0.70) NSKEWt 0.032*** 0.014*** 0.017*** (4.79) (4.95) (3.18) SIGMAt 0.381 0.870* 3.845*** (0.33) (1.77) (4.61) RETt 2.250*** 0.772*** 1.313*** (11.17) (9.65) (9.49) Constant -0.541*** 1.316*** 0.806*** 1.389*** -0.179*** 0.589*** (-86.30) (9.43) (320.48) (22.98) (-39.34) (5.64) Observations 26,768 26,768 26,768 26,768 26,768 26,768 AdjR2 0.056 0.130 0.059 0.098 0.031 0.049 Industry FE CONTROL CONTROL CONTROL CONTROL CONTROL CONTROL Year FE CONTROL CONTROL CONTROL CONTROL CONTROL CONTROL 5.2. Excluded fluctuating years In order to avoid the impact of stock price fluctuation year data on the empirical results of this paper, the data of 2008, 2009, 2015 and 2016 are excluded in this paper, and the regression results are shown in Table 5. Table 5. Robustness test2 (1) (2) (3) NSKEWt+1 DUVOLt+1 COUNTt+1 ANANUMt 0.074*** 0.022*** 0.042*** (11.779) (8.442) (8.69) LEVt -0.036 -0.031** 0.002 (-1.080) (-2.188) (0.08) SIZEt -0.059*** -0.015*** -0.028*** (-8.415) (-4.948) (-5.40) ROAt -0.970*** -0.302*** -0.497*** (-5.715) (-4.118) (-3.79) MTOBt 0.040*** 0.015*** 0.023*** (5.974) (4.937) (4.46) DTURNt -0.005 -0.014** -0.011 (-0.347) (-2.367) (-1.02) ABACCt 0.307*** 0.108*** 0.163*** (4.254) (3.548) (2.72) EPSt 0.081*** 0.019** 0.055*** (4.402) (2.389) (3.91) SEPRTt -0.000 0.000 -0.000 (-0.178) (0.584) (-0.75) NSKEWt 0.032*** 0.013*** 0.014** (4.288) (4.135) (2.38) SIGMAt -0.321 1.432** 3.209*** (-0.247) (2.547) (3.38) RETt 1.972*** 0.789*** 1.147*** (9.166) (9.184) (7.51) Constant 0.963*** 1.147*** 0.379*** (5.910) (16.354) (3.15) Observations 19,760 19,760 19,760 Adjusted R-squared 0.115 0.075 0.042 Industry FE CONTROL CONTROL CONTROL Year FE CONTROL CONTROL CONTROL 213 5.3. Firm fixed effects and high-dimensional fixed effects In order to eliminate individual differences, improve the explanatory power of the model, reduce the endogeneity problem of this paper, and improve the accuracy and reliability of the model estimation in this paper, the firm fixed effect and the high latitude fixed effect are added in this paper, and the results are shown in Table 6. Columns (1), (2), and (3) added the annual and individual fixed effects, with coefficients of 0.034, 0.011, and 0.019, respectively, and were significant at the 1% level, Columns (4), 5, and 6 add the high latitude fixed effect and firm fixed effect, and the stock price crash risk index coefficients are 0.036, 0.012 and 0.020, respectively, which are still significant at the level of 1%. Table 6. Robustness test3 (1) (2) (3) (4) (5) (6) NSKEWt+1 DUVOLt+1 COUNTt+1 NSKEWt+1 DUVOLt+1 COUNTt+1 ANANUMt 0.034*** 0.011*** 0.019*** 0.036*** 0.012*** 0.020*** (4.43) (3.46) (3.26) (4.60) (3.60) (3.39) LEVt -0.083* -0.041* -0.032 -0.062 -0.033 -0.025 (-1.66) (-1.95) (-0.84) (-1.21) (-1.53) (-0.63) SIZEt 0.026* 0.018*** 0.017 0.037*** 0.023*** 0.023** (1.92) (3.20) (1.64) (2.65) (3.82) (2.11) ROAt -0.728*** -0.202*** -0.343** -0.769*** -0.211*** -0.386*** (-4.21) (-2.76) (-2.51) (-4.42) (-2.87) (-2.81) MTOBt 0.063*** 0.027*** 0.034*** 0.062*** 0.027*** 0.032*** (9.10) (8.72) (6.27) (8.75) (8.43) (5.81) DTURNt 0.008 -0.008 -0.009 0.007 -0.009 -0.009 (0.59) (-1.48) (-0.90) (0.54) (-1.51) (-0.88) ABACCt 0.225*** 0.086*** 0.131** 0.236*** 0.091*** 0.134** (3.57) (3.13) (2.55) (3.68) (3.21) (2.53) EPSt 0.059*** 0.015* 0.032** 0.056*** 0.012 0.033** (2.86) (1.77) (2.02) (2.67) (1.44) (2.04) SEPRTt 0.001 0.001 0.000 0.001 0.001 0.000 (1.06) (1.47) (0.30) (0.89) (1.13) (0.10) NSKEWt -0.079*** -0.030*** -0.049*** -0.083*** -0.031*** -0.051*** (-11.42) (-10.38) (-8.86) (-11.99) (-10.67) (-9.15) SIGMAt -3.878*** 0.125 0.982 -3.241** 0.360 1.369 (-3.06) (0.24) (1.07) (-2.49) (0.66) (1.45) RETt 1.674*** 0.660*** 0.896*** 1.798*** 0.695*** 0.969*** (7.87) (7.94) (6.06) (8.16) (8.07) (6.31) Constant -0.823*** 0.430*** -0.563** -1.096*** 0.330** -0.692*** (-2.75) (3.37) (-2.45) (-3.53) (2.53) (-2.94) Observations 26,388 26,388 26,388 26,375 26,375 26,375 AdjR2 0.157 0.127 0.063 0.168 0.135 0.068 Year FE CONTROL CONTROL CONTROL ID FE CONTROL CONTROL CONTROL CONTROL CONTROL CONTROL Year*Ind FE CONTROL CONTROL CONTROL 6. The alleviating effect of trust in the local community It was mentioned earlier that regional social trust can suppress management's concealment of negative information (Liu et al., 2016) and can also enhance the accounting conservatism of companies in that region (Li et al., 2017; Shen , 2019), indicating that in areas with a higher level of social trust, the risk of stock price collapse for companies is reduced. In order to test this conjecture, this paper cross- multiplies two dummy variables that measure regional social trust with the explanatory variables of this paper, and the results are shown in Table 7. From the data in Table 7, it can be seen that the coefficients of the interaction terms between the regional social trust index TRUSTDUMM1 and the analyst attention index are -0.024, -0.010 and -0.013, respectively. The coefficients of the interaction terms between the TRUSTDUMM2 of the regional social trust index and the analyst attention index were -0.016, -0.008 and -0.014, respectively, and all of them were significant at the level of 1%, indicating that regional social trust would alleviate the positive effect between analysts' attention and the risk of stock price crash, and verify hypothesis H2. 214 Table 7. Mechanism Inspection (1) (2) (3) (4) (5) (6) NSKEWt+1 DUVOLt+1 COUNTt+1 NSKEWt+1 DUVOLt+1 COUNTt+1 ANANUMt 0.084*** 0.027*** 0.048*** 0.079*** 0.026*** 0.048*** (10.93) (8.48) (8.30) (10.46) (8.17) (8.37) TRUSTDUMM1 0.055*** 0.021*** 0.032** (3.13) (3.02) (2.43) ANANUMtΓ—TRUSTDUMM1 -0.024*** -0.010*** -0.013** (-2.86) (-2.86) (-1.99) TRUSTDUMM2 0.048*** 0.020*** 0.035*** (2.73) (2.78) (2.63) ANANUMtΓ—TRUSTDUMM2 -0.016* -0.008** -0.014** (-1.96) (-2.15) (-2.10) LEVt 0.013 -0.010 0.027 0.013 -0.010 0.027 (0.45) (-0.85) (1.15) (0.45) (-0.85) (1.15) SIZEt -0.075*** -0.024*** -0.037*** -0.075*** -0.024*** -0.037*** (-12.50) (-9.25) (-8.33) (-12.51) (-9.25) (-8.33) ROAt -0.786*** -0.244*** -0.362*** -0.789*** -0.245*** -0.362*** (-5.56) (-4.00) (-3.39) (-5.60) (-4.02) (-3.39) MTOBt 0.032*** 0.012*** 0.018*** 0.031*** 0.012*** 0.018*** (6.06) (5.20) (4.60) (6.04) (5.19) (4.57) DTURNt -0.011 -0.013*** -0.012 -0.011 -0.013*** -0.012 (-0.93) (-2.61) (-1.33) (-0.95) (-2.63) (-1.35) ABACCt 0.277*** 0.105*** 0.140*** 0.276*** 0.104*** 0.141*** (4.84) (4.22) (2.99) (4.83) (4.21) (2.99) EPSt 0.079*** 0.022*** 0.045*** 0.079*** 0.022*** 0.045*** (4.95) (3.23) (3.78) (4.99) (3.26) (3.79) SEPRTt -0.000 -0.000 -0.000 -0.000 -0.000 -0.000 (-0.83) (-0.22) (-0.86) (-0.83) (-0.22) (-0.85) NSKEWt 0.033*** 0.015*** 0.017*** 0.033*** 0.015*** 0.017*** (5.01) (5.20) (3.36) (5.03) (5.21) (3.36) SIGMAt 0.347 0.888* 3.832*** 0.345 0.889* 3.846*** (0.30) (1.80) (4.59) (0.30) (1.80) (4.60) RETt 2.242*** 0.774*** 1.309*** 2.241*** 0.774*** 1.311*** (11.10) (9.64) (9.44) (11.09) (9.64) (9.45) Constant 1.241*** 1.342*** 0.538*** 1.249*** 1.344*** 0.536*** (8.88) (22.22) (5.17) (8.92) (22.24) (5.15) Observations 26,768 26,768 26,768 26,768 26,768 26,768 AdjR2 0.129 0.097 0.048 0.129 0.097 0.048 Industry FE CONTROL CONTROL CONTROL CONTROL CONTROL CONTROL Year FE CONTROL CONTROL CONTROL CONTROL CONTROL CONTROL 7. Conclusion and Recommendations 7.1. Conclusion (1) The analysts in our country's financial market are unable to fully perform their role as information intermediaries, failing to act as interpreters and transmitters of information between investors and listed companies. The increased attention on analysts may lead company decision- makers to make more decisions to conceal negative news, thereby increasing the risk of a stock price collapse for enterprises. (2) In regions with a higher level of social trust, the positive relationship between analysts' attention and the risk of stock price crashes significantly weakens. This is because social trust, as an informal institution, suppresses the management of companies in these regions from concealing negative news, thereby enhancing the transparency of corporate information and the robustness of accounting, which in turn reduces the risk of stock price crashes for listed companies in these areas. 7.2. Recommendations (1) Currently, analysts in our country merely serve the role of conveying information from the capital market, and are unable to effectively analyze and interpret this information, thereby failing to fully exert their role as supervisors. It is necessary to enhance the professionalism of analysts, strengthen the regulation of analysts, and establish their authority and credibility. (2) Regional social trust, as an external institution, can reduce information asymmetry, inhibit management's behavior of concealing adverse information, enhance the transparency of accounting information, and thereby reduce the risk of abnormal declines in stock prices. Therefore, it is necessary to maintain a positive relationship between social trust, the quality of corporate governance, investor protection, and the degree of improvement of formal systems, in order to uphold the stability of the capital market. 215 References [1] Amy P. Hutton, Alan J. Marcus, Hassan Tehranian, Opaque financial reports, R2, and crash risk[J], Journal of Financial Economics, 2009 [2] Chunfang Cao, Changyuan Xia, Kam C. Chan, Social trust and stock price crash risk: Evidence from China[J]. International Review of Economics & Finance, 2016 [3] Cui Bing, San Chongwen. Media Reports, Analyst Attention, and the Risk of Stock Price Collapse[J]. Journal of Hubei University of Technology, 2022, 37 (06): 43-48. [4] Degeorge F, Ding Y, Jeanjean T, et al. Analyst coverage, earnings management and financial development: An international study[J]. Journal of Accounting and Public Policy, 2013 [5] Han Yanjin, Feng Xiaoqing, Song Jianbo. Analyst Attention and Stock Price Crash Risk Based on Information Generation Process[J]. Journal of Management, 2021 [6] Jeong-Bon Kim, Yinghua Li, Liandong Zhang, Corporate tax avoidance and stock price crash risk: Firm-level analysis[J]. Journal of financial Economics, 2011 [7] Li Chuntao, Zhao Yi, Xu Xin, Li Qingyuan. Pressing the Gourd and Floating the Dipper: Analyst Following and the Choice of Earnings Management Methods[J]. Financial Research, 2016, (04): 144-157. [8] Li Jin, Stewart C. Myers, R2 around the world: New theory and new tests[J], Journal of Financial Economics, 2006 [9] Liang Quanxi, Zeng Haijian. Reform of the Independent Director System, Independence of Independent Directors, and Stock Price Crash Risk[J]. Management World, 2016, (03): 144-159. [10] Liu Baohua, Luo Hong, Zhou Wei, Yang Xing. Social Trust and Stock Price Crash Risk[J]. Journal of Finance and Economics, 2016, (09): 53-66. [11] Meng Qingbin, Yang Junhua, and Lu Bing. The Information Content of Management Discussion and Analysis Disclosure and Stock Price Crash Risk: A Study Based on Text Vectorization Methods [J]. China Industrial Economics, 2017, (12): 132-150. [12] Nie Jun, Ran Rong. Social Trust in Regional Society and the Fulfillment of Corporate Social Responsibility by Business Clients [J]. Journal of Zhongnan University of Economics and Law, 2020, (05): 137-146+160. [13] Pan Yue, Dai Yiyi, Lin Chaoqun. Information Opacity, Analyst Coverage, and Individual Stock Crash Risk[J]. Financial Research, 2011, (09): 138-151. [14] Pei Ping, Fu Shun, Zhu Hongbing. Analyst Coverage, Cash Flow Risk, and Stock Price Crash[J]. Journal of East China Normal University (Philosophy and Social Sciences Edition), 2021, 53 (01): 142-154+174. [15] Shen Danlin. Social trust and corporate risk-taking [J]. Economic Management,2019 [16] Tian Xuezhen. The impact of analysts' attention on stock price crash risk: An empirical test based on China's A-share market[D].Hunan University,2020. [17] Xiaorong Li, Steven Shuye Wang, Xue Wang, Trust and stock price crash risk: Evidence from China[J]. Journal of Banking & Finance, 2017 [18] Xu Nianhang, Jiang Xuanyu, Yi Zhihong, et al. Analysts' conflict of interest, optimism bias and stock price crash risk[J].Economic Research Journal,2012 [19] Yang Wei, Song Min, Feng Ke. M&A Goodwill, Investor Overreaction, and Stock Price Bubbles and Crashes [J]. China Industrial Economics, 2018, (06): 156-173. [20] Zhou Mingshan, Lin Jing, Xu Nianxing. Analyst Tracking and Stock Price Synchronization: Evidence from the Perspective of Overreaction [J]. Journal of Management Science, 2016