RESEARCH

"Detecting accounting fraud using machine learning algorithms with MD&A textual information” with S. Chen and Y. Dong, R&R.

Abstract: In this paper, we evaluate the role of Management Discussion and Analysis (MD&A) in detecting accounting fraud in China. Employing advanced machine learning techniques, including the RUSBoost based on decision tree (RUSBoost-DT) algorithm, we examine the informativeness of MD&A to regulators, a crucial yet often overlooked information user in the capital market, in detecting fraudulent activities by Chinese listed firms. Using a comprehensive dataset spanning from 2007 to 2019, we reveal that incorporating MD&A information alongside traditional financial information does not significantly improve the detection of accounting fraud in China, in stark contrast to the empirical evidence in the U.S. (Purda and Skillicorn, 2015; Hoberg and Lewis, 2017; Brown et al., 2020). Furthermore, by exploiting an exogenous regulatory change to MD&A requirements in China, we document a significant improvement in the predictive effectiveness of MD&A information in the post-regulation period. The study further explores the usefulness of MD&A information in fraud detection between State-Owned Enterprises (SOEs) and non-SOEs. Overall, our study highlights the pressing need to enhance corporate disclosure quality, thereby facilitating the detection of fraud by regulators and safeguarding the integrity of the Chinese capital market.

"Can Q theory explain a firm's green investments?: Evidence from China," with S. Ding, X. Zhang and Y. Zhou. 

"Deep GARCH," with S. Yang and C. Cai.

Ding, S., Kim, M., Zhang, X., & Zhou, Y. (2025). The impact of cash flow uncertainty on investment-cash flow sensitivity in China: The debt financing channel. International Journal of Finance and Economics, forthcoming. https://dx.doi.org/10.1002/ijfe.3103

Abstract: Chinese firms' investment-cash flow sensitivity (ICFS) declined during the global financial crisis (GFC), contradicting the conventional financial constraint interpretation of ICFS. We analyse this phenomenon by examining how cash flow uncertainty affects financing investment. We find that ICFS reveals not only the relationship between investment and cash flow but also that between internal funds and debt financing. When internal funds and debt financing are complementary, cash flow uncertainty decreases ICFS more than when they are substitutes. The relationship between internal funds and debt financing weakens when cash flow uncertainty rises. A natural experiment based on the GFC and the post-GFC economic stimulus package confirms the causal relationship between cash flow uncertainty and ICFS.

Dong, Y., Kim, M., & Chen, X. (n.d.). Can corporate governance information facilitate accounting fraud detection in China? Evidence from machine learning. The International Journal of Accounting, Accepted.

Abstract: Using machine learning techniques, we find that incorporating corporate governance information with raw financial information improves accounting fraud detection. Specifically, our further analyses indicate that board structure and the personality traits of executives are crucial for identifying fraudulent activities across the five corporate governance categories. Moreover, chairperson age, performance-based composition, and ownership concentration are particularly relevant corporate governance factors when it comes to accounting fraud detection. Our findings suggest that non-financial information, such as corporate governance factors, provides incremental value beyond financial statements that can be particularly useful to regulators as important yet ignored information users in the capital market.

Zhang, X., Kim, M., Yan, C., & Zhao, Y. (2024). Default dependence in the insurance and banking sectors: A copula approach.  Journal of International Financial Markets, Institutions and Money, 91, 101911. https://doi.org/10.1016/j.intfin.2023.101911

Abstract: We employ a time-varying asymmetric copula model that combines the generalized autoregressive score model with the generalized hyperbolic skewed t copula to capture the dynamics and asymmetry of default dependence between insurers and banks. We identify the term structure of default dependence between these two sectors. The short-term and long-term dependence of default risk rise and converge during financial crises. We explore the determinants of the time-series variation in default dependence. While traditional macro variables can explain only a small portion of the variation in default dependence, we find a significant negative correlation between default dependence and global geopolitical risk.

Kim, M., Yang, J., Song, P., & Zhao, Y. (2020). The dependence structure between equity and foreign exchange markets and tail risk forecasts of foreign investments. Quantitative Finance, 21(5), 815–835. https://doi.org/10.1080/14697688.2020.1812701 

Abstract: Motivated by the importance of the dependence structure between equity and foreign exchange rates in international financial markets, we investigate whether modelling the dependence structure can help forecast the tail risk of foreign investments. We propose a new time-varying asymmetric copula for modelling the dependence structure and forecasting the tail risk. We conduct backtesting on our tail risk forecasts for 12 major developed and emerging markets. We find that modelling the dependence structure can improve the tail risk forecast and make risk management of foreign investments more robust.

Cai, C., Kim, M., Shin, Y., & Zhang, Q. (2019). FARVaR: Functional Autoregressive Value-at-Risk. Journal of Financial Econometrics, 17(2), 284–337. https://doi.org/10.1093/jjfinec/nby031

Abstract: Motivated by the stylized fact that intraday returns can provide additional information on the tail behavior of daily returns, we propose a functional autoregressive value-at-risk (VaR) approach which can directly incorporate such informational advantage into the daily VaR forecast. Our approach leads to greater flexibility in modeling the dynamic evolution of the density function of intraday returns and the ability to capture substantial swings in the tails following major events. We comprehensively evaluate our proposed model using intraday transaction data and demonstrate that it can improve coverage ability, reduce economic cost, and enhance statistical reliability in market risk management.

Ding, S., Kim, M., & Zhang, X. (2018). Do firms care about investment opportunities? Evidence from China. Journal of Corporate Finance, 52, 214-237. https://dx.doi.org/10.1016/j.jcorpfin.2018.07.003

Abstract: What drives a firm's investment decisions in China? While most literature focuses on the role of financial factors (such as cash flow), we explore this most important question in corporate finance from the perspective of economic fundamentals. Using a large number of proxies for investment opportunities and a variety of econometric approaches, our empirical results show that it is private firms that make the most of all types of investment opportunities in China. State-owned enterprises respond more to the investment opportunities from the supply side, but much less so to demand-side shocks and future profitability. Financial sector development is found to be conducive to the improvement of the investment efficiency of private firms by making them take better advantage of all types of investment opportunities in their decision-making. Our research calls for further institutional and financial sector reforms in China.

Cerrato, M., Crosby, J., Kim, M., & Zhao, Y. (2017). The joint credit risk of UK global‐systemically important banks. Journal of Futures Markets, 37(10), 964-988. https://dx.doi.org/10.1002/fut.21855

Abstract: We study the joint credit risk in the UK banking sector using the weekly CDS spreads of global systemically important banks over 2007–2015. We show that the time‐varying and asymmetric dependence structure of the CDS spread changes is closely related to the joint default probability that two or more banks simultaneously default. We are able to flexibly measure the joint credit risk at the high‐frequency level by applying the combination of the reduced‐form model and the GAS‐based dynamic asymmetric copula model to the CDS spreads. We also verify that much of the dependence structure of the CDS spread changes are driven by the market factors. Overall, our study demonstrates that the market factors are key inputs for the effective management of the systemic credit risk in the banking sector.

Cerrato, M., Crosby, J., Kim, M., & Zhao, Y. (2017). Relation between higher order comoments and dependence structure of equity portfolio. Journal of Empirical Finance, 40, 101-120. https://doi.org/10.1016/j.jempfin.2016.11.007

Abstract: We study a relation between higher order comoments and dependence structure of equity portfolio in the US and UK by relying on a simple portfolio approach where equity portfolios are sorted on the higher order comoments. We find that beta and coskewness are positively related with a copula correlation, whereas cokurtosis is negatively related with it. We also find that beta positively associates with an asymmetric tail dependence whilst coskewness negatively associates with it. Furthermore, two extreme equity portfolios sorted on the higher order comoments are closely correlated and their dependence structure is strongly time-varying and nonlinear. Backtesting results of value-at-risk and expected shortfall demonstrate the importance of dynamic modeling of asymmetric tail dependence in the risk management of extreme events.

Chaudhuri, K., Kim, M., & Shin, Y. (2016). Forecasting distributions of inflation rates: the functional auto-regressive approach. Journal of the Royal Statistical Society Series A: Statistics in Society, 179(1), 65-102. https://doi.org/10.1111/rssa.12109

Abstract: In line with recent developments in the statistical analysis of functional data, we develop the semiparametric functional auto-regressive modelling approach to the density forecasting analysis of national rates of inflation by using sectoral inflation rates in the UK over the period January 1997–September 2013. The pseudo-out-of-sample forecasting evaluation and test results provide an overall support to superior performance of our proposed models over the aggregate auto-regressive models and their statistical validity. The fan chart analysis and the probability event forecasting exercise provide further support for our approach in a qualitative sense, revealing that the modified functional auto-regressive models can provide a complementary tool for generating the density forecast of inflation, and for analysing the performance of a central bank in achieving announced inflation targets. As inflation targeting monetary policies are usually set with recourse to the medium-term forecasts, our proposed work may provide policy makers with an invaluably enriched information set.

Dang, V. A., Kim, M., & Shin, Y. (2015). In search of robust methods for dynamic panel data models in empirical corporate finance. Journal of Banking & Finance, 53, 84-98.

Abstract: We examine which methods are appropriate for estimating dynamic panel data models in empirical corporate finance. Our simulations show that the instrumental variable and GMM estimators are unreliable, and sensitive to the presence of unobserved heterogeneity, residual serial correlation, and changes in control parameters. The bias-corrected fixed-effects estimators, based on an analytical, bootstrap, or indirect inference approach, are found to be the most appropriate and robust methods. These estimators perform reasonably well even in models with fractional dependent variables censored at [0, 1]. We verify these results in two empirical applications, on dynamic capital structure and cash holdings.

Dang, V. A., Kim, M., & Shin, Y. (2014). Asymmetric adjustment toward optimal capital structure: Evidence from a crisis. International Review of Financial Analysis, 33, 226-242.

Abstract: We employ dynamic threshold partial adjustment models to study the asymmetries in firms' adjustments toward their target leverage. Using a sample of US firms over the period 2002–2012, we document a negative impact of the Global Financial Crisis on the speed of leverage adjustment. In our subperiod analysis, we find moderate evidence of cross-sectional heterogeneity in this speed, which seems more pronounced pre-crisis and provides little support for the financial constraint view. Over the pre-crisis period, more constrained firms, such as those with high growth, with large investment, of small size, and with volatile earnings, adjust their capital structures more quickly than their less constrained counterparts. These firms rely heavily on external funds to offset large financing deficits, suggesting that their higher adjustment speeds may be driven by lower adjustment costs that are shared with the transaction costs of accessing external capital markets. During the crisis, the speed of adjustment varies with the deviation from target leverage: only firms with sufficiently large deviations attempt to revert to the target, albeit slowly. Overall, our results provide new evidence of both cross-sectional and time-varying asymmetries in capital structure adjustments, which is consistent with the trade-off theory.

Chaudhuri, K., Greenwood‐Nimmo, M., Kim, M., & Shin, Y. (2013). On the Asymmetric U‐Shaped Relationship between Inflation, Inflation Uncertainty, and Relative Price Skewness in the UK. Journal of Money, Credit and Banking, 45(7), 1431-1449. https://dx.doi.org/10.1111/jmcb.12058

Abstract: We investigate the asymmetric relationships between aggregate inflation and the second and third moments of the cross-sectional distribution of relative prices using a modified Calvo pricing model with regime-dependent price rigidities. Calibration experiments reveal that the inflation-standard deviation and inflation-skewness relationships exhibit U-shaped asymmetries around the historical mean inflation rate. UK sectoral data support our results. We conclude that monetary policy should target an inflation rate proximate to the (common) minima of these nonlinear relationships and that core inflation measures should not be used for policy purposes as they exclude much of the information contained in the higher moments.

Dang, V. A., Kim, M., & Shin, Y. (2012). Asymmetric capital structure adjustments: New evidence from dynamic panel threshold models. Journal of Empirical Finance, 19(4), 465-482. https://doi.org/10.1016/j.jempfin.2012.04.004

Abstract: We develop a dynamic panel threshold model of capital structure to test the dynamic trade-off theory, allowing for asymmetries in firms' adjustments toward target leverage. Our novel estimation approach is able to consistently estimate heterogeneous speeds of adjustment in different regimes as well as to properly test for the threshold effect. We consider several proxies for adjustment costs that affect the asymmetries in capital structure adjustments and find evidence that firms with large financing imbalance (or a deficit), large investment or low earnings volatility adjust faster than those with the opposite characteristics. Firms not only adjust at different rates but also seem to adjust toward heterogeneous leverage targets. Moreover, we document a consistent pattern that firms undertaking quick adjustment are over-levered with a financing deficit and rely heavily on equity issues to make such adjustment.