Recently, ensemble-based machine learning models have been widely adopted and have demonstrated their effectiveness in bankruptcy prediction. However, these algorithms often function as black boxes, making it difficult to understand how they generate forecasts. This lack of transparency has led to growing interest in interpretability methods within artificial intelligence research.

In this paper, we assess the predictive performance of Random Forest, LightGBM, XGBoost, and NGBoost (Natural Gradient Boosting for probabilistic prediction) on French firms across various industries, with a forecasting horizon of one to five years. We then apply Shapley Additive Explanations (SHAP), a model-agnostic interpretability technique, to explain XGBoost, one of the best-performing models in our study. SHAP highlights the contribution of each feature to the model’s predictions, enabling a clearer understanding of how financial and macroeconomic factors influence bankruptcy risk. Moreover, it allows for the explanation of individual predictions, making black-box models more applicable in credit risk management.

We experimentally investigate whether and how the potential presence of algorithmic trading (AT) in human-only asset markets can influence humans’ price forecasts, trading activities and price dynamics. Two trading strategies commonly employed by high-frequency traders, spoofing (SP) – associated with market manipulation – and market making (MM) – seen as liquidity provision – are considered.

This study investigates how green investment assets improve optimal portfolio diversification in terms of tail downside risk. We use the wavelet conditional value-at-risk ratio to explore the benefits of adding green assets to conventional portfolios. We quantify risk based on the contagion between conventional stock market indices and green environmental assets, including a sustainability index, clean energy, and green bonds. Our findings emphasize the high variance between conventional stock pairs, providing evidence of contagion effects before and during the COVID-19 pandemic. We show that including clean energy and green bond indices in conventional portfolios reduces the extreme risk of portfolios. In addition, we find that the diversification benefits of clean energy, green bonds, and safe-haven investments apply especially in the short term during the pandemic. Finally, we show that the considered portfolios could not decrease long-term risk during the COVID-19 crisis because of the systematic risk spread. Our portfolio optimization design supports the superiority of clean energy and green bonds in portfolio diversification over the sustainability index. These insights can be used by portfolio managers to inform diversification in different investment horizons.

Increasing awareness of climate change and its potential consequences on financial markets has led to interest in the impact of climate risk on stock returns and portfolio composition, but few studies have focused on perceived climate risk pricing. This study is the first to introduce perceived climate risk as an additional factor in asset pricing models. The perceived climate risk is measured based on the climate change sentiment of Twitter dataset with 16 million unique tweets in the years 2010–2019. One of the main advantages of our proxy is that it allows us to capture both physical and transition climate risks. Our results show that perceived climate risk is priced into S&P 500 Index stock returns and is robust when different asset-pricing models are used. Our findings have implications for market participants, as understanding the relationship between perceived climate risk and asset prices is crucial for investors seeking to navigate the financial implications of climate change, and for policymakers aiming to promote sustainable financing and mitigate the potential damaging effects of climate risk on financial markets, and a pricing model that accurately incorporates perceived climate risk can facilitate this understanding.

Financial assets are monetary instruments that can be traded in financial markets, such as stocks, bonds, and mutual funds. They allow investors to generate income or capital appreciation. Unlike physical assets, financial assets derive their value from contractual claims.

A financial market is a marketplace where financial assets like stocks, bonds, and commodities are traded. These markets help businesses and governments raise capital and enable the efficient allocation of financial resources. Financial markets are vital for economic growth.

Fundamental analysis involves assessing a stock’s intrinsic value by reviewing financial statements, industry trends, and economic factors. It helps investors determine whether a stock is undervalued or overvalued, guiding their investment decisions.

Technical analysis involves studying historical price movements and trading volumes to predict future trends in financial assets. It relies on charts and technical indicators and does not consider the intrinsic value of a company.

Risk in investing refers to the chance that actual returns may differ from expected ones, possibly leading to a loss of capital. High-risk investments often offer higher potential returns but are more volatile. Diversification is a key method for managing risk by spreading investments across different asset classes.

When investors hold disproportionately high carbon emitters with associated increased carbon risk, a positive relationship exists between a firm’s carbon emissions and the association between the stock returns and dividend payment. If investors hold disproportionately high carbon emitters with the associated increased carbon risk stocks, the stock market reacts less positively (more negatively) to dividend increase (decrease) announcements. At the same time, if firms under-price their carbon risk, the stock market reacts less positively (more negatively) to dividend increase (decrease) announcements.