Economics Letters最新文献

We use stock-day level data on the market share of algorithmic trading to analyze whether algorithmic trading affects the frequency of mini flash crashes in the Austrian stock market. We use an instrumental variables approach and the Petrin and Train (2010) control function approach to address endogeneity concerns. We find no evidence that algorithmic trading significantly affects the probability of the occurrence of mini flash crashes.

Financial markets are known to respond to ECB policy announcements, yet these reactions could often reflect adjustments to heightened market expectations influenced by prior media coverage. News may have gained importance following the ECB's adoption of a “data-dependent” policy regime amidst persistent inflation in the Euro Area. Our empirical findings confirm this and underscore the complex interactions between media news and expectations in shaping market dynamics and reactions both before and during official policy announcements.

We investigate the belief-based mechanism that potentially drives the equity home bias puzzle using a novel experimental design. We find that geographical information causally shapes investor beliefs, leading to a home bias effect in beliefs. Participants expect higher returns and lower risk from local stocks compared to non-local ones. Loss probability strongly predicts risk perception, but its effect weakens when the stock is local. Yet, we do not find a home bias in investment choices.

Based on a sample of Chinese heavily polluting firms from 2008 to 2019, we examine the impact of the adoption of the emissions rights mortgage (ERM) policy on firm-level green investment. We find that an ERM policy significantly enhances firms’ green investment. This effect is more salient when financial constraints are higher, local marketization levels are higher, and local governments’ environmental concerns are higher. Our study provides empirical evidence and policy implementations for the further development of financial instruments based on emissions rights and other environmental rights, which show that ERM is a sound policy to promote a sustainable environment.

We examine the impact of mergers on product quality in the U.S. agri-food industry. Our difference-in-differences analyses provide evidence that following a one standard deviation increase in merger activity there is a 48.1% decrease in pounds of food recalled and an 18.7% decrease in the total number of recalls. A significant implication of our analysis is that increased concentration increases product quality.

Will the adoption of AI by businesses substitute for worker tasks or jobs? This is a core question for which relatively scarce evidence exists—especially in the wake of recent advances in generative AI. Using a new large-scale business survey by the U.S. Census Bureau, we find that AI use is having a much greater impact on worker tasks than on employment levels at the firm level. About 27% of firms using AI report replacing worker tasks, but only about 5% experience employment change due to AI use. These rates are expected to increase to nearly 35% and 12%, respectively, in the near future.

This paper argues that the performance of the Synthetic Control Method (SCM) can depend on the initial set of control units (donors) due to a bias–variance trade-off. A forward stepwise selection algorithm, validated via cross-validation, may reduce prediction error by suggesting to use fewer donors.

Survey-based indicators of labour shortages are often based on the number of respondents that experience a shortage – the prevalence – overlooking the crucial element of shortage intensity. Using data from a large representative business survey in Sweden, this paper constructs an establishment-level measure of labour shortage intensity, by considering the number of positions in shortage and total employment at the establishment. According to the data, the average intensity of labour shortages has decreased by 32% since 2007. A new aggregate indicator of relative labour shortages (RLS) that accounts for both the average intensity and prevalence of labour shortages indicated more subdued labour market conditions in Sweden during the recovery from the Great Recession and the COVID-19 pandemic than other indicators. Failing to account for the decline in the intensity of labour shortages mostly affects labour shortages in industry and less in services sectors.

We incorporate model uncertainty into the intertemporal realized utility model proposed by Ingersoll and Jin (2013). We find: (1) model uncertainty erodes the investor’s value and makes him more willing to take the risk; (2) when ambiguity aversion is sufficiently high, voluntary realization of losses will not occur; (3) model uncertainty accelerates sales, thus shortening the average holding period and strengthening the disposition effect; (4) the disposition effect is weaker in the lower volatility case under model uncertainty, which is consistent with the empirical results; (5) model uncertainty alters the effect of volatility on the sale decision.