De Nederlandsche Bank

We examine how households' euro-area inflation expectations reacted to the ECB's adoption of a symmetric inflation target in July 2021, and the subsequent inflation surge. Using a randomised control trial within a monthly representative survey, we elicit short- and long-term inflation expectations. We find that the ECB's strategy announcement had no material effect on expectations. In contrast, the sharp rise in actual inflation significantly raised point forecasts and expected probabilities of high inflation. These findings suggest that households' expectations became less well anchored and more sensitive to inflation realizations in the high-inflation period, likely reflecting inter alia reduced rational inattention.

We estimate the impact of tax shocks on output across different stages of the business cycle. We do this for a panel of nine advanced economies using a harmonized dataset of narratively identified exogenous tax changes and a smooth transition local projection model. The output response to an exogenous tax shock is significant, but only during economic expansions. In recessions, the tax multiplier is insignificant, both in the short- and long run. We also find that, during booms, output only responds to tax hikes and is unresponsive to tax cuts. The results on the state-dependent and asymmetric effects of tax shocks are robust to a number of alternative model specifications and definitions of the business cycle.

We explore the macroeconomic effects of climate policies promoting the green energy transition in the euro area using an extended version of the Euro Area and Global Economy (EAGLE) model. The model differentiates between brown and green energy sectors and incorporates carbon taxes and brown capital income taxes. We analyze scenarios with unilateral and globally coordinated carbon taxes, with and without revenue redistribution to green firms and financially constrained households. Carbon taxes act as negative supply shocks, raising inflation and lowering output, while subsidies to green energy firms reduce green energy prices, supporting the transition and easing recessions. Redistribution to constrained households boosts consumption but does not accelerate the green transition. Taxes on brown capital income lower both inflation and output by acting as demand shocks. Recycling revenue from this tax to subsidize green capital investment strengthens the shift to green energy and moderates economic contractions. Global coordination of carbon taxes delivers only modest additional macroeconomic effects compared to unilateral action, as substitution in energy use outweighs international spillovers. Sensitivity analyses confirm the robustness of these findings under alternative assumptions on price rigidity, substitution elasticities, and monetary policy.

We examine how investment funds use exchange-traded funds (ETFs) to manage their liquidity. We find that during the COVID-19 market turmoil, investment funds were selling more ETF shares than any other investor sector. Also, open-ended funds that faced larger outflows in March 2020 sold a higher share of their ETF holdings, consistent with the conjecture of open-ended funds selling ETF shares to raise cash in response to redemption requests. Moreover, ETFs with higher investment fund ownership experienced larger outflows in the primary market. Our findings point to a possible channel for liquidity contagion from open-ended funds to the ETF market.

Current anti-money laundering (AML) systems, predominantly rule-based, exhibit notable shortcomings in efficiently and precisely detecting instances of money laundering. As a result, there has been a recent surge toward exploring alternative approaches, particularly those utilizing machine learning. Since criminals often collaborate in their money laundering endeavors, accounting for diverse types of customer relations and links becomes crucial. In line with this, the present paper introduces a graph neural network (GNN) approach to identify money laundering activities within a large heterogeneous network constructed from real-world bank transactions and business role data belonging to DNB, Norway's largest bank. Specifically, we extend the homogeneous GNN method known as the Message Passing Neural Network (MPNN) to operate effectively on a heterogeneous graph. As part of this procedure, we propose a novel method for aggregating messages across different edges of the graph. Our findings highlight the importance of using an appropriate GNN architecture when combining information in heterogeneous graphs. The performance results of our model demonstrate great potential in enhancing the quality of electronic surveillance systems employed by banks to detect instances of money laundering. To the best of our knowledge, this is the first published work applying GNN on a large real-world heterogeneous network for anti-money laundering purposes.