Working Papers

Nowcasting GDP using tone-adjusted time varying news topics: Evidence from the financial press

Published:

We extract tone-adjusted, time-varying and hierarchically ordered topics from a large corpus of Dutch financial news and investigate whether these topics are useful for monitoring the business cycle and nowcasting GDP growth in the Netherlands. The financial newspaper articles span the period January 1985 up until January 2021. Our newspaper sentiment indicator has a high concordance with the business cycle. Further, we find newspaper sentiment increases the accuracy of our nowcast for GDP growth using a dynamic factor model, especially in periods of crisis. We conclude that our tone-adjusted newspaper topics contain valuable information not embodied in monthly indicators from statistical offices. Read more

Nowcasting GDP using machine learning methods

Published:

This paper compares the ability of several econometric and machine learning methods to nowcast GDP in (pseudo) real-time. The analysis takes the example of Dutch GDP over the years 1992-2018 using abroad data set of monthly indicators. It discusses the forecast accuracy but also analyzes the use of information from the large data set of regressors. We find that the random forest forecast provides the most accurate nowcasts while using the different variables in a relative stable and equal manner. Read more

Business investment in EU countries

Published:

The article analyses recent developments in business investment for a large group of EU countries, using a broad set of analytical tools and data sources. We find that the assessment of whether or not investment is currently low varies across benchmarks and countries. At the euro area level and for most countries, the level of business investment is broadly in line with the level of overall activity. However rates of capital stock growth have slowed down since the crisis. The main cyclical determinants of investment developments in the euro area include foreign and domestic demand, uncertainty and financial conditions. Uncertainty seems to have played a negative role during the financial and sovereign debt crises; however, given its low levels more recently, it has not acted as a drag on business investment overall during the recovery. Credit constraints appear to have hindered investment during the twin crises, especially in stressed countries. Aside from cyclical developments, important secular factors – relating to demographics, the changing nature and location of production, and the business environment – have influenced investment. Another factor that may have amplified the decline in private investment, particularly in countries that were hit hardest by the sovereign debt crisis, is the low level of public investment. This is because when public investment enhances the productivity of the private sector, there may be positive spillovers from the former to the latter, including across countries. Finally, intra-sector capital misallocation, measured as the within-sector dispersion across firms in the marginal revenue product of capital, has been increasing in Europe since 2002, which may in turn have exerted a significant drag on total factor productivity dynamics, and hence on aggregate output growth. Read more

Forecasting GDP growth in times of crisis: private sector forecasts versus statistical models

Published:

This chapter examines the forecasting accuracy of linear statistical models tailored to forecast GDP growth in the adjacent quarters, with a special focus on the dot-com recession of 2001-2002 and the financial crisis of 2008-2009. The forecasting accuracy of the mechanical models is confronted with the forecasting accuracy of professional analysts, and it is evaluated whether a combination of the two can improve forecasting performance. The analysis covers the Netherlands over the years 1995{2010. Overall, the recently proposed dynamic factor model showed the highest forecasting accuracy of all models considered. During the recent financial crisis, the forecasting accuracy of all models deteriorated. The dynamic factor model shows the smallest deterioration of all models considered. Interestingly, enhancing the forecasts of the dynamic factor model with the judgmental forecasts of professional analysts can increase the forecasting accuracy, but only during the financial crisis. Read more

Preferences for redistribution in the Netherlands

Published:

We investigate the determinants of Dutch households’ preferences for income redistribution, using survey data. Our results show that support for redistributive policies is related to self-interest, exposure to misfortune and risk-aversion. In addition, people who believe that prosperity is primarily due to luck rather than hard work to favour redistribution, indicating that equal opportunities are considered important. Interestingly, support for redistributive policies is positively related educated to eduction, while the impact of age is ambiguous. This is an important outcome, as it implies that globalisation and skill-biased technological progress may put less pressure on the Dutch social security system than previously assumed. Read more