We compile a unique dataset linking micro price data underlying the official Swedish producer price index with administrative firm level data and provide new evidence on the inflationary effects of global supply chain shocks. For identification, we interact exogenous shocks to global supply chains, obtained through a VAR model, with firm-specific import shares. Shocks to global supply chains lead to a significant and persistent increase in producer prices with a peak response after two years. Importantly, average responses mask heterogeneous responses across firms. Relatively larger firms, firms with lower labor costs and a higher market share raise prices more strongly.
Abstract Mathias Klein Firm heterogeneity in financial constraints is a quantitatively important driver of how monetary policy transmits to inflation. Using detailed microdata on Swedish public and private firms, and high-frequency monetary policy surprises around Riksbank announcements, we document that smaller, financially constrained firms adjust prices significantly less than larger firms in response to changes in monetary policy. This heterogeneous price response materially dampens the aggregate PPI inflation response to monetary policy. Models of customer markets and financial frictions can explain our findings: because the external finance premium rises after a monetary contraction, constrained firms cut prices less to preserve cash flows, sacrificing future market share. Additional evidence on heterogeneous sales, debt, marginal cost, and markup responses further supports this channel. We consider several alternative explanations, including differences in price adjustments, working capital, market share, and export share, but these cannot rationalize our main heterogeneity result.
We estimate the demand for transactional and non-transactional cash balances (banknotes and coins) in Canada, Denmark, Iceland, Sweden and Norway over the last decades exploiting the seasonality of cash demand. These countries share many features that are relevant for cash demand, but nevertheless show large differences in terms of aggregate cash balances. While Canada, Iceland and Denmark have seen increased aggregate cash balances, Norway and especially Sweden have seen a dramatic decline. We find that transactional balances have decreased somewhat in all of the countries and the differences in aggregated cash balances is due to differences in the development of non-transactional cash balances. We argue that different de facto legal tender status, crisis exposures, foreign demand and cash supply-side policies help explain these findings.
We investigate the growth-finance nexus in an endogenous growth model with search frictions and congestion effects in credit and innovation markets. The interaction between these two frictions generates a non-monotonic, inverted-U relationship between financial development and growth. Financial development exerts two opposing forces: a direct positive finance channel by easing access to funding and an indirect negative congestion channel, as greater financial activity draws more firms into R&D competition for scarce innovation resources. The resulting hump-shaped pattern implies that excessive financial development can slow technological progress. The mechanism remains robust when allowing for firm heterogeneity that can generate composition effects. In a calibration close to the U.S. economy, the impact of finance on growth is negative but quantitatively small, consistent with the observation that, over the last century, most advanced economies have experienced financial expansion alongside nearly constant GDP growth rates.
Predicting current and near-term macroeconomic developments using linear indicator models and factor models, together with Economic Tendency Survey data, is standard nowcasting practice. In the current article, it is investigated whether machine learning (ML) methods, when used together with a limited set of tendency survey confidence indicators, can improve the forecasts of Swedish quarterly GDP growth compared with linear indicator models and factor models. The results indicate that ML methods generally perform relatively well. In particular, gradient-boosted regression trees, random forests, and multilayer perceptron models are identified as some of the best performing models. The results indicate that ML models can be fruitfully applied in macroeconomic forecasting without employing vast amounts of data. One factor contributing to this could be the ability of ML methods to capture nonlinearities. Results also indicate that, when implementing ML models, care should be taken in determining how often central model parameters should be tuned and estimated.
