I study the macroeconomic effects of a shock to oil supply expectations. I show that information effects, which are pervasive in many contexts, are also a prominent feature of the oil market. Failing to account for them can lead to biased estimates of the effects of the shock of interest. I exploit the institutional features of OPEC to capture exogenous changes in financial markets' expectations about future oil supply. In particular, I show how to identify the shock using the co-movement between the price of oil futures and stock prices, computed in a narrow window around OPEC production quota announcements. This avenue of research is important because shocks to expected oil supply can have powerful effects on global economic activity and might require stabilisation from monetary and fiscal authorities.

Previous literature used high-frequency surprises in oil futures around OPEC announcements to identify a shock to oil supply expectations. Using a model of information frictions in the oil market, I show that these surprises capture not only revisions in market expectations of oil supply, but also revisions in expected oil demand. Any identification strategy based on surprises in oil futures alone would then conflate shocks to expected oil supply and demand, and cannot be used to inform policy. The intuition is the following. Assume that markets do not perfectly observe the strength of the economy. Moreover, assume that OPEC possesses private information about future oil supply that is relevant to the pricing decisions of the markets. Financial markets use the price of oil as a public signal to forecast the state of the economy, where potential deviations from expected oil supply act as noise. Since markets do not know whether a movement in oil price is due to noise or to fundamentals, in equilibrium, they place less weight on the forecasted component than they would under full information. An OPEC announcement, by revealing information about oil supply conditions, reduces the noise in the public signal, causing a revision in the weight that markets place on the forecasted component. This is captured in the oil price revision and gives rise to the identification problem.

The co-movement of oil futures and stock prices around OPEC announcements allows us to disentangle the demand and supply components of the surprises, thus obtaining an exogenous measure of shifts in oil supply expectations. This co-movement is informative because a shock to oil demand expectations moves both oil futures and stock prices in the same direction, while a shock to oil supply expectations moves them in opposite directions. The sign of the co-movement enables a clean identification of the two shocks. I find that a negative shock to oil supply expectations has deep and long-lasting stagflationary effects on global economic conditions. Importantly, the contractionary effects of the shock are deeper and more immediate that previously reported. Moreover, I find novel evidence that the shock also propagates through financial channels.

As a consequence of the international role of the Dollar as the world's dominant currency in financial markets, trade invoicing, and foreign exchange reserves, US monetary policy propagates internationally not only via the classical trade and exchange rate channels, but also via powerful financial channels. The strength of US spillover effects is such that it raises questions about the validity of the Trilemma of international macro, whereby a country with flexible exchange rates, absent capital controls, can use monetary policy to stabilise domestic conditions.

In this paper, we measure the spillover effects of conventional US monetary policy on the global economy and on a set of 30 advanced and emerging markets, quantifying the relative importance of the various channels of transmission, and studying how spillovers differ across country characteristics such as national income, exchange rate regime, degree of capital controls, share of trade invoiced in dollar, and share of national assets and liabilities denominated in dollars.

We identify US monetary policy shocks by employing a high-frequency identification strategy that avoids contamination from central bank's information effects. Specifically, we use as an external instrument the surprises in federal funds futures computed around Federal Open Market Committee (FOMC) announcements, after regressing them on FOMC's private information, as measured by the Greenbook forecasts. Our dataset spans the period 1990:1 to 2018:9 and covers a large set of global macro-financial aggregates as well as a set of harmonised country-specific indicators for 30 advanced and emerging economies. It contains over 150,000 data-points. We estimate a battery of large vector autoregressions, using Bayesian techniques, to trace the dynamic responses to the shock of the global economy and of the countries in our sample.

We report three novel stylised facts. First, a US monetary policy tightening has large and qualitatively homogeneous contractionary effects on both real and nominal variables in advanced and emerging economies alike. Second, flexible exchange rates cannot fully insulate domestic economies from the shock, due to movements in domestic risk premia that limit central banks' ability to control the yield curve. Third, financial channels dominate demand and exchange rate channels in the transmission to real variables, while the transmission via oil and commodity prices determines nominal spillovers. The latter is an important novel channel not previously reported in the literature. Our results raise questions about the desirability of international policy coordination and which policy tools are best suited to limit the destabilising effects of US monetary policy.

High-frequency surprises in asset prices computed around monetary policy announcements are frequently used as an instrument to identify the macroeconomic effects of monetary policy. These surprises can be decomposed in various factors, capturing different aspects of monetary policy. We focus on two factors: one that captures the surprise in the policy action (the target) and another capturing the surprise in the policy communication (the path). The path factor summarises the forward path of monetary policy and can be used to identify unconventional monetary policy shocks.

Monetary policy announcements not only reveal the policy stance of the central bank, but also the bank's own assessment of economic conditions. The high-frequency surprises might capture both revisions in market expectations about the policy stance and revisions about the state of the economy. The latter informational component acts as a confounding factor when the surprises are used to identify monetary policy shocks, as it induces a positive co-movement between interest rates and economic conditions. Miranda-Agrippino and Ricco (2021), focussing on US conventional monetary policy, decomposes surprises in the policy action into macroeconomic information and policy shocks by directly controlling for the information set of the central bank, as represented by the staff's internal forecasts.

This paper extends the methodology proposed in Miranda-Agrippino and Ricco (2021) to the study of US unconventional monetary policy. First, we extend to 2015:12 the high-frequency instrument for the identification of conventional monetary policy shocks, therefore including the zero lower bound (ZLB) period that followed the Great Recession. Second, we show that the same methodology can be used to decompose the policy and informational components in the target and path factors. This allows us to estimate domestic effects of US unconventional monetary policy that are not contaminated by informational effects. Finally, we compare our results to those obtained with an alternative methodology to separate the policy and informational components in the surprises. We show that, although the identified shocks have broadly similar properties when studied in a VAR model, the instruments obtained are very different across methodologies. The nature of this difference requires further study.

This paper studies how disagreement among professional forecasters affects the transmission of monetary policy in the US. Forecasters disagree about forecasts for any variable and at all forecast horizons, and the extent of disagreement varies over time. Since anchoring agents' expectations about the future path of key macro indicators is one of the most important tools of monetary stabilisation policy, we expect the transmission of monetary policy to differ between times of high and low disagreement.

We estimate a Bayesian threshold VAR for the US economy, where the threshold variable is a proxy for disagreement among professional forecasters. This model endogenously selects periods of high and low disagreements. The proxy for disagreement is the first principal component of the time series of cross-sectional standard deviations extracted from the Blue Chip Financial Forecasts. The structural shocks are identified using high-frequency instruments for conventional monetary policy and information. The instrument for information shocks is the component of high-frequency surprises in the federal funds futures computed around monetary policy announcements that is explained by the private information of the central bank, as measured by the Greenbook forecasts.

Although we do not find significant differences in the response of the economy to a monetary policy shock when disagreement is high compared to low, we show that information effects are strong only when disagreement is high. For a monetary policy shock, responses are unambiguously contractionary in both high and low disagreement regimes, and marginally stronger for the high disagreement regime. On the contrary, we find a clear asymmetry across regimes for the responses to an information shock. In this case we observe stark expansionary effects for the high-disagreement regime and borderline contractionary effects for the low-disagreement regime.

Residential land, being a non-reproducible factor in fixed supply, becomes more valuable as the economy grows. This has been suggested as a potential cause of the widening inequality in income and wealth in many advanced economies. At the same time, agents often find it more profitable to invest in land rather than in productive capital, with a subsequent decline in the long-run rate of economic growth. This poses important questions about the fiscal treatment of land both in terms of policies aimed at reducing inequality and at improving long-term economic conditions. In this paper, we build a multiple-agents dynamic endogenous growth model that includes land and housing to study the long term effects of land taxation on economic growth and inequality. In the model there are two social classes: the capitalists, who invest both in productive assets and in housing (which is a combination of residential land and housing structures) but do not provide labour services, and the workers, who invest only in housing and decide on the level of effort they provide. Results from numerical simulations of the calibrated model show that a revenue-neutral shift from income taxation to land taxation increases long term economic growth and reduces wealth inequality.