Abstract The spillover index developed by Diebold and Yilmaz (Economic Journal, 2009, vol. 119, pp. 158-171) is widely used to measure connectedness in economics and finance. Abrupt increases in the spillover index are thought to result from major economic and financial events, but formal evidence of this relationship is scarce. We develop a bootstrap-based technique to evaluate the probability that the value of the spillover index changes following an exogenously defined event. We revisit the original dataset from Diebold and Yilmaz and obtain qualified support for their finding that the spillover index increases in a timely manner in the wake of the adverse shocks. Keywords: Spillover index, advserse shocks, influential events, bootstrap-after-bootstrap procedure. JEL Codes: C32, C58, G15. Full text coming soon!

Abstract We use a cycle dating algorithm to identify five commodity price cycles over the period January 1999 to May 2021, the peaks of which occur in November 2000, September 2005, July 2008, February 2012 and October 2018. In all five cases, we provide strong evidence of a change in interdependence among the commodity market and a large cross-section of global equity markets between the expansionary and contractionary phases of the cycles. We show that the probability of interdependence change is related to changes in the economic characteristics of each market, including in per capita GDP, the size of manufacturing value added and characteristics of the trade in oil. Keywords: Commodity price cycle, interdependence change tests, Logit regression. JEL Codes: C58, E32, G15. Full text coming soon!

Abstract We investigate the relationship between R&D expenditure and investment using a nonlinear autoregressive distributed lag (NARDL) model. We propose an analytically tractable two-step NARDL (2SNARDL) estimation framework, in which the parameters of the long-run relationship are estimated first using the fully-modified least squares estimator before the dynamic parameters are estimated by ordinary least squares in the next step. We show that the 2SNARDL estimator is consistent for the parameters of the NARDL model under mild regularity conditions. We derive the relevant limit distributions and develop Wald test statistics for the hypotheses of short-run and long-run parameter asymmetry. We apply the 2SNARDL procedure and show that R&D expenditure is asymmetrically associated with physical investment in the U.S. Keywords: R&D Expenditure; Physical Investment; Nonlinear Autoregressive Distributed Lag (NARDL) Model; Two-step Estimation; Wald Test Statistic. JEL Codes: C22, E22, O32. Full text

Abstract Monetary theory typically assumes that the pass-through from policy-controlled interest rates to longer term rates and yields is complete, rapid and symmetric. We investigate these assumptions by applying the Nonlinear ARDL (NARDL) model advanced by Shin, Yu and Greenwood-Nimmo (2013) to the analysis of interest rate pass-through in the U.S. We find five stylised findings as follows: (i) rate cuts are passed through more completely than rate hikes in the long-run; (ii) rate hikes are passed through more rapidly than rate cuts in the short-run; (iii) pass-through has become less complete during the Great Moderation; (iv) the speed of adjustment has increased during the Great Moderation; and (v) the 2003-5 decoupling 'conundrum' identified by Alan Greenspan is statistically indistinct from previous monetary contractions. Where pass-through exhibits such complexities, our findings suggest that demand management policies involving manipulation of the short-term interest rate may encounter significant difficulties in the absence of regulatory reform. Keywords: Nonlinear Autoregressive Distributed Lag (NARDL) Model, Asymmetric Interest Rate Pass-through, Great Moderation, Greenspan Decoupling Conundrum. JEL Codes: C22, E43, E52. SSRN: http://ssrn.com/abstract=1894621 Full text

Abstract We derive a new method of modelling the Taylor Rule in a system setting which expressly accounts for its combination of I(0) and I(1) series. Using a long sample of US data, our model provides modest support for an inertial Taylor-type rule. However, estimation across rolling windows indicates that the in ation and output preferences of the Fed have varied significantly through time, presumably reflecting the prevailing economic and political conditions, its chairmanship, and the composition of the FOMC. Our most significant finding is that the Taylor Principle was robustly upheld under Volcker, often upheld pre-Volcker but rarely observed post-Volcker. Keywords: System Estimation with Mixed I(0) and I(1) Variables, Long-Run Structural Modelling, Rolling Estimation, Taylor Rule. JEL Codes: C13, C51, E58, N10. SSRN: http://ssrn.com/abstract=1810643 Full text

Abstract We identify three general forms of asymmetry that may characterise a wide range of economic processes: reaction asymmetry, adjustment asymmetry and locational asymmetry. The first relates to the possibility that the long-run response of one variable to another may be regime sensitive. The second refers to the case in which the pattern of dynamic adjustment to long-run equilibrium may be state-contingent. The third relates to the notion that an economic relationship may depend upon which conditional quantile of the dependent variable a given observation belongs. Based on a synthesis of the nonlinear ARDL model developed by Shin, Yu and Greenwood-Nimmo (2009) and the quantile regression approach of Koenker and Bassett (1978), we develop a new empirical framework capable of coherently and simultaneously modelling these three asymmetries. The application of this model to US monetary policymaking over the period 1964q2-2008q2 reveals the following phenomena: (i.) the Fed responds more aggressively to positive than negative inflation gaps and also to negative than positive output gaps; (ii.) the Fed responds more rapidly to inflationary pressures than disinflationary pressures and to recessionary pressures than expansionary pressures; and (iii.) the degree of policy aggression is a monotonically increasing function of the conditional quantile of the interest rate. Keywords: Nonlinear ARDL Model, Quantile Regression, Dynamic Multipliers, Asymmetric Central Bank Preferences. JEL Codes: C22, C51, E58. Full text

Abstract We employ the cointegrating VAR framework with exogenous I(1) variables initiated by Pesaran et al. (2000), Pesaran and Shin (2002) and Shin (2009) to reexamine four popular monetary models of the exchange rate. Unlike existing studies, we carefully account for parameter instability and do not impose any arbitrary restrictions on the exchange rate-fundamental relationship (EFR) the exchange rate and macro fundamentals. Using the data for the GBP/USD bilateral exchange rate, we find that there exists a long-run relationship between the exchange rate and fundamentals. However, the results from our counterfactual exercises show that it takes at least a year or two for the EFR to converge to equilibrium after shocks. This finding explains why the exchange rate are often found to be disconnected from fundamentals in the short run. We also note that the adjustment speed of the EFR towards equilibrium following shocks is faster in the stickyprice models than it is in the flexible-price models. This underlies the breakdown of PPP in the short run as a key reason for the persistent adjustment of the EFR in the flexible-price models. Importantly, the monetary models in our study can explain between 40% and 60% of quarterly exchange rate movements and generate a better forecasting performance than that of the random walk in in-sample and out-of-sample forecasting exercises. This is a significant finding given the documented poor performance of macro exchange rate models. Keywords: Exchange rate, monetary models, cointegrating VAR, structural breaking, exchange rate forecasting. JEL Codes: C22, F31, G15. Fully revised version coming soon!

Abstract Through its many incarnations the Phillips curve has remained a fundamental cornerstone of the macroeconomic debate, a position that is assured by its role in informing the design and conduct of stabilisation policies. However, the evolution of the Phillips curve is an ongoing process and a number of recent studies have suggested that it may exhibit various non-linearities. This observation has profound implications for monetary policy as nonlinearity in the underlying Phillips curve will result in nonlinear monetary policy even when policymakers' preferences are strictly quadratic. We estimate a threshold-ARDL model of the Canadian Phillips curve and find overwhelming support for nonlinearity in relation to the rate of change of unemployment. Our results suggest that the output-unemployment tradeoff is steep when unemployment is changing rapidly but insignificant when unemployment is changing slowly. Hence, we conclude that there is substantial scope for the Bank of Canada to pursue opportunistic expansionary policies when the unemployment rate is in the relatively stable regime. Keywords: 3 Regime Nonlinear Phillips Curve, Asymmetric ARDL Model, Threshold Modelling, Opportunistic Monetary Policy, Behavioural Economics. JEL Codes: C22, E52, E61. Full text

Abstract The ‘threat’ to monetary policy posed by e–money was a high profile issue at the turn of the millennium. This paper argues that the debate surrounding this issue, although largely resolved, has diverted attention from a number of important and less fanciful issues associated with the electronification of the payments system. These include the revival of the spectre of old–fashioned bank–runs, the threats to systemic security posed by unregulated offshore issuers and issues surrounding anonymity and the right to privacy. The extent to which these issues can be considered important depends upon the degree of usage of e–money products. A comparative analysis of the Euro Area and Singapore reveals that e–money usage remains minimal in most European countries, certainly in terms of total transaction value. Furthermore, a range of forecasting exercises indicate modest growth potential in the medium–term. These results lead to either of two conclusions. The reactionist view would hold that e–money usage is not widespread and that the current regulatory framework is, therefore, sufficient. The proactive view, by contrast, would ask whether the regulatory framework may be responsible for the slow uptake of the new technology and whether it could be revised to provide stronger incentives for all parties concerned. Keywords: Electronic Money (E–Money), Monetary Policy, Forecasting, Cross–Country Analysis. JEL Codes: C53, E52, G20. Full text