with Mike Seiferling (UCL), Abeer ElBahrawy (City, University London), and Keith Chan (University of Cambridge)
In the aftermath of the financial crisis, the shortcomings of existing financial systems became widely criticised, leading to an unprecedented wave of interest in new ways of efficiently executing economic transactions while ensuring high levels of transparency and accountability. The popularity of cryptocurrencies and their potential for “disrupting” and improving traditional financial systems have led to an ever-expanding list of media commentaries, research papers, and policy reports. With over 2,000 in existence, cryptocurrencies have become progressively embraced by speculative investors and growing market caps, but have yet to be adopted by the wider public as a viable form of money due to practical technical challenges along with a lack of trust in issuing authorities and understanding of how to use them. This research programme provides a more comprehensive overview of how cryptocurrency could be used for the betterment of society, how they currently function and how the general public use, understand and trust cryptocurrencies across Europe and the Americas.
Research in Progress
Teaching An Old Horse Some New Tricks: Soft power and exchange rate prediction.
with Serhan Cevik (IMF)
It is well known that exchange rate fluctuations are very difficult to predict using traditional economic models. In fact, the consistent outperformance of the random walk over economic models in forecasting exchange rates has become known as the Meese and Rogoff Puzzle. In reviews of recent advances in the empirical international finance literature, Rossi (2013) and Cheung et al. (2019) have concluded that although some models can perform better than the random walk in specific settings, the overall results from recent models are not very successful. In summarizing the evidence from their exhaustive analysis, Cheung et al. (2019) conclude that the “models that have become popular in the last fifteen years or so might not be much better than the older ones”. Given the poor performance of standard models — based exclusively on macro-financial variables — in predicting exchange rate fluctuations, we introduce a neglected set of “soft power” factors capturing a country’s demographic, institutional, political, and social underpinnings to shed some light on the components that may be missing in traditional studies of exchange rate prediction.
Currency GEOVOL and the PPP Puzzle.
with Susana Campos-Martins (Oxford)
Most of the discussion about the Purchasing Power Parity (PPP) Puzzle of Rogoff (1996) has pertained to the reversion speed of deviations from PPP. Much less attention, however, has been given to the other component of the puzzle: the high short-term volatilities of real exchange rates. In this paper, we provide a framework that is capable of explaining the econometric sources of this short-term volatility. First, we study the drivers of real exchange rate volatilities using a Cross-Sectionally Augmented Autoregressive Distributed Lag (CR-ARDL) panel framework and the conditional covariance matrices of the system with nominal exchange rates and price differentials. This analysis indicates that, for both emerging and developed markets, common factors are the main drivers of volatility. With this result in hand, we propose a novel econometric framework (based on Campos-Martins and Engle (Forthcoming)) that explains the sources of this short-term volatility as geopolitical shocks (GEOVOL).
Pairs Trading is commonly referred to as the most straightforward example of a market neutral trading strategy. In this study, we analyse how market news and idiosyncratic news affect the profitability of this statistical arbitrage trading technique. We propose a conditional covariance framework based on Kroner and Ng (1998) extension of the BEKK Generalized Autoregressive Conditional Heteroskedasticity (GARCH) model to analyse the dependence of second moments between different portfolios of pairs and the returns of the market. In contradiction to what is generally assumed about the market neutrality of this strategy, our preliminary results indicate the existence of significant spillovers from market news to different portfolios of pairs. In fact, market shocks seem to play a significant role in explaining conditional second moments of the returns of these portfolios, particularly in more turbulent periods. Not only do market shocks have a significant impact on the volatility of different portfolios, but they also appear to influence the correlation of this strategy with the market. In a second step of the study, we analyse the contribution of both idiosyncratic and market components to pairs volatility over time in an asynchronous panel of pairs. This analysis shows that the volatility of the pairs strategy has become more dependent on idiosyncratic rather than market shocks. In this sense, although Pairs Trading cannot be said to be market neutral from a second moments perspective by looking at the full sample, the strategy has certainly become more market neutral as markets have evolved over time.
Data for replication:
The objective of this paper is to study the relationship between high-frequency trading and volatility. Specifically, how high-frequency traders differ from other types of traders when both total and efficient components of volatility are considered. By using a vector autoregression (VAR) framework to model information in equity markets and a dataset that includes all trades for stocks traded on NASDAQ between the 22nd of February 2010 and the 26th of February 2010, this study concludes that high-frequency traders are, from a volatility perspective, more efficient than non-high-frequency traders. Although trades where high-frequency traders are the liquidity seeking side represent more than half the number of trades and the total volume, these trades are responsible for less than one third of the total volatility. More importantly, not only do high-frequency traders generate less than half the total volatility of the other types of traders, but they contribute ten times more than non-high-frequency traders to price discovery, that is, to efficient price volatility.