Trade imbalance network and currency risk premia




In this paper, I propose a new approach to thoroughly explore the link between network centrality and currency risk premia.  My paper will contribute to the understanding of the role of international trade network being a channel for risk transmission and its importance for asset prices.

According to GM2015, it is the trade imbalance, deficit of a country’s current account that affects the country’s currency depreciation in face of a large shock. We therefore propose a new trade imbalance network to capture the importance of each edge, based on the deficit of balance of payment). 

Previous researchfocuses on testing uncovered interest rate parity (UIP) and explaining the forward premium puzzle (Lustig &Verdelhan, 2007),  however, simply ignores how economic sources, i.e., trade and trade imbalance affect the foreign exchange returns.  In particular, the existing literature ignores that a country’s fundamentals depend not only on the quality of the fundamentals of its direct trading partners but also indirectly on the quality of those trading partners’ trading partners. Richmond (2018) is the only paper studying how trade network centrality help explain the currency risk premium. His focus is the importance of a country in the global network of total trade. However, it is the trade deficit causing the currency depreciated in face of a large shock (GM2015).


Therefore, instead of using the network on the total trade, I build a network on trade imbalance, and the link between countries is the deficit balance of trade between them.  An edge directs from one country to another if the latter country has trade deficit with the former one.  My network includes over 66 countries and varies annually from 2001 to 2017. The countries are selected following Corte,Riddiough& Sarno (2016)who test the GM2015 theory. If a country has trade deficit with many trading partners, which also have trade deficit with their trading partners, this country is central  in my network of trade imbalance.


I then sort currencies into portfolios based on the calculated centrality measure. Sorting into portfolios reduce idiosyncratic currency risks (Lustig & Verdelhan, 2007). Using the portfolio sorts, I construct a central-minus-peripheral (CMP) portfolio by buying the currencies of the most central countries in the network and selling those of the most peripheral ones. We test also whether the CMP portfolio is a risk factor that explains the returns to currency carry trade strategy. This project contributes to the existing literature by introducing a trade imbalance network approach. We also propose a new risk factor to explain the currency risk premia.

International Trade and sovereign yieldrisks (with Caihong Xu & Xiaoxia Ye)




This project proposes new approaches to thoroughly investigate the impact of the international trade on international sovereign yield. In particular, we study how trade linkages could provide a channel through which sovereign yield risks spread between the importing and exporting countries. We establish trade networks based on the total trade and the composition of international trade: commodity goods v.s. high technology goods, then use it to investigate how interest rate risk embodied in the sovereign yield curves can be transmitted among exporting and importing countries.

The futures market microstructure invariance (With Lars Norden & Caihong Xu)




How can we reconcile the activities of the different types of futures traders into one theoretical framework? Kyle and Obizhaeva (2016) propose that the market microstructure invariance (MMI) theory does the job. It stipulates that the distributions of risk transfers and transaction costs are constant over trading time. Andersen, Bondarenko, Kyle & Obizhaeva (2018) translate the MMI into an intraday trading invariance (ITI) theory, in which they assume that the volatility per trade is proportional to expected trade size. The intuition is that traders trade more often, and in smaller lots, when the volatility is high. We analyze how well these theories stand up in a futures hedging demand-supply framework, with, on the one hand, long-term investors with hedging needs, and, on the other hand, HFTs and other liquidity suppliers with intraday trading horizons. In particular, we study the relationship between volatility and trade size at times when futures hedging pressure is large. We expect that an increased hedging pressure will create higher transaction costs than predicted by the MMI and the ITI.

Long- and Short-Run Volatility Spillover in European Stock Markets (with Hossein Asgharian & charlotte Christiansen)


In this project, we plan to investigate the volatility spillover from the global stock market (US) to local European stock markets, namely France, Germany, Italy, Netherlands, Spain, Sweden, and the UK. We will propose a new framework where we combine the mixed data sampling (MIDAS) volatility approach of Engle, Ghysels & Sohn (2013) with the volatility spillover model of Bekaert & Harvey (1997). This will enable us to examine both short-run and long-run volatility spillover. To our knowledge, we will be the first to investigate the short-run and long-run volatility spillover. We will divide the unexpected return to the local country into long-run and short-run global and local effects. From this, we will measure the local country’s variance ratio related the long-run and short-run global and local effects. We will examine the potential effects from the recent financial crisis and the European sovereign debt crisis and we will study whether volatility spillover is asymmetric between positive and negative shocks.