eThesis - Kauppakorkeakoulun sähköiset gradut

» hse_ethesis_13829.pdf   koko: 17 MB (17069201)

Objective of the study:

In this paper, I study the effects of funding constraints with respect to the uncovered interest parity (UIP) violations, i.e. the excess returns emerging from the traditional carry strategies. More specifically, I examine the impact of the realized firm-level idiosyncratic aggregate average uncertainty in the United States (U.S.) economy as well as the realized firm-level idiosyncratic average uncertainty in the U.S. financial sector on carry trade excess returns. Moreover, I conduct a sub-period analysis with respect to the surge in the amount of speculative capital since early 2000 and the financialization in order to understand how sensitive the speculative community is to the unexpected changes in systemic risk. Finally, in addition to contributing to existing research and opening new avenues for future research, I re-examine and confirm existing literature on the uncovered interest parity (UIP) violations, the role of learning in the forward premium puzzle, and the linkage between currency carry trades and currency crash risk.

Date and methodology:

The data set consists of daily spot and forward rates for 9 currencies with respect to the USD dollar from January 1996 to February 2014. In addition, I collect daily data for all listed U.S. stocks and their daily returns from CRSP for the same period in order to construct the idiosyncratic firm-level risk metrics. To test the hypotheses, I estimate several multivariate OLS regressions with varying specifications and perform numerous robustness checks.

Findings of the study:

The multivariate model based upon the idiosyncratic financial sector uncertainty is statistically significant and explains 14.3% of the excess return variability of the High-minus-Low (HML3) portfolios that comprise a long position in the top three currencies and a short position in the bottom three currencies. The funding model consists of two independent explanatory variables that are both related to the realized idiosyncratic average firm-level uncertainty in the U.S. financial sector: a contemporaneous change and a 6-month moving average.

Moreover, a multivariate funding model based upon the conventional TED spread, a typical measure of funding constraints, and the idiosyncratic banking sector uncertainty explains 21.4% of the High-minus-Low (HML3) excess return variability. The marginal contribution of the latter is statistically significant once I take into account the loss in degrees of freedom.

Additionally, a single explanatory variable that proxies the effects of unexpected funding shocks, a normalized specification, explains 11.4% of the High-minus-Low (HML3) excess return variability for the full period from 1996 to 2013. In comparison, the contemporaneous TED spread explains 11.0% of the variation for the same period. The correlation coefficient among the two is 0.56. Furthermore, the explanatory power of the normalized specification increases from 11.4% to 22.0% in a multi-dimensional setting.