What trigger jumps? Negative versus Positive jumps
Negative equity returns lead to higher future volatility than positive returns. This stylized fact is analyzed in low frequency framework (Bollerslev, Engle and Diebold, 1994; Andersen, Bollerslev, Christoffersen and Diebold, 2006) and then extended to a high frequency setting by (Bollerslev, Litvinova and Tauchen, 2006; Barnhoff-Nielsen, Kinnebrock, and Shephard, 2010; Visser, 2008; Chen and Ghysels (2011).
Barnhoff-Nielsen, Kinnebrock, and Shephard (2010) introduced the realized semivariance which allows decomposing the realized variance into a component that relates only to positive high frequency returns and a component that relates only to negative high frequency returns. This decomposition is carried out using the signed returns. Semivariance has been shown to outperform the variance and covariance in asset pricing. Others show that the covariance outperforms the variance in calculating the optimal hedge ratio (hedging a spot asset with a futures asset).
Accurate quantification the impact of a contemporaneous jump on future volatility is crucial as some test of jumps identification use a rolling window that accounts for future volatility (Lee-Mykland, 2008). The impact of jumps on future volatility depends critically upon the sign of the jumps; averaging across both positive and negative jumps the impact is around zero, while separating the jumps by sign leads to significant predictive gains. But what triggers jumps in financial data?
Empirical evidence supports the strong role of economic fundamentals in explaining the dynamics of bond and equity markets. However, fundamentals are found to carry far less explanatory power in explaining exchange rate movements. This phenomenon is termed the ‘news puzzle’ (Evans and Lyons, 2008). Due to the failure of fundamentals in explaining the movements of financial assets (exchange rate, bonds, stock index) several questions arise:
- To what extent do currency jumps and cojumps correspond with macro news?
- What is the direction of the relationship between financial assets price movements and news
- Are some types of macro news more influent?
- What is the speed of market response and jump resolution to news release?
- Do ‘good news’ cause positive jumps?
Earlier studies used low frequency data (daily, monthly) to analyze the role of monetary and macro news on currency markets (Klein et al, 1991; Evans and Lyons, 2008).
More recent studies employ high-frequency data to examine intra-day adjustment of exchange rates (Andersen et al, 2003; Laurent et al, 2011). The reaction of exchange rate (price) is quick while that of the variance is slow. The reaction is asymmetric where adverse news is found to have a larger impact (Andersen et al, 2003). Lahaye et al (2011) identify jumps and cojumps and relate them to macro news. They find that the likelihood for a news to cause currency jumps is lower than that to found in the bond and stock markets. The weak power of fundamentals in explaining currency jumps may be to the fact that researchers consider only US. Macroeconomic news and ignore the impact of domestic (or local or regional) news may have on currency markets.