We first propose a reduced-form model in discrete time for S&P 500 volatility showing that the forecasting performance can be significantly improved by introducing persistent leverage effect with long-range dependence similar to of itself. also find strongly significant positive impact lagged jumps on volatility, which however is absorbed more quickly. then estimate continuous-time stochastic models are able reproduce statistical features captured discrete-time model. show single-factor...

We introduce a novel stochastic quantity, named excess idle time (EXIT), measuring the extent of sluggishness in observed high-frequency financial prices.Using limit theory robust to market microstructure noise, we provide econometric support for fact that transaction prices are, coherently with liquidity and asymmetric information theories price determination, generally stickier than implied by ubiquitous semimartingale assumptions (and its noise-contaminated counterpart).EXIT provides,...

Asset prices can be stale. We define price staleness as a lack of adjustments yielding zero returns (i.e., zeros). The term idleness (respectively, near idleness) is, instead, used to when trading activity is absent close absent). Using statistical and pricing metrics, we show that zeros are genuine economic phenomenon linked the dynamics volume and, therefore, liquidity. Zeros are, in general, not result institutional features, like discreteness. In essence, spells or stylized facts...

Epps [17] reported empirical evidence that stock correlations decrease when sampling frequency increases. This phenomenon, named effect, has been observed in several markets. In this paper, the dynamics underlying effect are investigated. Using Monte Carlo simulations and analysis of high foreign exchange rate price data, it is shown can largely be explained by two factors: non-synchronicity observations existing lead-lag relationship between asset prices. order to compute co-volatilities,...

We present a methodology for detecting flash crashes by identifying short-term V- shaped price reversals. Our approach, based on drift burst test statistics, aligns with the SEC's forensic definition of market access rule violations, highlighting its potential as surveillance tool. Flash have become more frequent over past decade and are typically accompanied high volumes, volatility, an increase in odd-lot trades. They likely to occur following periods elevated impact, low heightened...

In this paper, new fully nonparametric estimators of the diffusion coefficient continuous time models are introduced. The based on Fourier analysis state variable trajectory observed and estimation quadratic variation between observations by means realized volatility. proposed shown to be consistent asymptotically normally distributed. Moreover, estimator can iterated get a estimate in bivariate model which one is volatility other. unbiased small samples using Monte Carlo simulations used...

Download This Paper Open PDF in Browser Add to My Library Share: Permalink Using these links will ensure access this page indefinitely Copy URL DOI

This study reconsiders the role of jumps for volatility forecasting by showing that have a positive and mostly significant impact on future volatility. result becomes apparent once is correctly separated into its continuous discontinuous component. To this purpose, we introduce concept threshold bipower variation, which based joint use variation estimation. We show generalization (threshold multipower variation) admits feasible central limit theorem in presence jumps, not attainable standard...

We have measured the hyperfine field at $^{100}\mathrm{Rh}$ impurity nuclei in a ferromagnetic Ni host region just below Curie temperature, using time-differential perturbed $\ensuremath{\gamma}\ensuremath{\gamma}$ angular correlations. obtain single-valued critical exponent of $\ensuremath{\beta}=0.385\ifmmode\pm\else\textpm\fi{}0.005$.

Even moderate amounts of zero returns in financial data, associated with stale prices, are heavily detrimental for reliable jump inference. We harness staleness-robust estimators to reappraise the statistical features jumps markets. find that much less frequent and contributing price variation than what found by empirical literature so far. In particular, finding volatility is driven a pure process actually shown be an artifact due staleness.