http://hdl.handle.net/2451/25932 Search the Channel search http://archive.nyu.edu/simple-search http://hdl.handle.net/2451/26944 Title: Vector Multiplicative Error Models: Representation and Inference<br/><br/>Cipollini, Fabrizio; Engle, Robert F.; Gallo, Giampiero M.<br/><br/>Abstract: The Multiplicative Error Model introduced by Engle (2002) for positivevalued processes is specified as the product of a (conditionallyautoregressive) scale factor and an innovation process with positivesupport. In this paper we propose a multivariate extension of such amodel, by taking into consideration the possibility that the vectorinnovation process be contemporaneously correlated. The estimationprocedure is hindered by the lack of probability density functions formultivariate positive valued random variables. We suggest the use ofcopula functions and of estimating equations to jointly estimate theparameters of the scale factors and of the correlations of theinnovation processes. Empirical applications on volatility indicatorsare used to illustrate the gains over the equation by equation procedure. http://hdl.handle.net/2451/26940 Title: The Underlying Dynamics of Credit Correlations<br/><br/>Berd, Arthur; Engle, Robert; Voronov, Artem<br/><br/>Abstract: We propose a hybrid model of portfolio credit risk where the dynamics ofthe underlying latent variables is governed by a one factor GARCHprocess. The distinctive feature of such processes is that the long-termaggregate return distributions can substantially deviate from theasymptotic Gaussian limit for very long horizons. We introduce thenotion of correlation spectrum as a convenient tool for comparingportfolio credit loss generating models and pricing synthetic CDOtranches. Analyzing alternative specifications of the underlyingdynamics, we conclude that the asymmetric models with TARCH volatilityspecification are the preferred choice for generating significant andpersistent credit correlation skews. http://hdl.handle.net/2451/26933 Title: The Spline GARCH Model for Unconditional Volatility and its GlobalMacroeconomic Causes<br/><br/>Engle, Robert F.; Rangel, J. Gonzalo<br/><br/>Abstract: We introduce a new model to measure unconditional volatility, theSpline-GARCH. The model is applied to equity markets for 50 countriesfor up to 50 years of daily data. Macroeconomic determinants ofunconditional volatility are investigated. It is found that volatilityin macroeconomic factors such as gdp growth, inflation and short terminterest rates are important explanatory variables that increasevolatility. There is evidence that high inflation and low growth ofoutput are positive determinants. Volatility is higher for emergingmarkets and for markets with small numbers of listings but also forlarge economies. http://hdl.handle.net/2451/26934 Title: The Rise in Firm-Level Volatility: Causes and Consequences<br/><br/>Comin, Diego; Philippon, Thomas<br/><br/>Abstract: We study the increase in firm level risk and how it relates to thedecrease in aggregate risk [..] http://hdl.handle.net/2451/26938 Title: The Properties of Automatic Gets Modelling<br/><br/>Hendry, David F.; Krolzig, Martin<br/><br/>Abstract: After reviewing the simulation performance of general-to-specificautomatic regression model selection, as embodied in PcGets, we show howmodel selection can be non-distortionary: approximately unbiased&lsquo;selection estimates&rsquo; are derived, with reported standarderrors close to the sampling standard deviations of the estimated DGPparameters, and a near-unbiased goodness-of-fit measure. The handling oftheory-based restrictions, non-stationarity, and problems posed bycollinear data are considered. Finally, we consider how PcGets canhandle three &lsquo;intractable&rsquo; problems: more variables thanobservations in regression analysis; perfectly collinear regressors; andmodelling simultaneous equations without a priori restrictions. http://hdl.handle.net/2451/26927 Title: Stochastic Skew in Currency Options<br/><br/>Carr, Peter; Wu, Liuren<br/><br/>Abstract: We document the behavior of over-the-counter currency option pricesacross moneyness, maturity, and calendar time on two of the mostactively traded currency pairs over the past eight years. We find thatthe risk-neutral distribution of currency returns is relativelysymmetric on average. However, on any given date, the conditionalcurrency return distribution can show strong asymmetry. This asymmetryvaries greatly over time and often switch directions. We design andestimate a class of models that capture these unique features of thecurrency options prices and perform much better than traditionaljump-diffusion stochastic volatility models. http://hdl.handle.net/2451/26935 Title: Price Discovery in Tick Time<br/><br/>Frijnsy, Bart; Schotmanz, Peter<br/><br/>Abstract: In this paper we propose a tick time model for the quote setting processon Nasdaq using a time series of all quote updates by the most activedealers and ECNs (Electronic Communication Networks). The model includesduration effects in the volatility of the efficient price and in thecovariance of quote updates with the efficient price. As a measure ofprice discovery we define information shares in tick time. Whenaggregated to calendar time they provide an alternative for theHasbrouck (1995) information shares. In the empirical analysis wecompare quotes from two ECNs (Island and Instinet), and three wholesalemarket makers for 20 actively traded stocks with varying liquidity. Wefind that volatility does not increase with the duration between quoteupdates, and that longer quote durations lead to lower price discovery.In terms of price discovery we find that ECNs tend to dominate theliquid stocks, whereas market makers are important for less liquid stocks. http://hdl.handle.net/2451/26928 Title: Modelling Round-the-Clock Price Discovery for Cross-Listed Stocks usingState Space Methods<br/><br/>Menkveld, Albert J.; Koopman, Siem Jan; Lucas, Andr&eacute;<br/><br/>Abstract: U.S. trading in non-U.S. stocks has grown dramatically. Round-the-clock,these stocks trade in the home market, in the U.S. market and,potentially, in both markets simultaneously. We develop a generalmethodology based on a state space model to study 24-hour pricediscovery in a multiple markets setting. As opposed to the standardvariance ratio approach, this model deals naturally with (i)simultaneous quotes in an overlap, (ii) missing observations in anon-overlap, (iii) noise due to transitory microstructure effects, and(iv) contemporaneous correlation in returns due to market-wide factors.We provide an application of our model to Dutch-U.S. stocks. Ourfindings suggest a minor role for the NYSE in price discovery for Dutchshares, in spite of its non-trivial and growing market share. Theresults differ significantly from the variance ratio approach. http://hdl.handle.net/2451/26930 Title: Individual Investor Sentiment and Stock Returns<br/><br/>Kaniel, Ron; Saar, Gideon; Titman, Sheridan<br/><br/>Abstract: This paper investigates a unique dataset that enables us to determinethe aggregate buy and sell volume of individual investors for a largecross-section of NYSE stocks. We find that individuals trade as if theyare contrarians, and that the stocks that individuals buy exhibitpositive excess returns in the following month. These patterns areconsistent with the idea that risk-averse individuals provide liquidityto meet institutional demand for immediacy. We further examine therelation between individual investor sentiment and short-horizon(weekly) return reversals that have been documented in the literature.Our results reveal that individual investor sentiment predicts futurereturns, and that the information content of investor sentiment isdistinct from that of past returns or past volume. Furthermore, thetrading of individuals predicts weekly returns in the post-2000 era forstocks of all sizes, while past return seems to have lost its predictivepower for all but small stocks over the same time period. Lastly, wenote that there is very little cross-sectional correlation of ourindividual sentiment measure across the stocks in our sample. http://hdl.handle.net/2451/26942 Title: High Frequency Multiplicative Component GARCH<br/><br/>Engle, Robert F.; Sokalska, Magdalena E.; Chanda, Ananda<br/><br/>Abstract: This paper proposes a new way of modeling and forecasting intradayreturns. We decompose the volatility of high frequency asset returnsinto components that may be easily interpreted and estimated. Theconditional variance is expressed as a product of daily, diurnal andstochastic intraday volatility components. This model is applied to acomprehensive sample consisting of 10-minute returns on more than 2500US equities. We apply a number of different specifications. Apart frombuilding a new model, we obtain several interesting forecasting results.In particular, it turns out that forecasts obtained from the pooledcross section of companies seem to outperform the correspondingforecasts from company-by-company estimation. http://hdl.handle.net/2451/26926 Title: A Tale of Two Indices<br/><br/>Carr, Peter; Wu, Liuren<br/><br/>Abstract: In 1993, the Chicago Board of Options Exchange (CBOE) introduced theCOBE Volatility Index (VIX). This index has become the de factobenchmark for stock market volatility. On September 22, 2003, the CBOErevamped the definition and calculation of the VIX, and back-calculatedthe new VIX up to 1990 based on historical option prices. The CBOE isalso planning to launch futures and options on the new VIX. In thispaper, we describe the major differences between the old and the newVIXs, derive the theoretical underpinnings for the two indices, anddiscuss the practical motivation for the recent switch. We also studythe historical behaviors of the two indices.