Pierre Giot

The Logarithmic Acd Model: An Application to the Bid-Ask Quote Process of Three NYSE Stocks

This paper introduces the logarithmic autoregressive conditional duration (Log-ACD) model and compares it with the ACD model of Engle and Russell [1998]. The logarithmic version allows to introduce in the model additional variables without sign restrictions on their coefficients. We apply the Log-ACD model to price durations relative to the bid-ask quote process of three securities listed on the New York Stock Exchange, and we investigate the influence of some characteristics of the trade process (trading intensity, average volume per trade and average spread) on the bid-ask quote process.

Relationships Between Implied Volatility Indexes and Stock Index Returns

There is a negative and statistically significant relationship between the returns of the S&P 100 and the Nasdaq 100 stock indexes and their corresponding implied volatility indexes, VIX and VXN. For the S&P 100, the relationship is asymmetric, as negative stock index returns are more associated than positive returns with greater changes in VIX. VIX changes when negative stock index returns are observed are greater in low–volatility periods. For the Nasdaq 100, the asymmetric effect is rather weak, but the VXN and underlying index co–movement is also somewhat muted in high–volatility trading environments. There is some evidence that positive forward–looking returns are to be expected for long positions triggered by extremely high levels of the implied volatility indexes.

Market risk in commodity markets: a VaR approach

We put forward Value-at-Risk models relevant for commodity traders who have long and short trading positions in commodity markets. In a five-year out-of-sample study on aluminium, copper, nickel, Brent crude oil and WTI crude oil daily cash prices and cocoa nearby futures contracts, we assess the performance of the RiskMetrics, skewed Student APARCH and skewed student ARCH models. While the skewed Student APARCH model performs best in all cases, the skewed Student ARCH model delivers good results and its estimation does not require non-linear optimization procedures. As such this new model could be relatively easily integrated in a spreadsheet-like environment and used by market practitioners.

IPOs, Trade Sales and Liquidations: Modelling Venture Capital Exits Using Survival Analysis

Using a detailed sample made up of more than 20,000 investment rounds, we analyze the time to ‘IPO’, ‘trade sale’ and ‘liquidation’ for about 6,000 venture backed firms. We model these exit times using competing risks models. Biotech and internet firms have the fastest IPO exits. Internet firms are also the fastest to liquidate, while biotech firms are however the slowest. The conditional probability for IPOs are clearly non-monotonous with respect to time. As time flows, venture capital-backed firms first exhibit an increased likelihood of exiting to an IPO. However, after having reached a plateau, investments that have not yet exited have fewer and fewer possibilities of IPO exits as time increases. The bubble period from 1998 to 2000 was an ‘easy money’ period where venture capitalists gave much more money to firms, many of which did not offer outstanding growth potential as they tended to liquidate much faster than in normal times.

News Announcements, Market Activity and Volatility in the Euro/Dollar Foreign Exchange Market

This paper deals with the impact of nine categories of scheduled and unscheduled news announcements on the Euro/Dollar return volatility. We highlight and analyze the pre-announcement, contemporaneous and postannouncement reactions. Using high-frequency intraday data and within the framework of ARCH-type and realized volatility models, we show that volatility increases in the pre-announcement periods, particularly before scheduled events. Market activity also significantly impacts return volatility as expected by the theoretical literature on order flow.

Econometric Modelling of Stock Market Intraday Activity

Acknowledgments. Introduction. 1. Market Microstructure, Trading Mechanisms and Exchanges. 2. NYSE TAQ Database and Financial Durations. 3. Intraday Duration Models. 4. Empirical Results and Extensions. 5. Intraday Volatility and Value-At-Risk. About the Authors. Index.

Market risk models for intraday data

Abstract In this paper, market risk at an intraday time horizon is quantified using normal GARCH, Student GARCH, RiskMetrics and high-frequency duration (log-ACD) models set in the framework of the conditional VaR methodology. Because of the small time horizon of the intraday returns (15 and 30 minute returns in this paper), an evaluation of intraday market risk can be useful to market participants (traders, market makers) involved in frequent trading. As expected, the volatility features an important intraday seasonality, which must be removed prior to using the market risk models. The four models are applied to intraday returns data for three stocks traded on the New York Stock Exchange and it is shown that the Student GARCH model performs best. The use of price durations as a measure of risk on time is commented upon.

Time transformations, intraday data and volatility models

In this paper, we focus on the trade and quote data for the IBM stock traded at the NYSE.We present two different frameworks for analyzing this dataset. First, using regularly sampled observations, we characterize the intraday volatility of the mid-point of the bid-ask quotes by estimating GARCH and EGARCH models, with intraday seasonalitybeing accounted for. We also highlight the impact of characteristics of the trade process (traded volume, number of trades and average volume per trade) on the volatility specifications. Secondly, we deal directly with the irregularly spaced data. We review two time transformations that allowa thinning of the original dataset such that new durations are defined. The newly defined price and volume durations are characterized and the performance of the Log-ACD model for modelling these durations is assessed. Moreover, price durations allowan easy computation of intraday volatility and this method compares favorablyto ARCH estimations.

Implied Volatility Indexes and Daily Value at Risk Models

Value at risk, though flawed as a concept, has become one of the most common ways to summarize risk exposure for an investment position. Computing VaR requires the volatility of the underlying asset as an input. The two general approaches are to use historical volatility or some other estimator based on past returns, such as GARCH; or, alternatively, to compute implied volatility from observed market option prices. The Chicago Board Options Exchange computes and publishes indexes of implied volatilities from options on the stocks in the S&P 500 (VIX) and the Nasdaq index (VXN), respectively. In this article, Giot compares the performance of volatility forecasts drawn from historical returns and from option implied volatilities, using metrics that are appropriate for VaR estimates, such as the empirical failure rate relative to the target VaR percentage, independence, and conditional coverage. Implied volatility is found to perform relatively well, but combining GARCH and implied volatility often improves on the results from either one alone.

The Moments of Log-ACD Models

We provide existence conditions and analytical expressions of the moments of logarithmic autoregressive conditional duration (Log-ACD) models. We focus on the dispersion index and the autocorrelation function and compare them with those of ACD (Engle and Russell 1998) and SCD models. Using duration data for several stocks traded on the New York Stock Exchange, we compare the models in terms of their ability at fitting some stylized facts.