Overnight Returns Research Articles

Asymmetry in the Price Impact of Trades in an High-Frequency Microstructure Model with Jumps

We estimate a general microstructure model of the transitory and permanent impact of order flow on stock prices. Jumps are detected in both the transaction price (observation equation) and fundamental value (state equation). The model's parameters and variances are updated in real time. Prices can be altered by both the size and direction of trades, and the effects of buy-initiated and sell-initiated trades are different. We estimate this model using tick-by-tick data for 12 large-capitalization stocks traded on the Euronext-Paris Bourse. We find that, at tick frequency, the overnight return, the intraday jumps, and the continuous innovations represent approximately 7%, 8.5%, and 36.7% of the total variation of stock returns. The microstructure model explains on average 47.7% of the total variation. Once jumps are filtered and parameters are estimated in real time, we also find that the price impact of trades is symmetric on average. However, the price of highly liquid stocks with a large proportion of sell-initiated orders tends to be more sensitive to buy trades, whereas the price of less liquid stocks with a large proportion of buy-initiated orders tends to be more sensitive to sell trades.

Overnight stock returns and realized volatility

The information flow in modern financial markets is continuous, but major stock exchanges are open for trading for only a limited number of hours. No consensus has yet emerged on how to deal with overnight returns when calculating and forecasting realized volatility in markets where trading does not take place 24 hours a day. Based on a recently introduced formal testing procedure, we find that for the S&P 500 index, a realized volatility estimator that optimally incorporates overnight information is more accurate in-sample. In contrast, estimators that do not incorporate overnight information are more accurate for individual stocks. We also show that accounting for overnight returns may affect the conclusions drawn in an out-of-sample horserace of forecasting models. Finally, there is considerably less variation in the selection of the best out-of-sample forecasting model when only the most accurate in-sample RV estimators are considered.

Predictability in Bond ETF Returns

We study the persistence of bond ETF premiums and discounts. Following a day of high or low premiums or discounts over NAV, ETFs tend to maintain a premium or discount for up to 30 days. Premiums and discounts also predict distinct patterns of returns after daily closing. Overnight returns are negative following a high premium, while ETFs with large discounts are followed by positive overnight returns. The large discount ETFs have substantially higher returns than high premium ETFs over the subsequent thirty days. We find that traditional liquidity measures, along with prior deviations from NAV, are significant in explaining a fund’s premiums/discounts. Finally, we examine a long-short portfolio strategy to exploit the observed deviations from NAV, and find it generates an alpha of .96% per month or about 11.5% per year.

Overnight Returns and Realized Volatility

The information flow in modern financial markets is continuous, but major stock exchanges are open for trading for only a limited number of hours. No consensus has emerged on how to deal with overnight returns when calculating and forecasting realized volatility in markets where trading does not take place 24 hours a day. Based on a recently introduced formal testing procedure, we find that for the S&P 500 index, a realized volatility estimator that optimally incorporates overnight information is more accurate in-sample. In contrast, estimators that do not incorporate overnight information are more accurate for individual stocks. We also show that accounting for overnight returns may a ffect the conclusions drawn in an out-of-sample horserace of forecasting models. Finally, there is considerably less variation in the selection of the best out-of-sample forecasting model when only the most accurate in-sample RV estimators are considered.

A leader of the world commodity futures markets in the making? The case of China's commodity futures

We use the daily data of 16 commodity futures contracts traded in China and corresponding foreign markets (the US, the UK, Japan, and Malaysia) to analyze the linkages between markets. Several findings are noteworthy. First, trading returns of foreign markets, such as the US, have significant impact on China's overnight (close-to-open) returns and vice-versa. Second, daytime (open-to-close) returns of many Chinese commodity futures contracts are not led by foreign daytime returns. Finally, the close-to-close returns analysis suggests that there are no significant lead-lag relationships between the Chinese and foreign markets. These results suggest that (1) the Chinese commodity futures markets are information-efficient, and (2) they are likely to be driven by local market dynamics occurring during the daytime trading session.

A semi-Markov model for price returns

We study the high frequency price dynamics of traded stocks by a model of returns using a semi-Markov approach. More precisely we assume that the intraday returns are described by a discrete time homogeneous semi-Markov process and the overnight returns are modeled by a Markov chain. Based on this assumptions we derived the equations for the first passage time distribution and the volatility autocorrelation function. Theoretical results have been compared with empirical findings from real data. In particular we analyzed high frequency data from the Italian stock market from 1 January 2007 until the end of December 2010. The semi-Markov hypothesis is also tested through a nonparametric test of hypothesis.

Paying Attention: Overnight Returns and the Hidden Cost of Buying at the Open

AbstractWe find a strong tendency for positive returns during the overnight period followed by reversals during the trading day. This behavior is driven by an opening price that is high relative to intraday prices. It is concentrated among stocks that have recently attracted the attention of retail investors, it is more pronounced for stocks that are difficult to value and costly to arbitrage, and it is greater during periods of high overall retail investor sentiment. The additional implicit transaction costs for retail traders who buy high-attention stocks near the open frequently exceed the effective half spread.

Information Transmission between Nighttime Trading and Daytime Trading: Evidence from KOSPI 200 Futures

Trading of KOSPI 200 futures on CME Globex platform, which was launched in November 2009, starts at 18:00 and closes at 05:00 in the next morning. This paper investigates the information transmission between daytime trading of KOSPI 200 futures on KRX and nighttime trading of KOSPI 200 futures on CME Globex by using GARCH (1, 1). The main findings of this paper are summarized as followings; Firstly, the statistically significant spillover effect from open-to-close returns of KOSPI 200 Futures on Globex to the overnight returns of KOSPI 200 futures on KRX is found but not to the daytime returns. Moreover, I find the spillover effects from daytime returns of KOSPI 200 futures on KRX to close-to-open return and open-to-close return of KOSPI 200 Futures on Globex. Meanwhile, this information transmission between two markets of common underlying asset shows more strongly statistical significance during highly liquid period. Secondly, daytime traders of KOSPI 200 futures on KRX recognize the price of KOSPI 200 futures on CME Globex as more valuable information than volatility of US stock market. Overall empirical evidence suggests that KOSPI 200 futures on CME Globex has leading role in price discovery process of daytime trading of KOSPI 200 futures on KRX despite of immaturity and low liquidity.

Management Insights

Management Insights

(Study on the Impact of Us Stock Market's Shock Due to Global Financial Crisis on Korean Financial Markets)

Korean Abstract: 본 연구는 글로벌 금융위기에 따른 미국 주식시장의 충격이 한국 금융시장 특히 주식 및 외환시장에 미친 영향을 분석하였다. 특히 글로벌 금융위기 기간과 금융위기 전후 기간을 비교하였으며, 전체 주식시장뿐만 아니라 금융업종(은행, 증권, 보험)에 미친 영향도 함께 분석하였다. 분석방법으로는 GARCH 모형과 GJR-GARCH 모형을 이용하여 대칭적 - 비대칭적 정보 전이효과를 함께 분석하였다. 분석 결과 첫째, 미국 주식시장의 주가변동이 한국 주식시장의 야간수익률에 유의한 양( )의 영향을, 주간수익률에는 금융위기 기간에서만 유의한 음(-)의 영향을 미치고 있다. 또한 미국 주식시장과 한국 주식시장 간에는 비대칭적 전이효과가 존재하는 것으로 나타나고 있다. 둘째, 미국 주식시장의 주가변동이 한국 외환시장 원달러환율의 야간수익률에 유의한 음(-)의 영향을 미치고 있으나, 주간수익률에는 유의한 영향이 없었다. 또한, 금융위기를 거치면서 미국 주식시장에서 한국 원달러환율의 야간수익률에 대한 전이효과는 증가한 것으로 나타났다. 셋째, 미국 주식시장의 주가변동이 개별 금융업종에 미친 영향을 분석한 결과 주식시장 전체에 미친 영향보다 더 큰 영향을 미친 것으로 나타났다. 넷째, 미국 주식시장 주가변동은 한국 주식시장에서의 외국인 순매매와 높은 양( )의 상관관계를 보이고 있으며, 이러한 관계는 금융위기를 거치면서 더욱 증가하였다. 또한 국내 주식시장에서의 외국인 순매매는 주가수익률과 유의한 양( )의 관계를 가지는 반면, 원달러환율과는 유의한 음(-)의 상관관계를 보이고 있으며, 금융위기를 거치면서 국내 금융시장에서 외국인의 역할이 더욱 증대되고 있는 것으로 나타났다. 결국 금번 글로벌 금융위기에서 미국 주식시장의 충격이 한국 금융시장에 커다란 영향을 미쳤으며, 그 과정에서 외국인 자금이탈이 커다란 역할을 했음을 알 수 있다. English Abstract: Study on the impact of US stock market’s shock due toWe empirically examined the spillover effects of US stock market on Korean stock and foreign exchange market during the global financial crisis. The time period of this study was divided into three periods that are global financial crisis period and pre- and post crisis period. Using the methodology of GARCH and GJR-GARCH, we examined symmetric and asymmetric information spillover effect. The major empirical findings are as follow: First, the spillover effect of US stock market on the overnight return of Korean stock market was statistically significant positive, and spillover effect on daytime return was significant negative only during the financial crisis period. Second, the spillover effect of US stock market on Won/dollar overnight return in Korea was statistically significant negative, but impact on its daytime return was not significant. Third, the impact of US stock market on Korean financial industries’ stock index(banks, securities, insurance) was arger than Korean total stock market index. Fourth, there was relatively high positive correlation between the change in US stock index and the foreign investors' net sales in Korean stock market. The foreign investors’ net sales in Korean stock market has positive correlation with the return of Korean stock market and has negative correlation with the return of Won/dollar exchange rate. These relations are stronger throughout the financial crisis. As a results, foreign investors has a important role in the contagion process of financial crisis in Korean financial market. global financial crisis on Korean financial markets.

Realized Volatility and Overnight Returns

No consensus has emerged on how to deal with overnight returns when calculating realized volatility in markets where trading does not take place 24 hours a day. This paper explores several common volatility applications, investigating how the chosen treatment of overnight returns affects the results. For example, the selection of the best volatility forecasting model depends on the way overnight returns are incorporated into realized volatility. The evidence favours weighted estimators over those that have been more commonly used in the existing literature. The definition of overnight returns is particularly challenging for the S&P 500 index, and we propose two alternative measures for its overnight return.​

Lead‐lag relationship, volatility asymmetry, and overreaction phenomenon

PurposeThe purpose of the paper is to investigate the price interrelationship between the Taiwanese and US financial markets.Design/methodology/approachThe trivariate GJR‐GARCH (1,1) model and event study were employed to investigate volatility asymmetry and overreaction phenomenon, respectively.FindingsThe empirical results show that return volatility reveals the asymmetric phenomenon, and the holding period returns on US index futures from the opening of the US index futures electronic trading to the opening of the Taiwanese stock market are an important reference for investors in the Taiwanese stock market. Additionally, the paper presents an overreaction of the Taiwan Stock Exchange Capitalization Weighted Stock Index to a drastic price rise of E‐min NASDAQ 100 Index futures at the opening of the Taiwanese stock market.Research limitations/implicationsThis paper deletes the observations arising from the different national holidays of the USA and Taiwan, to have the same number of observations in both markets, which might contaminate the empirical results.Practical implicationsInvestors in the Taiwanese stock market tend to pay more attention to the fluctuations in the share prices of high‐technological companies in the USA.Originality/valueMost of the previous studies regarding price transmission between the Taiwanese and US stock markets focused mainly on the Taiwanese market reactions to the overnight returns of the US market. This paper enlarges the current field by examining the lead‐lag relationship, the volatility asymmetry, and the overreaction phenomenon between the Taiwanese and US financial markets according to the most updated US stock index information.

Paying Attention: Overnight Returns and the Hidden Cost of Buying at the Open

Using 13 years of intraday data for U.S. stocks, we find a strong tendency for positive returns during the overnight period followed by reversals during the trading day. This behavior is driven by an opening price that is high relative to intraday prices. We find this temporary price inflation at the open is concentrated among stocks that have recently attracted the attention of retail investors, and these high attention stocks have high levels of net retail buying at the start of the trading day. In addition, we document that the sensitivity of opening prices to retail investor attention is more pronounced for stocks that are difficult to value and costly to arbitrage, and is greater during periods of high overall retail investor sentiment. The additional implicit transaction costs for retail traders who buy high attention stocks near the open frequently exceed the effective half spread.

Modelling Overnight and Daytime Returns Using a Multivariate Garch-Copula Model

We introduce a multivariate GARCH-Copula model to describe joint dynamics of overnight and daytime returns for multiple assets. The conditional mean and variance of individual overnight and daytime returns depend on their previous realizations through a variant of GARCH specification, and two Student's t copulas describe joint distributions of both returns respectively. We employ both constant and time-varying correlation matrices for the t copulas and with the time-varying case the dependence structure of both returns depends on their previous dependence structures through a DCC specification. We estimate the model by a two-step procedure, where marginal distributions are estimated in the first step and copulas in the second. We apply our model to overnight and daytime returns of SPDR ETFs of nine major sectors and briefly illustrate its use in risk management and asset allocation. Our empirical results show higher mean, lower variance, fatter tails and lower correlations for overnight returns than daytime returns. Daytime returns are significantly negatively correlated with previous overnight returns. Moreover, daytime returns depend on previous overnight returns in both conditional variance and correlation matrix (through a DCC specification). Most of our empirical findings are consistent with the asymmetric information argument in the market microstructure literature. With respect to econometric modelling, our results show a DCC specification for correlation matrices of t copulas significantly improves the fit of data and enables the model to account for time-varying dependence structure.

Volume, Intraday and Overnight Returns for Volatility Prediction: Profitability or Accuracy?

This article presents a comprehensive analysis of the relative ability of three information sets --- daily trading volume, intraday returns and overnight returns --- to predict equity volatility. We investigate the extent to which statistical accuracy of one-day-ahead forecasts translates into economic gains for technical traders. Various profitability criteria and utility-based switching fees indicate that the largest gains stem from combining historical daily returns with volume information. Using common statistical loss functions, the largest degree of predictive power is found instead in intraday returns. Our analysis thus reinforces the view that statistical significance does not have a direct mapping onto economic value. As a byproduct, we show that buying the stock when the forecasted volatility is extremely high appears largely profitable, suggesting a strong return-risk relationship in turbulent conditions.

How Information Transmits to Equity Markets: Earnings Announcements and Volatility

We explore the impact of earnings announcements on equity markets, using intraday price data for the DJIA stocks. We find on a daily basis, an abnormally high volatility only within one day following the overnight announcement. On an intraday basis, a striking volatility spike stands out during the overnight period, implying that the earnings news is fully reflected in the opening price. The continued high volatility during the first several minutes of trading seems to be driven by noise traders, given the one-minute returns uncorrelated with the overnight return. Using the standardized earnings surprise (SES), we find the overnight return highly dependent on good and bad news and no relationship between the 9:30-9:31 return and news content. The evidence along with the CAAR patterns surrounding the earnings announcement confirms that the equity market is completely awake and busy in reflecting the new information in the opening trading price.

A reduced form framework for modeling volatility of speculative prices based on realized variation measures

Building on realized variance and bipower variation measures constructed from high-frequency financial prices, we propose a simple reduced form framework for effectively incorporating intraday data into the modeling of daily return volatility. We decompose the total daily return variability into the continuous sample path variance, the variation arising from discontinuous jumps that occur during the trading day, as well as the overnight return variance. Our empirical results, based on long samples of high-frequency equity and bond futures returns, suggest that the dynamic dependencies in the daily continuous sample path variability are well described by an approximate long-memory HAR–GARCH model, while the overnight returns may be modeled by an augmented GARCH type structure. The dynamic dependencies in the non-parametrically identified significant jumps appear to be well described by the combination of an ACH model for the time-varying jump intensities coupled with a relatively simple log-linear structure for the jump sizes. Finally, we discuss how the resulting reduced form model structure for each of the three components may be used in the construction of out-of-sample forecasts for the total return volatility.

Intraday linkages between the Spanish and the US stock markets: evidence of an overreaction effect

This paper focuses on short-term information transmission between the US stock market, properly the DOW index, and the main Spanish stock index, IBEX-35, in its early and final hours. We follow the approaches of Lin, Engle and Ito (1994), Susmel and Engle (1994) and Baur and Jung (2005) who use a GARCH model to analyze the influence of the previous daytime and overnight returns and volatility of the DOW upon the overnight returns and daytime returns of the IBEX from Open-to-3:30 and from 3:30-to-Close. The results suggest that the Spanish stock market usually has a low price movement till Wall Street opens. Additionally, they indicate that the Spanish market reacts quickly to the news, basically in the first four minutes following the opening of the US market. Furthermore, we find the existence of an overreaction effect during the two hours before the closing of the Spanish market.

Market Madness? The Case of Mad Money

We use the popular television show Mad Money hosted by Jim Cramer to test theories of attention and limits to arbitrage. Stock recommendations on Mad Money constitute attention shocks to a large audience of individual traders. We find that stock recommendations lead to large overnight returns which subsequently reverse over the next few months. The spike-reversal pattern is strongest among small, illiquid stocks that are hard-to-arbitrage. Using daily Nielsen ratings as a direct measure of attention, we find the overnight return is strongest when high income viewership is high. We also find weak price effects among sell recommendations. Taken together, the evidence supports the retail attention hypothesis of Barber and Odean (2008) and illustrates the potential role of media in generating mispricing.

Modelling Overnight and Daytime Returns Using a Multivariate GARCH-Copula Model

We introduce a multivariate GARCH-Copula model to describe joint dynamics of overnight and daytime returns for multiple assets. The conditional mean and variance of individual overnight and daytime returns depend on their previous realizations through a variant of GARCH specification, and two Student’s t copulas describe joint distributions of both returns respectively. We employ both constant and time-varying correlation matrices for the t copulas and with the time-varying case the dependence structure of both returns depends on their previous dependence structures through a DCC specification. We estimate the model by a two-step procedure, where marginal distributions are estimated in the first step and copulas in the second. We apply our model to overnight and daytime returns of SPDR ETFs of nine major sectors and briefly illustrate its use in risk management and asset allocation. Our empirical results show higher mean, lower variance, fatter tails and lower correlations for overnight returns than daytime returns. Daytime returns are significantly negatively correlated with previous overnight returns. Moreover, daytime returns depend on previous overnight returns in both conditional variance and correlation matrix (through a DCC specification). Most of our empirical findings are consistent with the asymmetric information argument in the market microstructure literature. With respect to econometric modelling, our results show a DCC specification for correlation matrices of t copulas significantly improves the fit of data and enables the model to account for time-varying dependence structure.