High-frequency Data Set Research Articles

Panel Data and Experimental Design

How should researchers design experiments to detect treatment effects with panel data? In this paper, we derive analytical expressions for the variance of panel estimators under non-i.i.d. error structures, which inform power calculations in panel data settings. Using Monte Carlo simulation, we demonstrate that, with correlated errors, traditional methods for experimental design result in experiments that are incorrectly powered with proper inference. Failing to account for serial correlation yields overpowered experiments in short panels and under-powered experiments in long panels. Using both data from a randomized experiment in China and a high-frequency data set of U.S. electricity consumption, we show that these results hold in real-world settings. Our theoretical results enable us to achieve correctly powered experiments in both simulated and real data. This paper provides researchers with the tools to design well-powered experiments in panel data settings.

An Introduction to Applications of Wavelet Benchmarking with Seasonal Adjustment

Summary Before adjustment, low and high frequency data sets from national accounts are frequently inconsistent. Benchmarking is the procedure used by economic agencies to make such data sets consistent. It typically involves adjusting the high frequency time series (e.g. quarterly data) so that they become consistent with the lower frequency version (e.g. annual data). Various methods have been developed to approach this problem of inconsistency between data sets. The paper introduces a new statistical procedure, namely wavelet benchmarking. Wavelet properties allow high and low frequency processes to be jointly analysed and we show that benchmarking can be formulated and approached succinctly in the wavelet domain. Furthermore the time and frequency localization properties of wavelets are ideal for handling more complicated benchmarking problems. The versatility of the procedure is demonstrated by using simulation studies where we provide evidence showing that it substantially outperforms currently used methods. Finally, we apply this novel method of wavelet benchmarking to official data from the UK's Office for National Statistics.

Karst aquifer characterization using storm event analysis for Black Dragon springshed, Beijing, China

Beijing is a megacity suffering from acute water shortage. The characterization of karst aquifers in Beijing's outlying districts is critical to inform sustainable policies and practices in groundwater development. We captured spring hydrogeochemical dynamics during an intense storm event (721 Storm) by installing a data logger in the weir, which presents an opportunity for defining and characterizing the recharging sources of discharge from Black Dragon (BD) springshed. Multivariate techniques (including PCA, 2D-HCA heat map, and piper diagram) were used to characterize the waters and investigate their geochemical evolution. The results in combination with hydrological evidence suggest BD Spring was dominated by interaction with dolomite host limestone rock during dry season, whereas it was fed by a second water supply (dolomite aquifer) during storm events. The infiltration coefficients were calculated based on the water balance model, and it was found that the 721 Storm has resulted in less infiltration due to the high rainfall intensity. We presented a novel methodology, combining high-frequency data sets and routine manual flow measurements with traditional geochemical sampling techniques. This methodology appears to be very efficient at characterizing a karst aquifer and seems to be potentially applicable to other karst aquifers.

Continuous High-Frequency Measurements of the Drilling Process Provide New Insights Into Drilling-System Response and Transitions Between Vibration Modes

Summary The oil-and-gas industry has become increasingly interested in drilling dynamics and vibration as causes of drilling inefficiency and reduced drilling performance. Generally, drillstring vibration is measured with shock-and-vibration sensors installed in measurement-while-drilling (MWD) tools, logging-while-drilling (LWD) tools, and rotary steerable systems (RSS). Although these tools provide valuable real-time and recorded-mode information on the dynamic conditions, they are not generally designed to capture continuous high-frequency (HF) mechanics and dynamics data, and burst data may miss important information about the evolution of the system response and state. A downhole mechanics measurement tool has been developed that makes a comprehensive suite of measurements of the drilling process, including forces, accelerations, rotational speed, pressures, and temperatures. In addition to providing information in real time, the tool has the capability to capture long durations of continuous data at frequencies between 50 and 2,000 Hz. The recorded-mode information obtained has provided significant insight into the response of the drilling system to starting rotation; drilling procedures and parameter modifications; and exposure to excitation from sources including, but not limited to, rig heave, bottomhole-assembly (BHA) component imbalance, and bit/rock interaction. A wide range of occurrences has been captured in which the drilling system switches from a dominant vibration mode, typically torsional (downhole rotation-velocity oscillations or stick/slip) into a different mode, such as axial (bit bounce) or lateral (whirl). Transitions between different types of whirl have also been recorded. Several cases were studied to investigate the evolution and response of drilling-system behavior on the basis of in-depth interpretation of relatively long durations (minutes to hours) of HF data sets in the operational context. The findings verify the value of using continuous HF vibration data to understand the drilling system and to increase drilling performance.

Roughing up beta: Continuous versus discontinuous betas and the cross section of expected stock returns

We investigate how individual equity prices respond to continuous and jumpy market price moves and how these different market price risks, or betas, are priced in the cross section of expected stock returns. Based on a novel high-frequency data set of almost 1,000 stocks over two decades, we find that the two rough betas associated with intraday discontinuous and overnight returns entail significant risk premiums, while the intraday continuous beta does not. These higher risk premiums for the discontinuous and overnight market betas remain significant after controlling for a long list of other firm characteristics and explanatory variables.

Determining the Integrated Volatility via Limit Order Books with Multiple Records

The integrated volatility plays an important role in risk management and portfolio selection, the estimation methods regarding to the quantity have been widely investigated, either under low frequency data or high frequency data, or a combination of both. In this paper, we propose a measure for the integrated volatility via limit order books data with multiple records. The estimator is valid under mild conditions and it is easy to implement. The finite sample performance of proposed estimator has been verified by simulation studies and we apply the method to some real high frequency data-sets as well.

Implied volatility smirk and future stock returns: evidence from the German market

Purpose – The purpose of this paper is to investigate the relationship between option’s implied volatility smirk (IVS) and excess returns in the Germany’s leading stock index Deutscher-Aktien Index (DAX) 30. Design/methodology/approach – The study defines the IVS as the difference in implied volatility derived from out-of-the-money put options and at-the-money call options. This study employs the ordinary least square regression with Newey-West correction to analyse the relationship between IVS and excess DAX 30 index returns in Germany. Findings – The authors find that the German market adjusts information in an efficient way. Consequently, there is no information linkage between option volatility smirk and market index returns over the nine years sample period after considering the control variables, global financial crisis dummies, and the subsample test. Research limitations/implications – This study finds that the option market and the DAX 30 index are informationally efficient. Implications of the findings are that the investors cannot profit from the information contained in the IVS since the information is simultaneously incorporated into option prices and the stock index prices. The findings of this study are applicable to other markets with European options and for market participants who seek to exploit short-term market divergence from efficiency. Originality/value – The relationship between IVS and stock price changes has not been investigated sufficiently in academic literature. This study looks at this relationship in the context of European options using high-frequency transactions data. Prior studies look at this relationship for only American options using daily data. Pricing efficiency of the European option market using high-frequency data have not been studied in the prior literature. The authors find different results for the German market based on this high-frequency data set.

Evaluation of sampling strategies for estimating ammonia emission factors for pig fattening facilities

Determining ammonia emission factors (EF) for fattening pig facilities is important from both a regulatory and a research point of view. However, measurements to determine an EF can be time consuming and costly. Several reduced sampling strategies were developed in the past to reduce the costs and measuring time, by taking into account parameters that influence NH3 emissions. A methodology to evaluate the precision and accuracy of estimated EFs solely as a function of the sampling frequency and strategy is demonstrated. This evaluation was done by using two long-term, high frequency datasets which both contained measurements during two consecutive pig fattening periods. These datasets were subjected to simulated sampling strategies. Long-term, low-frequency grab sampling proved to be more accurate than short-term monitoring. Repetitive short-term sampling events result in increased precision, but as this entails higher investment in time and money it is imperative to strike the balance between desired precision and available resources. A method to help as set guidelines to decide upon the number of short-term sampling events or the length of a long-term, low-frequency monitoring strategy is presented.

Generation of cubic power-law for high frequency intra-day returns: Maximum Tsallis entropy framework

A theoretical framework based on the maximum Tsallis entropy is proposed to explain the tail behavior of the intra-day stock returns, providing a rationale for the cubic law behavior for high frequency data. The specification of first two time-dependent moment constraints yields a q-Gaussian distribution for the intra-day stock returns. The value of the parameter q is estimated by minimizing appropriately modified Jensen–Shannon (JS) divergence in Tsallis entropy framework between q-Gaussian distribution and empirical NASDAQ 100 data. The estimated value of q yields the well-known empirically observed cubic law tail behavior of the intra-day stock returns which has been observed for high frequency data sets. To validate the cubic law stylized fact, five more data sets from high frequency NASDAQ 100, S&P 500 and NYSE index have been examined and it is found that the cubic law operates.

The intraday determination of liquidity in the NYSE LIFFE equity option markets

We exploit an extensive high-frequency data set of all individual equity options trading at New York Stock Exchange London International Financial Futures and Options Exchange (Amsterdam, London and Paris) in order to study the determination of liquidity during the trading day. In particular, we focus on two main aspects of option liquidity: (i) the intraday behaviour of equity option liquidity and its determinants and (ii) the influence of macroeconomic events and commonality on intraday equity option liquidity. Inventory management models cannot explain the intraday variation in option spreads and depths. Instead, we show that the option liquidity measures are strongly correlated with option volatility. Increases in volatility are associated with decreases in liquidity, a finding that is in line with information asymmetry models and the derivatives hedging theory. However, the relationship between spreads and volume varies across the three markets. Option liquidity reacts strongly to macroeconomic news announcements, especially US events. The average systematic liquidity component is 12% for Amsterdam, 14% for London and 16% for Paris.

Price discovery and the effects of fragmentation on market quality: evidence from Cypriot cross-listed stocks

Using a novel high-frequency data set, we examine the contribution of Greek trading to the price discovery process of a pair of Cypriot blue-chip, cross-listed stocks during overlapping trading hours. Additionally, we investigate the effects of market fragmentation on the home market’s quality, as measured by microstructure-based liquidity measures. Contrary to earlier studies from other markets, our findings show that foreign stock exchanges can act as the leading contributors to price discovery and can concentrate the majority of trading activity and produce the lowest transaction costs. Our results also show that market fragmentation can lead to negative effects on market liquidity.

2D/3D multi-phase Fresnel volume rays and applications to simultaneously update both velocity model and reflector geometry

Theoretically, Fresnel volume ray theory is more suitable for handling real seismic propagation problems because the traveltime depends not only on the velocity distribution along a traditional geometric ray but also on the velocity distribution within a vicinal region (referred to as first Fresnel volume, abbreviated as FFV) which embraces the geometric ray. In this study, we used an exact solution to calculate multi-phase FFV rays for both 2-D and 3-D cases and introduced a normalized coefficient to account for different contributions inside the FFV ray on the traveltimes. Furthermore, we draw a new formula to calculate the partial traveltime derivatives with respective to the velocity variations and depth changes of the reflectors and finally present a simultaneous inversion method for updating both velocity field and reflector geometry by using these multi-phase FFV rays for both in 2-D and 3-D cases. Using synthetic data examples, we compare the reconstructions of the velocity field and the reflector geometry using the FFV ray tomographic methods and the traditional ray tomography approaches. The simulated inversion results for both 2-D and 3-D cases show that the FFV ray tomographic method is advantageous over the traditional ray tomography method, especially when the ray density is relatively low. The other advantage for the FFV ray tomography method is that it can capture the coarse velocity structure and reflector geometry by starting with a low-frequency data set and then map the fine velocity structure and the detailed reflector geometry by using a high-frequency data set.

Analysis of the mixing processes in the subtropical Advancetown Lake, Australia

Summary This paper presents an extensive investigation of the mixing processes occurring in the subtropical monomictic Advancetown Lake, which is the main water body supplying the Gold Coast City in Australia. Meteorological, chemical and physical data were collected from weather stations, laboratory analysis of grab samples and an in-situ Vertical Profiling System (VPS), for the period 2008–2012. This comprehensive, high frequency dataset was utilised to develop a one-dimensional model of the vertical transport and mixing processes occurring along the water column. Multivariate analysis revealed that air temperature and rain forecasts enabled a reliable prediction of the strength of the lake stratification. Vertical diffusion is the main process driving vertical mixing, particularly during winter circulation. However, a high reservoir volume and warm winters can limit the degree of winter mixing, causing only partial circulation to occur, as was the case in 2013. This research study provides a comprehensive approach for understanding and predicting mixing processes for similar lakes, whenever high-frequency data are available from VPS or other autonomous water monitoring systems.

Heterotrophic denitrification constrains the upper limit of dissolved N2O-nitrate concentration ratio in agricultural groundwater

Indirect emission of nitrous oxide (N2O) through groundwater and surface drainage is commonly estimated based on the apparent positive relationship between the concentrations of nitrate ([NO3−]) and dissolved N2O ([N2O]) for regional and national assessments. Field observations of the ratio of [N2O]–[NO3−], however, rarely follow such relationship due to the complexity in N2O dynamics. Here we hypothesized that the temporal variation in [N2O] involves a phase shift as a function of [NO3−] during heterotrophic denitrification (HD) based on the fact that high [NO3−] inhibits N2O reduction to dinitrogen. We tested this hypothesis using the long-term, high frequency dataset of subsurface drainage water from the lysimeters under a cultivated soil in Japan. We identified the cases where [N2O] increased under high [NO3−] conditions (phase 1) and decreased under low [NO3−] conditions (phase 2) during the non-rainy periods. When [NO3−] exceeded 20.4 mg N L−1 across the entire dataset, we found the positive correlation between [N2O] and the concentration of dissolved carbon dioxide ([CO2]), an end-product of HD, in accordance with phase 1. When [NO3−] was 28.3 mg C L−1, we found the positive correlation between [N2O] and [NO3−], which agrees with phase 2. With this phase shift, [N2O]–[NO3−] ratio increased (phase 1) and reached a plateau (phase 2). Our results suggest that the progress of HD, especially phase 1, contributes to the increases in [N2O]–[NO3−] ratio and thus site-specific emission factor of N2O, EF5g.

The Transmission of the Financial Crisis in 1907: An Empirical Investigation

Using an extensive high-frequency data set, we investigate the transmission of financial crisis specifically focusing on the Panic of 1907, the final severe panic of the National Banking Era (1863-1913). We trace the transmission of the crisis from New York City trust companies to the New York City national banks through direct and indirect interconnections. Trust companies held cash balances at national banks, and these balances were liquidated as trust companies suffered depositor runs. Secondly, trust companies and national banks were notable creditors to the New York Stock Exchange; when trusts were suffering runs, the call loan market on the stock exchange seized. The crisis spread to the interior banks after the New York Clearing House banks restricted the convertibility of deposits into cash. Bond returns were sharply negative in the two weeks following the suspension. We highlight commonalities between the Panic of 1907 and the financial crisis of 2007-2009.

Application of Weather Generation to High Frequency and High Resolution Gridded Datasets in Sao Paulo

The adjusted stochastic weather generator to disaggregate the monthly climate averages into daily timescale is investigated due to the limitations of Climate Models in representing the daily weather fluctuations. We use weather generator for daily precipitation and maximum and minimum temperature are investigated for Sao Paulo State. We use the adjusted stochastic weather generator to disaggregate the monthly climate averages into daily timescale due to the limitations of Climate Models in representing the daily weather fluctuations. The weather generator WGEN was adjusted considering the statistical distribution of temperature and precipitation from gauges to the present climate. This derived dataset was first applied to investigate the ability of a Regional Climate Model version 4 (RegCM4) in reproduce the spatial variability of those distributions. Then, we developed to generate realizations of daily temperature and precipitation for gridding datasets over Sao Paulo while preserving the spatial and temporal distribution as well as produce gridded dataset of the proprieties that define the precipitation and temperature intensity and frequency. For temperature, we consider topographies to take the interpolation using distance square inverse. Comparison and evaluating generated time series with observations and regional model shows that performance of the weather generator tested in term of the frequency and extreme events reproduced the observed statistical distribution.

Pressure autoregulation monitoring and cerebral perfusion pressure target recommendation in patients with severe traumatic brain injury based on minute-by-minute monitoring data.

In severe traumatic brain injury, a universal target for cerebral perfusion pressure (CPP) has been abandoned. Attempts to identify a dynamic CPP target based on the patient's cerebrovascular autoregulatory capacity have been promising so far. Bedside monitoring of pressure autoregulatory capacity has become possible by a number of methods, Czosnyka's pressure reactivity index (PRx) being the most frequently used. The PRx is calculated as the moving correlation coefficient between 40 consecutive 5-second averages of intracranial pressure (ICP) and mean arterial blood pressure (MABP) values. Plotting PRx against CPP produces a U-shaped curve in roughly two-thirds of monitoring time, with the bottom of this curve representing a CPP range corresponding with optimal autoregulatory capacity (CPPopt). In retrospective series, keeping CPP close to CPPopt corresponded with better outcomes. Monitoring of PRx requires high-frequency signal processing. The aim of the present study is to investigate how the processing of the information on cerebrovascular pressure reactivity that can be obtained from routine minute-by-minute ICP and MABP data can be enhanced to enable CPPopt recommendations that do not differ from those obtained by the PRx method, show the same associations with outcome, and can be generated in more than two-thirds of monitoring time. The low-frequency autoregulation index (LAx) was defined as the moving minute-by-minute ICP/MABP correlation coefficient calculated over time intervals varying from 3 to 120 minutes. The CPPopt calculation was based on LAx-CPP plots and done for time windows between 1 and 24 hours and for each LAx type. The resulting matrix of CPPopts were then averaged in a weighted manner, with the weight based on the goodness of fit of a U-shape and the lower value of the LAx corresponding to the U-bottom, to result in a final CPPopt recommendation. The association between actual CPP/CPPopt and outcome was assessed in the multicenter Brain Monitoring with Information Technology Research Group (BrainIT) database (n = 180). In the Leuven-Tübingen database (60-Hz waveform data, n = 21), LAx- and PRx-based CPPopts were compared. In the BrainIT database, CPPopt recommendations were generated in 95% of monitoring time. Actual CPP being close to LAx-based CPPopt was associated with increased survival. In a multivariate model using the Corticosteroid Randomization After Significant Head Injury (CRASH) model as covariates, the average absolute difference between actual CPP and CPPopt was independently associated with increased mortality. In the high-frequency data set no significant difference was observed between PRx-based and LAx-based CPPopts. The new method issued a CPPopt recommendation in 97% of monitoring time, as opposed to 44% for PRx-based CPPopt. Minute-by-minute ICP/MABP data contain relevant information for autoregulation monitoring. In this study, the authors' new method based on minute-by-minute data resolution allowed for CPPopt calculation in nearly the entire monitoring time. This will facilitate the use of pressure reactivity monitoring in all ICUs.

Temporal Variability of Indoor Air Concentrations under Natural Conditions in a House Overlying a Dilute Chlorinated Solvent Groundwater Plume

Current vapor intrusion (VI) pathway assessment heavily weights concentrations from infrequent (monthly-seasonal) 24 h indoor air samples. This study collected a long-term and high-frequency data set that can be used to assess indoor air sampling strategies for answering key pathway assessment questions like: "Is VI occurring?", and "Will VI impacts exceed thresholds of concern?". Indoor air sampling was conducted for 2.5 years at 2-4 h intervals in a house overlying a dilute chlorinated solvent plume (10-50 μg/L TCE). Indoor air concentrations varied by 3 orders of magnitude (<0.01-10 ppbv TCE) with two recurring behaviors. The VI-active behavior, which was prevalent in fall, winter, and spring involved time-varying impacts intermixed with sporadic periods of inactivity; the VI-dormant behavior, which was prevalent in the summer, involved long periods of inactivity with sporadic VI impacts. These data were used to study outcomes of three simple sparse data sampling plans; the probabilities of false-negative and false-positive decisions were dependent on the ratio of the (action level/true mean of the data), the number of exceedances needed, and the sampling strategy. The analysis also suggested a significant potential for poor characterization of long-term mean concentrations with sparse sampling plans. The results point to a need for additional dense data sets and further investigation into the robustness of possible VI assessment paradigms. As this is the first data set of its kind, it is unknown if the results are representative of other VI-sites.

Labor Supply Decisions of Singaporean Cab Drivers

We use a unique administrative dataset of over 10,000 taxi drivers in Singapore to study the labor supply decisions of these drivers. Our study uses a high frequency dataset to test whether cabdrivers exhibit reference-dependence preferences. We estimate cabdrivers’ day and day-of-the-week hours and income targets using ex-post data, and find that the targets have a strong and positive impact on cabdrivers’ probability of stopping work. Cabdrivers with high cumulative hours and/or cumulative income for the day are more likely to stop work earlier when they reach their hour and income targets. We also find that cabdrivers’ income targets are independent of ex-post positive and negative daily income shocks, and that their targeting behavior is persistent. The results support the idea that rational expectations underpin reference-dependence in the labor supply decisions of cabdrivers in our sample.

Risk spillovers in international equity portfolios

Abstract We define risk spillover as the dependence of a given asset variance on the past covariances and variances of other assets. Building on this idea, we propose the use of a highly flexible and tractable model to forecast the volatility of an international equity portfolio. According to the risk management strategy proposed, portfolio risk is seen as a specific combination of daily realized variances and covariances extracted from a high frequency dataset, which includes equities and currencies. In this framework, we focus on the risk spillovers across equities within the same sector (sector spillover), and from currencies to international equities (currency spillover). We compare these specific risk spillovers to a more general framework (full spillover) whereby we allow for lagged dependence across all variances and covariances. The forecasting analysis shows that considering only sector- and currency-risk spillovers, rather than full spillovers, improves performance, both in economic and statistical terms.