Intraday Trading Data Research Articles

Hacks and the price synchronicity of bitcoin and ether

We use intraday trading data from the Kraken exchange to calculate the daily price synchronicity of Bitcoin and Ether from February 2018 to December 2022. We then use a comprehensive report provided by christalblockchain.com to investigate the impact of hacks on price synchronicity between the top two cryptocurrencies. Our results show that price synchronicity, as measured by the realized correlation, is consistently positive throughout the sample period, with only one (negative) exception. We further uncover a positive relationship between hacking events and the future price synchronicity of Bitcoin and Ether. This result is robust to an alternative price synchronicity measure.

The Impact of Shortening the Matching Time on The Cancellation of Orders by Investors

This article mainly discusses the impact of the improvement of market transparency and liquidity on investor withdrawal behavior after the Taiwan Stock Exchange shortened the matching seconds from 10 seconds to 5 seconds. We use the intraday trading data of 5 months before and after December 2014 for actual measurement. We use the ratio of canceled orders as an indicator and use the paired sample T-test to examine whether there is a significant difference in the ratio of canceled orders between institutional investors and individual investors before and after the system change. The empirical results show that, except for the significant increase in the cancellation rate in Step 1 of legal persons and retail investors, all others have decreased significantly. Since the possibility of the cancellation in Step 1 being caused by manipulation is not high, it can be inferred that it is a purely wrong order, and the decrease in the ratio shows that the new system is helpful to the entire market, because investors have become more cautious when placing orders.

Learning Financial Networks with High-frequency Trade Data.

Financial networks are typically estimated by applying standard time series analyses to price-based economic variables collected at low-frequency (e.g., daily or monthly stock returns or realized volatility). These networks are used for risk monitoring and for studying information flows in financial markets. High-frequency intraday trade data sets may provide additional insights into network linkages by leveraging high-resolution information. However, such data sets pose significant modeling challenges due to their asynchronous nature, complex dynamics, and nonstationarity. To tackle these challenges, we estimate financial networks using random forests, a state-of-the-art machine learning algorithm which offers excellent prediction accuracy without expensive hyperparameter optimization. The edges in our network are determined by using microstructure measures of one firm to forecast the sign of the change in a market measure such as the realized volatility of another firm. We first investigate the evolution of network connectivity in the period leading up to the U.S. financial crisis of 2007-09. We find that the networks have the highest density in 2007, with high degree connectivity associated with Lehman Brothers in 2006. A second analysis into the nature of linkages among firms suggests that larger firms tend to offer better predictive power than smaller firms, a finding qualitatively consistent with prior works in the market microstructure literature.

The Impact of COVID-19 on Stock Market in China

This paper studies the impact of the COVID-19 on the stock ambiguity, risks, liquidity, and stock prices in China stock market, before and after the outbreak of COVID-19 during the Chinese Spring Festival holidays in 2020. We measure stock ambiguity using the intraday trading data. The outbreak of COVID-19 has a significant impact on the average stock ambiguity, risk, and illiquidity in China and induces structural break in the market average ambiguity. However, the equity premium and liquidity premium change little during the same period. The market average stock ambiguity and risks decrease, and stock liquidity improves to pre-pandemic levels as the pandemic is under control in China. The market average stock ambiguity and risks in China increase again when the confirmed new cases in the U.S. surge in the second half of 2020. We also find a “flight-to-liquidity” phenomenon, and the equally-weighted (value-weighted) 20-trading-day liquidity premium declined significantly to about –4.42% (–6.48%) during the fourth quarter of 2020.

The Effects of Trader Type on Market Maker’s Effective Bid-Ask Spread

Using intraday trading data on U.S. Treasury note futures, we analyze the effects of trader type on market maker’s effective bid-ask spread. We find that market makers tend to experience losses when price uncertainty increases in the market, even though market maker’s effective BASs are positive on average. When the overall trading volume increases, the effects of economies of scale and competition between market makers and exchange member institutions cause market makers to narrow the effective bid-ask spread. However, when off-exchange traders increase their trading volume, market makers respond by lowering their buy prices and increasing sell prices, thereby widening the effective bid-ask spread. The results reveal that market makers protect themselves against risk when trading with off-exchange traders.

Manipulation in the bond market and the role of investment funds: Evidence from an emerging market

The aim of our research is to test for manipulation in the bond market, and to establish whether the practice of benchmarking investment funds encourages such interference. We analyse end-of-day and intraday trading data from the Borsa Istanbul Bond Securities Market between 2014 and 2018 along with specific data from investment fund sector over the same period. Our results indicate that treasury bonds have most likely been manipulated at the months' end by trading activities of investment funds. Price and volume rose significantly in the run up to the end of the month before falling off in the days following.

Empirical Stock Return Prediction as an Extension of Merton’s Asset Return Model

We extend Merton’s 1976 asset return analysis which relied on intraday trade data to estimate volatility by introducing drift estimation and equity premium dynamics leading to predictability of short run daily returns under appropriate conditions. We find empirical support for the identification of trade dates associated with commencement of change in return behavior as compared to recent past behavior. We show that drift level, change in drift, equity risk premium level, change in equity risk premium level and local variance shock are individually and jointly predictive to determine macro epoch return behavior, as well as identify optimal buy/sell trade dates and price. Our empirical model can be viewed as a super-set framework congruent with Merton’s (1976) findings as well as Lo’s (1995) findings as a subset. Empirical results concur in theory, distributional arguments and time series analysis of previous work on predictability of equity return processes.

Temporal Effects in Index CDS Trading and Liquidity

Using high-frequency intraday trading and quoting data, we study the temporal effects in index credit default swap (CDS) trading and liquidity. We find strong intraday variations in index CDS trading activities and liquidity. Unlike the U-shaped trading pattern in the equity market, index CDSs exhibit a hump-shaped intraday trading pattern. Trading costs and return volatility are substantially lower in active trading hours, indicating that investors tend to trade when trading costs and price informativeness are low. We also observe strong day-of-the-week effects. Overall, these findings improve our understanding of the trading costs and liquidity in the over-the-counter derivatives market.

Stock Market prediction on High frequency data using Long-Short Term Memory

Abstract High Frequency Trading (HFT) is part of algorithmic trading, and one of the biggest changes that happened in the last 15 years. HFT or nanotrading represents the ability, for a trader, to take orders within very short delays. This paper presents a model based on technical indicators with Long Short Term Memory in order to forecast the price of a stock one-minute, five-minutes and ten-minutes ahead. First, we get the S&P500 intraday trading data from Kaggle, then we calculate technical indicators and finally, we train the regression Long-Short Term Memory model. Based on the price history, alongside technical analysis indicators and strategies, this model is executed, and the results are analyzed based on performance metrics and profitability. Experiment results show that the proposed method is effective as well as suitable for prediction a few minutes before.

Investigating the Dynamics Between Price Volatility, Price Discovery, and Criminality in Cryptocurrency Markets

This paper identifies several stylised facts relating to the volatility and price discovery process from eight cryptocurrencies utilising an empirical analysis of intra-day trading data to uncover four main results. First, cryptocurrencies exhibit weekend-volatility effects while intra-day volatility is found to be influenced by international trading times, periods of substantial volatility in the markets for oil, and GBP/USD and cybercrime events. Secondly, a thorough investigation of recent cybercriminality identifies that cryptocurrency hacks are found to increase both the volatility of the currency hacked and the correlations across the hacked currency and other cryptocurrencies. Thirdly, hacks significantly reduce price discovery sourced within the hacked currency relative to other cryptocurrencies. Finally, there are abnormal returns associated with the hacks observed in the hours prior to the actual hacking event, which reverts to zero at the time of the public announcement of the hack.

Overreaction in Trading

Using intraday trading data during the 2008 financial crisis, from the Standard and Poor’s Depository Receipt (SPDR) market, we test for evidence of the informational advantage of traders. In addition, we examine the effect of pricing error on trade price. If traders are rational, and have accurate information, they will only purchase an asset at a premium (discount) if they have reason to believe that the fundamental value of that asset will increase (decrease). Our results show that the trading price of the SPDR does not significantly predict the movement of underlying asset values. This finding is consistent with traders overreacting to disparities between price and underlying value during the financial crisis.

The significance of calendar effects in the electricity market

How to balance supply and demand has become a long-term question in the electricity market, and anomalies related to calendar issues are critical factors to affect the resource allocation. This paper introduces a test method to assess the significance of all possible calendar effects in different time frequencies. We implement our test method to the largest electricity trading platform in the United States. Using the high-frequency intraday trading data, we assess the calendar effects in different time frequencies (Day-of-the-week, Hour-of-the-day, Month-of-the-year, Day-of-the-month and season). Our results confirm that calendar effects exist in every dimension of time frequency, and specify those calendar effects with statistical significance. Moreover, this study discovers commonalities between electricity markets and financial markets, which makes it feasible to apply the management of financial markets to electricity markets. Besides, the detected calendar effects depict periodic patterns of market inequilibrium and facilitate the implementation of corresponding technical solutions in electricity markets.

Stock Splits and Attracting Attention: Are Investors Buying Abnormally More Around Stock Split Announcements?

In this research I study whether stock splits attract market’s attention by exploring how investors are trading around event announcement dates. By employing high frequency intraday trading data from NYSE Trades and Quotes (TAQ) database I compute net abnormal buying around split announcements. The empirical tests on a matched pair sample of splitting and matching firms show that stock splits serve as attention attracting tool and investors are buying abnormally more around the announcements. Additional analysis confirms this finding – abnormal buying is significantly higher for larger splits. Furthermore, investors are more attracted to the splits that deliver higher subsequent long run stock performance.

NYSE closure and global equity trading: The case of cross-listed stocks

Cross-listing in the US opens a new information channel for cross-listed stocks and makes the cross-listed stocks more dependent on the US market as a source of information and intermarket competition for order flow. We test this proposition by examining the effects of closures of the NYSE on trading of cross-listed stocks in their home markets. We use daily and intraday trading data for stocks from 36 countries that cross-list on the NYSE during the period 2009–2018 to estimate measures of stock liquidity and information efficiency of stock prices on normal days and on days when the NYSE is closed due to US public holidays and natural hazards. We document that, when the NYSE is closed, stocks that cross-list on the NYSE see a significant reduction in stock liquidity and information efficiency of stock prices in their home markets, over and above the reduction experienced by their domestic counterparts. In addition, we investigate several channels that could explain the documented reduction in liquidity and price efficiency - information, substitute and arbitrage channels. We document that cross-listed stocks are more affected by NYSE closure when they have greater dependence on the US market, and when arbitrage strategies are easier to implement, providing some evidence for the information channel and arbitrage channel arguments.

Global price discovery in the Australian dollar market and its determinants

With Australia's growing importance in the global economy, the Australian dollar (AUD) has started to play a significant role in the global Foreign Exchange (FX) market. Using intraday trading data over the period of January 1999 to December 2013, we examine the determinants of price discovery in the AUD markets at two different time horizons. We find that the short-run determinants of price discovery include macroeconomic news, order flows and market state variables (i.e., return, volatility, trading volume and bid/ask spread). After controlling for the cross-region information flow and dynamic self-dependence, we find that more favorable market states and more unexpected order flows on macroeconomic news announcement days make a significant contribution to price discovery in the AUD market. Furthermore, we found that a higher level of market integration and consolidation contributes to price discovery process in the long-run.

NYSE Closure and Global Equity Trading: The Case of Cross-Listed and Domestic Stocks

Cross-listing in the US opens a new information channel for cross-listed stocks and makes the cross-listed stocks more dependent on the US market as a source of information and intermarket competition for order flow. We test this proposition by examining the effects of closures of the NYSE on trading of cross-listed stocks in their home markets. We use daily and intraday trading data for stocks from 36 countries that cross-list on the NYSE during the period 2009-2018 to estimate measures of stock liquidity and information efficiency of stock prices on normal days and on days when the NYSE is closed due to US public holidays and natural hazards. We document that, when the NYSE is closed, stocks that cross-list on the NYSE see a significant reduction in stock liquidity and information efficiency of stock prices in their home markets, over and above the reduction experienced by their domestic counterparts. In addition, we investigate several channels that could explain the documented reduction in liquidity and price efficiency - information, substitute and arbitrage channels. We document that cross-listed stocks are more affected by NYSE closure when they have greater dependence on the US market, and when arbitrage strategies are easier to implement, providing some evidence for the information channel and arbitrage channel arguments.

Herd Behaviour: How Decisive is the Noise in the NSE and BSE Stock Markets?

Do investors in the stock markets act/react on true information or noise? Do they believe on their own information or simply herd? The study seeks to explore these typical research queries from the behavioral finance perspectives. In particular, it develops a new theory of herding behavior and extends the models of Banerjee (1992) and Bikhchandani, Hirshleifer, and Welch (1992). The study also empirically tests the same on the Indian context with the high frequency intraday trading data for the real trade-time or time-stamp, trade-volume, and trade-price of ten sample scripts listed for their trading in both markets - the Bombay Stock Exchange (BSE) and the National stock Exchange (NSE). The study contributes to the literature with original findings. It shows that investors in the two Indian stock markets show crowd of positive and negative herding as well significantly and there is huge noise along with information in the markets equilibrium pricing mechanism.

A comparison of bid-ask spread proxies: evidence from Borsa Istanbul futures

We analyze the performance of five different methods appearing in the market microstructure literature in predicting effective and quoted bid-ask spreads (Roll, LOT Mixed, Effective Tick, High-Low and Closing Percent Quoted Spread proxies). With data from index futures, currency futures and gold futures traded in Borsa Istanbul and taking percent effective and percent quoted spreads obtained from intraday trade and quote data as benchmarks, we calculate and compare the correlations and root mean square errors of the spread measures. Results show that none of the proxies is successful enough in estimating effective or quoted spread although under normal market conditions, Effective Tick appears to perform best. Keywords: bid-ask spread, Borsa Istanbul, futures market JEL Classification: G10, G14, G23

Herd Behaviour: How Decisive is the Noise in the NSE and BSE Stock Markets?

Do investors in the stock markets act/react on true information or noise? Do they believe on their own information or simply herd? The study seeks to explore these typical research queries from the behavioral finance perspectives. In particular, it develops a new theory of herding behavior and extends the models of Banerjee (1992) and Bikhchandani, Hirshleifer, and Welch (1992). The study also empirically tests the same on the Indian context with the high frequency intraday trading data for the real trade-time or time-stamp, trade-volume, and trade-price of ten sample scripts listed for their trading in both markets - the Bombay Stock Exchange (BSE) and the National stock Exchange (NSE). The study contributes to the literature with original findings. It shows that investors in the two Indian stock markets show crowd of positive and negative herding as well significantly and there is huge noise along with information in the markets equilibrium pricing mechanism.

A PIN per day shows what news convey: the intraday probability of informed trading

This paper develops a new intraday estimation procedure for the sequential microstructure trading model initially proposed by Easley et al. (Rev Financ Stud 10:805–835, 1997a). Using a full year of intraday trading data for the top 100 German stocks, we demonstrate how the new estimation procedure eliminates or significantly reduces the shortcomings of the original approach in recent, high-frequency trading environments. We slice a trading day in buckets of several minutes’ length to obtain one estimate of the composite variable probability of informed trading (PIN) per day. This approach makes PIN applicable in short horizon event studies. Convergence rates are above 95 % even for the most liquid stocks and the model’s underlying assumptions of independence for the arrival of traders and information events are fulfilled to a much higher degree than in the original approach. An empirical application in an event study type setting demonstrates how official announcements stipulated in German insider trading legislation significantly reduce information asymmetry upon public disclosure.