High Frequency Trading Research Articles

HFTViz: Visualization for the exploration of high frequency trading data

High Frequency Trading (HFT), mainly based on high speed infrastructure, is a significant element of the trading industry. However, trading machines generate enormous quantities of trading messages that are difficult to explore for financial researchers and traders. Visualization tools of financial data usually focus on portfolio management and the analysis of the relationships between risk and return. Beside risk-return relationship, there are other aspects that attract financial researchers like liquidity and moments of flash crashes in the market. HFT researchers can extract these aspects from HFT data since it shows every detail of the market movement. In this paper, we present HFTViz, a visualization tool designed to help financial researchers explore the HFT dataset provided by NASDAQ exchange. HFTViz provides a comprehensive dashboard aimed at facilitate HFT data exploration. HFTViz contains two sections. It first proposes an overview of the market on a specific date. After selecting desired stocks from overview visualization to investigate in detail, HFTViz also provides a detailed view of the trading messages, the trading volumes and the liquidity measures. In a case study gathering five domain experts, we illustrate the usefulness of HFTViz.

High-frequency trading and market quality: The case of a “slightly exposed” market

Impacts of high-frequency trading (HFT) on market quality and various actors have been broadly studied. However, what happens when HFT is not a prominent figure in a market remains relatively unexplored. The paper seeks to answer this question focusing on 30 blue chip stocks in an emerging market, Borsa Istanbul, through Dec 2015 to Mar 2017. Despite a low share in the overall activity, HFT has observable effects, i.e. liquidity provision by non-HFT traders significantly reduces with HFT. Moreover, HFT generates profits on both positive and negative return days. Yet, HFT activity does not have an impact on volatility. These findings raise concerns regarding HFT and show potential externalities are not specific to the markets with HFT dominance.

The adaptive market hypothesis and high frequency trading

This paper uses NASDAQ order book data for the S&P 500 exchange traded fund (SPY) to examine the relationship between one-minute, informational market efficiency and high frequency trading (HFT). We find that the level of efficiency varies widely over time and appears to cluster. Periods of high efficiency are followed by periods of low efficiency and vice versa. Further, we find that HFT activity is higher during periods of low efficiency. This supports the argument that HFTs seek profits and risk reduction by actively processing information, through limit order additions and cancellations, during periods of lower efficiency and revert to more passive market-making and rebate-generation during periods of higher efficiency. These findings support the argument that the adaptive market hypothesis (AMH) is an appropriate description of how prices evolve to incorporate information.

TradeBot: Bandit learning for hyper-parameters optimization of high frequency trading strategy

Quantitative trading takes advantage of mathematical functions for automatically making stock or futures trading decisions. Specifically, various trading strategies that proposed by human-experts are associated with weight hyper-parameters to determine the probability of selecting a specific strategy according to market conditions. Prior work manually adjusting the weight hyper-parameters is error-prone, because the essential advantage of quantitative trading, i.e., automation, is lost. In this paper, we propose a dynamic parameter tuning algorithm, i.e., TradeBot, based on bandit learning for quantitative trading. We consider sequentially selecting hyper-parameters of rules for trading as a bandit game, where a set of hyper-parameters of trading rule is considered as an action. A novel reward-agnostic Upper Confidence Bound bandit method is proposed to solve the automatically trading problem with a reward function estimated by inverse reinforcement learning. Experimental results on China Commodity Futures Market Data show state-of-the-art performance. To our best knowledge, this is one of the first work deployed in the online trading system via reinforcement learning, in published literature.

Are high frequency traders responsible for extreme price movements?

This is the first paper to investigate the association between losses to liquidity providers and price jumps, separately for high frequency traders [HFT] and low frequency traders. Highly reliable data from Nasdaq identifies the trade direction of each trade and the trades that have HFT participation. To determine if trades executed by high or low frequency traders are more frequently associated with a price jump, we use the volume-synchronised probability of informed trading [VPIN] metric. We find that VPIN rapidly increases prior to a price jump but does not peak until after a price jump. HFT liquidity demand and supply prior to a price jump shows that HFT reduce their trading prior to price jumps. Low frequency traders hence remain as the main market participants during price jumps and are the information channel that leads to price jumps. Regulators and broker–dealers can implement strategies to protect investors against price jumps when VPIN rapidly increases and the proportion of HFT to other traders falls.

Attention: How high-frequency trading improves price efficiency following earnings announcements

Recent research indicates that high-frequency trading (HFT) helps incorporate fundamental information into prices, but how this happens is unclear. We examine reduced attention constraints as an important channel through which HFT enhances price efficiency. Using multiple proxies of attention constraints, we find that price inefficiencies are reduced by 65%–100% when high-frequency traders (HFTs) trade following low attention earnings announcements: initial price responses are larger and post-earnings-announcement drift is reduced. Results are not driven by firm size or announcement time-of-day. Our findings highlight how limited attention, a human bias affecting asset prices, is mitigated when machines trade.

Quantum Prisoner's Dilemma and High Frequency Trading on the Quantum Cloud.

High-frequency trading (HFT) offers an excellent use case and a potential killer application of the commercially available, first generation quasi-quantum computers. To this end, we offer here a simple game-theoretic model of HFT as the famous two player game, Prisoner’s Dilemma. We explore the implementation of HFT as an instance of Prisoner’s Dilemma on the (quasi) quantum cloud using the Eisert, Wilkens, and Lewenstein quantum mediated communication protocol, and how this implementation can not only increase transaction speed but also improve the lot of the players in HFT. Using cooperative game-theoretic reasoning, we also note that in the near future when the internet is properly quantum, players will be able to achieve Pareto-optimality in HFT as an instance of reinforced machine learning.

"Overview of Blockchain Technology: Consensus, Architecture, and Its Future Trends"

The blockchain technology that underpins Bitcoin has lately gained a lot of traction. Blockchain is a decentralized transaction ledger that is immutable. Blockchain applications are emerging in a variety of fields, including financial services, reputation systems, and the Internet of Things (IoT), among others. Nonetheless, concerns with blockchain knowledge, like as security or scalability, have yet to be solved. Bitcoin can only process 7 transactions per second, making it unsuitable for highfrequency trading. Larger blocks, on the other hand, imply more storage capacity and less network transfer. Because fewer individuals desire to own such a large blockchain, this will ultimately lead to centralization. This article provides a comprehensive description of blockchain technology. The researcher first provides an overview of blockchain architecture and then compares many commonly utilized consensus algorithms in different blockchains. A brief summary of technical difficulties and current advances is also included. They also highlight possible future blockchain developments.

Performance Evaluation of High Frequency Trading using a Deep Learning

Performance Evaluation of High Frequency Trading using a Deep Learning

Intraday volatility smile: Effects of fragmentation and high frequency trading on price efficiency

In 2007, the European Markets in Financial Instruments Directive ended the national concentration rule. As a result, market fragmentation has accelerated across multiple trading venues. Spatial fragmentation might create opportunities and incentives for High Frequency arbitrageurs to fill the void left by the lack of Reg NMS type order routing requirements in Europe, without neglecting market integrity. This paper examines intra-day volatility and price efficiency through the metric of the normalized volatility ratio for the years 2006, 2012 and 2013 for Euronext Paris, BATS and Chi-X Europe. Our findings show that price determination remains inefficient at market opening due to the complexity of price discovery activity following a period of non-trading and heavy information releases. However, we demonstrate that an active participation of high-frequency traders significantly improves market efficiency at opening session.

Can HFT profit in Chinese stock market?

We develop a new method to proxy the High-Frequency Trading (HFT) activity of Chinese stock market by searching order-trade data with low latency and message linkage. We find that daily average percent of HFT detected by the method is 3.8029% between September 2014 and December 2020. Meanwhile, stocks with the highest participation of HFT manifest best profitability, and the profit advantage declines gradually after 2019.

Censored expectation maximization algorithm for mixtures: Application to intertrade waiting times

In a previous analysis the problem induced by “zero-inflation” in time series data (caused by high frequency trading in the electronic order book) was handled by left-truncating the waiting times between consecutive limit orders . We demonstrated, using rigorous statistical methods, that the truncated Weibull distribution describes the corresponding stochastic dynamics for the entire range of inter-arrival limit order waiting times, except for a region close to zero. However, since the truncated Weibull distribution was not able to describe the prodigious “zero-inflated” probability mass in the neighbourhood of zero (making up approximately 50% of the data for limit orders), it became clear that the entire probability distribution must be a mixture distribution of which the Weibull distribution is a significant part. To investigate this idea, we use a “censored expectation–maximization algorithm” to analyse the intertrade waiting times data for four selected stocks trading on the London Stock Exchange. The intertrade waiting times usually have a much lower percentage of zero inflation, typically around 2.5%. Making use of this new method and testing various mixture models, we show that the desired mixture consists of the Weibull distribution with the universal shape parameter of β≃0.57 plus an additional exponential distribution. This is the same value for the shape parameter found already in our previous study. The “1 exponential + 1 Weibull” mixture describes the intertrade waiting times extremely well at all time scales. While the Weibull component dominates in the transition and tail regions the exponential distribution explains the “zero-inflated” excess mass.

Exponential Gradient with Momentum for Online Portfolio Selection

Online portfolio selection is a fundamental research problem, which has drawn extensive investigations in both machine learning and computational finance communities. The evolution of electronic trading has contributed to the growing prevalence of High-Frequency Trading (HFT) in recent years. Generally, HFT requires trading strategies to be fast in execution. However, the existing online portfolio selection strategies fail to either satisfy the demand for high execution speed or make effective utilization of historical data. In response, we propose a framework named Exponential Gradient with Momentum (EGM) which integrates EG with an acknowledged optimization method in stochastic learning, i.e., momentum. Specifically, momentum boosts the performance of EG by making full use of historical information. Most essentially, EGM can execute with only constant memory and running time in the number of assets per trading period, thus overcoming the drawback of most online strategies. The theoretical analysis reveals that EGM bounds the regret sublinearly. The extensive experiments conducted on four real-world datasets demonstrate that EGM outperforms relevant strategies with respect to comprehensive evaluation metrics.

Quantifying the High-Frequency Trading “Arms Race”

Abstract We use stock exchange message data to quantify the negative aspect of high-frequency trading, known as “latency arbitrage.” The key difference between message data and widely familiar limit order book data is that message data contain attempts to trade or cancel that fail. This allows the researcher to observe both winners and losers in a race, whereas in limit order book data you cannot see the losers, so you cannot directly see the races. We find that latency arbitrage races are very frequent (about one per minute per symbol for FTSE 100 stocks), extremely fast (the modal race lasts 5–10 millionths of a second), and account for a remarkably large portion of overall trading volume (about 20%). Race participation is concentrated, with the top six firms accounting for over 80% of all race wins and losses. The average race is worth just a small amount (about half a price tick), but because of the large volumes the stakes add up. Our main estimates suggest that races constitute roughly one-third of price impact and the effective spread (key microstructure measures of the cost of liquidity), that latency arbitrage imposes a roughly 0.5 basis point tax on trading, that market designs that eliminate latency arbitrage would reduce the market’s cost of liquidity by 17%, and that the total sums at stake are on the order of $5 billion per year in global equity markets alone.

The Ethics of Financial Market Making and Its Implications for High-Frequency Trading

During the last 20 years, the financial sector has undergone an unprecedented transformation due to new regulations and the implementation of several technological advancements. The combination of regulation and technology has brought about new financial processes that have fundamentally changed how financial market making is done. This paper studies the ethics of financial market making and its implications for one of the most controversial financial innovations of modern times, namely high-frequency trading (HFT). We claim that the Aristotelian distinction between natural chrematistics, which is aimed at serving the real economy, and unnatural chrematistics, whose ultimate purpose is wealth accumulation, can be a useful criterion to assess the ethics of financial market making and the goodness of an innovation as HFT, and how it can serve the common good of society. This approach can be defined as ‘purpose oriented’ or ‘purpose fulfillment’.

Profitability and liquidity provision of HFTs during large price shocks: Does relative tick size matter?

Using tick data identifying high-frequency traders (HFTs), this paper studies the trading profits and liquidity provision of HFTs during large price shocks. Previous studies report mixed evidence on whether HFTs provide or take liquidity in such events. Empirical findings in this paper suggest that relative tick size matters: HFTs provide liquidity when the shock is idiosyncratic and the relative tick size is large, but in this case, they do not earn profits from trading. On average, HFTs can earn profits against non-HFTs because they aggressively take liquidity and trade in the direction of the shocks for stocks of small relative tick sizes.

High-frequency trading and networked markets

Financial markets have undergone a deep reorganization during the last 20 y. A mixture of technological innovation and regulatory constraints has promoted the diffusion of market fragmentation and high-frequency trading. The new stock market has changed the traditional ecology of market participants and market professionals, and financial markets have evolved into complex sociotechnical institutions characterized by a great heterogeneity in the time scales of market members' interactions that cover more than eight orders of magnitude. We analyze three different datasets for two highly studied market venues recorded in 2004 to 2006, 2010 to 2011, and 2018. Using methods of complex network theory, we show that transactions between specific couples of market members are systematically and persistently overexpressed or underexpressed. Contemporary stock markets are therefore networked markets where liquidity provision of market members has statistically detectable preferences or avoidances with respect to some market members over time with a degree of persistence that can cover several months. We show a sizable increase in both the number and persistence of networked relationships between market members in most recent years and how technological and regulatory innovations affect the networked nature of the markets. Our study also shows that the portfolio of strategic trading decisions of high-frequency traders has evolved over the years, adding to the liquidity provision other market activities that consume market liquidity.

Do geopolitical tensions instigate mindless following in stock markets? An empirical enquiry into the indices of CNX Nifty HFT

Geopolitical tensions between nations play a crucial role in triggering volatility and affecting the investors’ behavior in stock markets. This empirical work attempts to detect the traces of herding and bubble embedded in the Indian stock indices of CNX Nifty 50 and CNX Nifty 100 (both in High-Frequency Trading domains) during the latest events of geopolitical tensions escalated between India-China and India-Pakistan. An event window approach is employed to capture the impact of these events on herding behavior and information uncertainty in the considered stock indices. Multifractal Detrended Fluctuation Analysis (MFDFA) is applied to compute the Hurst value in all the trading days of the event window. The results of both indices exhibit conclusive evidence of herding and bubble formation during the overall period of geopolitical tensions between India-China and India-Pakistan. However, the degree of herding in the stock indices intensifies to a profound pattern when the tensions between India and China escalated into deadly violent clashes, and also during the heightened tensions between India and Pakistan that eventually ended up in airstrikes across the boundaries. The overall level of information uncertainty depicted by entropy is within control. The volatility in these stock indices has been confirmed to follow a unidirectional pattern. AcknowledgementsThe authors express their sincere thanks of gratitude to Dr. Bikramaditya Ghosh (Professor, Department of Finance and Analytics, RV Institute of Management, Bangalore, India) for his instrumental role in encouraging and motivating them to accomplish this research task. The authors also extend their sincere thanks to Dr. Manu K.S. (Assistant Professor, School of business and management, CHRIST (Deemed to be university), Bangalore, India) for his continued support throughout this empirical investigation.

Relativistic Option Pricing

The change of information near light speed, advances in high-speed trading, spatial arbitrage strategies and foreseen space exploration, suggest the need to consider the effects of the theory of relativity in finance models. Time and space, under certain circumstances, are not dissociated and can no longer be interpreted as Euclidean. This paper provides an overview of the research made in this field while formally defining the key notions of spacetime, proper time and an understanding of how time dilation impacts financial models. We illustrate how special relativity modifies option pricing and hedging, under the Black–Scholes model, when market participants are in two different reference frames. In particular, we look into maturity and volatility relativistic effects.

Machine Learning and Speed in High-Frequency Trading

The creative destruction wrought by high-frequency algorithmic tradinghas raised increasing concerns about the eect of machine learning behaviorsand ultra high-frequency trading in finnancial markets. By employing a geneticalgorithm with a classifer system as an adaptive learning tool, we address someof these concerns by studying a dynamic limit order market model with asymmetricinformation and varying speeds of high-frequency trading (HFT). We showthat HFT benefits uninformed traders, improves information effciency but reducesmarket liquidity. We find that there is a trade-off where a competition effect erodesthe information and speed advantages of high-frequency traders, increasing tradingspeeds of HF traders, reduces market liquidity but generates a hump-shaped relationshipto the protability of high-frequency traders and information eciency.This research finds there may be potential benefits to throttling the trading speedarms race to improve market efficiency. We also find that strategic algorithmictrading compensates for diminishments in speed advantages, providing an insighton machine behavior in the FinTech age.