High Frequency Trading Research Articles

Deep Learning Modeling of the Limit Order Book: A Comparative Perspective

The present work addresses theoretical and practical questions in the domain of Deep Learning for High Frequency Trading. State-of-the-art models such as Random models, Logistic Regressions, LSTMs, LSTMs equipped with an Attention mask, CNN-LSTMs and MLPs are reviewed and compared on the same tasks, feature space, and dataset and clustered according to pairwise similarity and performance metrics. The underlying dimensions of the modeling techniques are hence investigated to understand whether these are intrinsic to the Limit Order Book’s dynamics. We observe that the Multilayer Perceptron performs comparably to or better than state-of-the-art CNN-LSTM architectures indicating that dynamic spatial and temporal dimensions are a good approximation of the LOB’s dynamics, but not necessarily the true underlying dimensions.

Striving to make better decision quicker in cloud: big data event trading in high frequency trading perspective

Striving to make better decision quicker in cloud: big data event trading in high frequency trading perspective

Frequent Batch Auctions Under Liquidity Constraints

We exploit European regulatory interventions to investigate the effects of sub-second periodic auctions on market quality under dark trading restrictions. The restrictions are linked to an observable increase in periodic auctions and an economically meaningful loss of liquidity. While periodic auctions ameliorate illiquidity, their effects are significantly less than those of the restrictions; therefore, the combined effects of periodic auctions’ increases and the restrictions are general declines in liquidity and informational efficiency. However, consistent with theory, periodic auctions are linked to reductions in adverse selection costs, thereby underscoring their potential to address latency arbitrage and the technological arms race.

High frequency traders and the price process

Using a dataset that uniquely identifies counterparties to each S&P500 eMini transaction, we classify each market participant as high or low frequency, and each transaction, by the speed of the traders involved. We investigate empirically the comparative influence of high and low frequency traders on the price process, and conversely the influence of the price process on the trading of high and low frequency traders. We find that high frequency traders have a particularly high success rate on each transaction, measured by the likelihood that the following price change will go in their direction as well as by the amount of time they have to wait to realize their gain, when trading against low frequency traders. Contrary to common wisdom, we find that high frequency traders’ activity does not induce volatility or jumps. In fact, it is their absence that is problematic: volatility and jumps are more prevalent in periods when they trade less intensely. Conversely, we find that spikes in volatility and jumps cause high frequency traders to trade less intensely, decreasing their provision of liquidity. Finally, looking at the market microstructure noise component to the price model, we find that higher level of noise generates trading opportunities for high frequency traders and lead them to increase their trading activity.

Optimization of Backtesting Techniques in Automated High Frequency Trading Systems Using the d-Backtest PS Method

Trading strategies intended for high frequency trading in Forex markets are executed by cutting-edge automated trading systems. Such systems implement algorithmic trading strategies and are configured with predefined optimized parameters in order to generate entry and exit orders and execute trades on trading platforms. Three high-frequency automated trading systems were developed in the current research, using the MACD (oscillator), the SMA (moving average) and the PIVOT points (price crossover) technical indicators. The systems traded on hourly time frames, employing historical data of closing prices and the parameter optimization for each system was done using the d-Backtest PS method over weekly periods. With this work we intend to extend the methods of parameter selection for automated trading systems in high frequency trading. Through this research and the interpretation and evaluation of its results, we conclude that backtesting parameters’ optimization, especially through the d-Backtest PS method, is much more profitable than the default values of the parameters and that the optimization of parameters yields the highest profits through the implementation of restrictive relationships among them. It is also observed that the selection of the most profitable parameters of a trading system can be unrestricted, rendering the validation of the minor divergence occurring among slightly varying prices redundant. Meanwhile, other conclusions that can be drawn are that the most profitable classification system employed by the d-Backtest PS method is calibrated by means of two validation periods and that the most efficient profitability ratio between historical data period and validation period is 6:1 (in- and out-of-the-sample ratio).

Liquidity commonality and high frequency trading: Evidence from the French stock market

High frequency trading (HFT) depends on sophisticated algorithms to closely monitor price changes across securities. Theory predicts this technological advantage should translate into market-wide liquidity co-variation, by transmitting information-based liquidity shocks. Using a dataset of orders and trades from the French stock market, we investigate whether HFT algorithms constitute a source of systematic liquidity risk. We demonstrate that, across securities, the liquidity offered by high frequency traders is significantly less diverse than that of traditional traders; this finding is in line with the cross-asset learning hypothesis. The excessive co-movement in liquidity is also partly explained by common market making rules. In periods of increased market stress, we find HFT, designated market making, and order size to be important sources of liquidity commonality. Our results have policy implications for market regulators in Paris, suggesting the inclusion of maximum spread-limit rules in market making contracts will reduce the possibility of liquidity drying up when markets are in turmoil.

A Multifactor Regime-Switching Model for Inter-Trade Durations in the Limit Order Market

This paper studies inter-trade durations in the NASDAQ limit order market and finds that inter-trade durations in ultra-high frequency have two modes. One mode is to the order of approximately 10^{-4} seconds, and the other is to the order of 10^0 seconds. This phenomenon and other empirical evidence suggest that there are two regimes associated with the dynamics of inter-trade durations, and the regime switchings are driven by the changes of high-frequency traders (HFTs) between providing and taking liquidity. To find how the two modes depend on information in the limit order book (LOB), we propose a two-state multi-factor regime-switching (MF-RSD) model for inter-trade durations, in which the probabilities transition matrices are time-varying and depend on some lagged LOB factors. The MF-RSD model has good in-sample fitness and the superior out-of-sample performance, compared with some benchmark duration models. Our findings of the effects of LOB factors on the inter-trade durations help to understand more about the high-frequency market micro-structure.

고빈도 경쟁시장에서 시장조성자제도의 의무ㆍ보상기준 개선방향

고빈도 경쟁시장에서 시장조성자제도의 의무ㆍ보상기준 개선방향

The dynamics of ex-ante weighted spread: an empirical analysis

We model the evolution of the ex-ante weighted spread (EWS) embedded in an open Limit Order Book (LOB) and investigate the impact of observed market-related variables on the spread. Our modeling involves decomposing the joint distribution of the weighted spread into simple and interpretable distributions. Our main results have several implications: (i) EWS features high persistence in autocorrelation; (ii) lower-level LOB remains liquid even after a high trade imbalance; (iii) lower- and higher-level LOB react to temporal spread change and trade imbalance in different ways; and (iv) both trade durations and quote durations have seasonality effects. We also show, through a simple high frequency trading exercise, that the use of the model can be economically important. Further, our model provides an estimation of market resilience.

THE IMPACT OF AUTOMATED TRADING SYSTEMS ON FINANCIAL MARKET STABILITY

The way in which financial markets operate has substantially been changed by the development of information technology. Automation of trading systems in financial markets represents the last phase of depersonalizing activities previously done by traders. Algorithmic trading development enabled computers to determine the moment and the way of executing sales orders. Computers still do not make autonomous decisions regarding the choice of instruments to be traded or trading criteria. They implement the strategy a trader has decided on, choosing a favorable moment. This reduces the impact of human emotions on decision making and enables overcoming possible problems which arise due to neglecting or lack of concentration. High-frequency trading enables the execution of algorithmic operations at a high speed. The main goal of the paper is to determine advantages and dangers produced by algorithmic stock trading.

Extracting Signals from High-Frequency Trading with Digital Signal Processing Tools

As algorithms replace a growing number of tasks performed by humans in the markets, there have been growing concerns about an increased likelihood of cascading events, similar to the Flash Crash of May 6, 2010. To address these concerns, researchers have employed a number of scientific data analysis tools to monitor the risk of such cascading events. As an example, the authors of this article investigate the natural gas (NG) futures market in the frequency domain and the interaction between weather forecasts and NG price data. They observe that Fourier components with high frequencies have become more prominent in recent years and are much stronger than could be expected from an analytical model of the market. Additionally, a significant amount of trading activity occurs in the first few seconds of every minute, which is a tell-tale sign of time-based algorithmic trading. To illustrate the potential of cascading events, the authors further study how weather forecasts drive NG prices and show that, after separating the time series by season to account for the different mechanisms that relate temperature to NG price, the temperature forecast is indeed cointegrated with NG price. They also show that the variations in temperature forecasts contribute to a significant percentage of the average daily price fluctuations, which confirms the possibility that a forecast error could significantly affect the price of NG futures. TOPICS:Statistical methods, simulations, big data/machine learning Key Findings • High-frequency components in the trading data are stronger than expected from a model assuming uniform trading during market hours. • The dominance of the high-frequency components have been increasing over the years. • Relatively small changes in temperature could create a large price fluctuation in natural gas futures contracts.

Order Placement Strategies in High-Frequency Markets

I analyze order placement strategies by high-frequency traders (HFT) and non-HFT on Euronext Paris. Virtually all orders are limit orders submitted by fast traders, in most cases acting as market makers. Fast traders display high limit to market order ratios and their orders have lower ll rates, higher cancellation rates, and shorter lifetime than non-HFT orders. Liquidity taking orders are less costly for HFT than non-HFT whereas liquidity supplying orders are cheaper for non-HFT than HFT. As volatility rises, the ratio of limit order to market orders increases (decreases) for fast (slow) traders but fast traders become more conservative in their limit order submissions.

Deep LSTM with Reinforcement Learning Layer for Financial Trend Prediction in FX High Frequency Trading Systems

High-frequency trading is a method of intervention on the financial markets that uses sophisticated software tools, and sometimes also hardware, with which to implement high-frequency negotiations, guided by mathematical algorithms, that act on markets for shares, options, bonds, derivative instruments, commodities, and so on. HFT strategies have reached considerable volumes of commercial traffic, so much so that it is estimated that they are responsible for most of the transaction traffic of some stock exchanges, with percentages that, in some cases, exceed 70% of the total. One of the main issues of the HFT systems is the prediction of the medium-short term trend. For this reason, many algorithms have been proposed in literature. The author proposes in this work the use of an algorithm based both on supervised Deep Learning and on a Reinforcement Learning algorithm for forecasting the short-term trend in the currency FOREX (FOReign EXchange) market to maximize the return on investment in an HFT algorithm. With an average accuracy of about 85%, the proposed algorithm is able to predict the medium-short term trend of a currency cross based on the historical trend of this and by means of correlation data with other currency crosses using techniques known in the financial field with the term arbitrage. The final part of the proposed pipeline includes a grid trading engine which, based on the aforementioned trend predictions, will perform high frequency operations in order to maximize profit and minimize drawdown. The trading system has been validated over several financial years and on the EUR/USD cross confirming the high performance in terms of Return of Investment (98.23%) in addition to a reduced drawdown (15.97 %) which confirms its financial sustainability.

Coherent Preferences in Stock Market

In this paper, we study individual trading behaviors by cumulative trading volume distribution over a price range. We select intraday volume distribution as individual revealed preferences over a price range and determine beliefs by the maximum volume price in stock market. We propose a coherent preferences hypothesis in stock market where momentum, reversal, and interactive trading behaviors coexist in the framework of a price-volume probability wave differential equation. Then, we examine it by a set of explicit models of coherent preferences in interaction using tick by tick high frequency trading data of an ETF in Chinese stock market against a large number of the price-volume distributions. We evidence that subject individuals behave coherent preferences significantly. Finally, we find that individuals behave disposition effect or inverse disposition effect, and explain them by asymmetry and nonlinear coherent preferences. The bounded rational hypothesis combines individual preferences, beliefs, and cognitive limits, and has a set of unified and better models available in behavioral finance. It can help us to understand excessive price change or bubble and study nonlinear equilibrium and risk in financial market.

High-frequency trading and the weekly natural gas storage report

PurposeThis paper aims to investigate the quickness, and test the accuracy, of liquidity taking high-frequency traders (HFT). This gives us important insights into a class of market participant who has come to be very influential in present markets.Design/methodology/approachThe authors use the weekly natural gas (NG) storage report for the test because the information contained in the release often has a large effect on prices. Moreover, the NG market is heavily traded and liquid, and prone to high volatility. These factors make trading in this market attractive to HFT. The authors test for the profitability of those who trade in the first milliseconds after the report’s release; and for information leakage prior to the report.FindingsThe authors find those who trade within the first 50 ms accurately incorporate the information contained in the storage report into prices, and earn the majority of profits. In fact, HFT profits are decreasing in the time it takes them to trade after the announcement (measured to 200 ms). Further tests find no evidence of informed trading prior to the release of the report, and so the HFT reaction to the report incorporates the information contained therein into prices.Originality/valueThis is one of the few analyzes of the profitability of liquidity-taking HFT, and the only analysis that uses millisecond NG data. The data used is the exchanges original FIX/FAST messages.

A Descriptive Study of High-Frequency Trade and Quote Option Data

This paper provides a guide to high frequency option trade and quote data disseminated by the Options Price Reporting Authority (OPRA). We present a comprehensive overview of the U.S. option market, including details on market regulation and the trading processes for all 16 constituent option exchanges. We review the existing literature that utilizes high-frequency options data, summarize the general structure of the OPRA dataset and present a thorough empirical description of the observed option trades and quotes for a selected sample of underlying assets that contains more than 25 billion records. We outline several types of irregular observations and provide recommendations for data filtering and cleaning. Finally, we illustrate the usefulness of the high frequency option data with two empirical applications: option-implied variance estimation and risk-neutral density estimation. Both applications highlight the rich information content of the high frequency OPRA data.

The Need for Speed: Does High-Frequency Trading Make or Break Equity Markets?

Advances in technology and introduction of new regulations have transformed the equity market over last few decades, giving rise to a new breed of trading—high-frequency trading (HFT). This article investigates transformations that have shaped today’s equity trading ecology in detail with a particular focus on HFT. HFT accounts for an increasingly significant share of equity trading volumes, and is associated with lower transaction costs, greater liquidity, improved price discovery, and overall market efficiency. At times of increasing order toxicity, however, HFTs may withdraw liquidity due to higher probability of adverse selection. In this new era of equity trading, further research on a complete ecosystem of liquidity provision that ensures full provision of liquidity even during stressed periods is much needed. TOPICS:Fundamental equity analysis, accounting and ratio analysis, technical analysis

High-Frequency Trading Activities and Brokerage Firms Effect : Empirical Evidence From the Indonesia Stock Exchange

This research studies the trading activity of type of traders through their brokers. Order imbalance is believed to be a better proxy for explaining trading activity. This paper presents some empirical test that on brokerage level analysis exhibit information paradigm in Indonesia which market makers and specialist are not available. We divide imbalances into groups of samples (all stocks and most liquid stocks), trader type (foreign or domestic) and size of brokerage firm (small to big). Our results show that order imbalances generally have a positive serial correlation for all the traders and brokers analyzed. However, we find that the determinant of order imbalances is a particular phenomenon at the brokerage level, whose results differ from our market-wide analysis. We do not find that previous order imbalances can predict market returns across trader type and brokerage class. In contrast, for the inventory paradigm, the evidence from the brokerage level analysis indicates that information dissemination is induced order imbalance by brokerage house.

Effectiveness of filter trading as an intraday trading rule

PurposeFilter trading is a technical trading rule that has been used extensively to test the efficient market hypothesis in the context of long-term trading. In this paper, the authors adopt the rule to analyze intraday trading, in which an open position is not left overnight. This paper aims to explore the relationship between intraday filter trading profitability and intraday realized volatilities. The bivariate thin plate spline (TPS) model is chosen to fit the predictor-response surface for high frequency data from the Hang Seng index futures (HSIF) market. The hypotheses follow the adaptive market hypothesis, arguing that intraday filter trading differs in profitability under different market conditions as measured by realized volatility, and furthermore, the optimal filter size for trading on each day is related to the realized volatility. The empirical results furnish new evidence that range-based realized volatilities (RaV) are more efficient in identifying trading profit than return-based volatilities (ReV). These results shed light on the efficiency of intraday high frequency trading in the HSIF market. Some trading suggestions are given based on the findings.Design/methodology/approachAmong all the factors that affect the profit of filter trading, intraday realized volatility stands out as an important predictor. The authors explore several intraday volatilities measures using range-based or return-based methods of estimation. The authors then study how the filter trading profit will depend on realized volatility and how the optimal filter size is related to the realized volatility. The bivariate TPS model is used to model the predictor-response relationship.FindingsThe empirical results show that range-based realized volatility has a higher predictive power on filter rule trading profit than the return-based realized volatility.Originality/valueFirst, the authors contribute to the literature by investigating the profitability of the filter trading rule on high frequency tick-by-tick data of HSIF market. Second, the authors test the assumption that the magnitude of the intraday momentum trading profit depends on the realized volatilities and aims to identify a relationship between them. Furthermore, the authors consider several intraday realized volatilities and find the RaV have the higher prediction power than ReV. Finally, the authors find some relationship between the optimal filter size and the realized volatilities. Based on the observations, the authors also give some trading suggestions to the intraday filter traders.

High-Frequency Trading with Fractional Brownian Motion

In the high-frequency limit, conditional expected increments of fractional Brownian motion converge to a white noise, shedding their dependence on the path history and the forecasting horizon, and making dynamic optimization problems tractable. We find an explicit formula for locally mean-variance optimal strategies and their performance for an asset price that follows fractional Brownian motion. Without trading costs, risk-adjusted profits are linear in the trading horizon and rise asymmetrically as the Hurst exponent departs from Brownian motion, remaining finite as the exponent reaches zero while diverging as it approaches one. Trading costs penalize numerous portfolio updates from short-lived signals, leading to a finite trading frequency, which can be chosen so that the effect of trading costs is arbitrarily small, depending on the required speed of convergence to the high-frequency limit.