Large language models (LLMs) have demonstrated remarkable proficiency in understanding and generating human-like texts, which may potentially revolutionize the finance industry. However, existing LLMs often fall short in the financial field, which is mainly attributed to the disparities between general text data and financial text data. Unfortunately, there is only a limited number of financial text datasets available, and BloombergGPT, the first financial LLM (FinLLM), is close-sourced (only the training logs were released). In light of this, we aim to democratize Internet-scale financial data for LLMs, which is an open challenge due to diverse data sources, low signal-to-noise ratio, and high time-validity. To address the challenges, we introduce an open-sourced and data-centric framework, Financial Generative Pre-trained Transformer (FinGPT), that automates the collection and curation of real-time financial data from 34 diverse sources on the Internet, providing researchers and practitioners with accessible and transparent resources to develop their FinLLMs. Additionally, we propose a simple yet effective strategy for fine-tuning FinLLM using the inherent feedback from the market, dubbed Reinforcement Learning with Stock Prices (RLSP). We also adopt the Low-rank Adaptation (LoRA, QLoRA) method that enables users to customize their own FinLLMs from general-purpose LLMs at a low cost. Finally, we showcase several FinGPT applications, including robo-advisor, sentiment analysis for algorithmic trading, and low-code development. FinGPT aims to democratize FinLLMs, stimulate innovation, and unlock new opportunities in open finance. The codes have been open-sourced.
The use of machine learning in algorithmic trading systems is increasingly common. In a typical set-up, supervised learning is used to predict the future prices of assets, and those predictions drive a simple trading and execution strategy. This is quite effective when the predictions have sufficient signal, markets are liquid, and transaction costs are low. However, those conditions often do not hold in thinly traded financial markets and markets for differentiated assets such as real estate or vehicles. In these markets, the trading strategy must consider the long-term effects of taking positions that are relatively more difficult to change. In this work, we propose a Reinforcement Learning (RL) algorithm that trades based on signals from a learned predictive model and addresses these challenges. We test our algorithm on 20+ years of equity data from Bursa Malaysia.
This paper investigates the temporal patterns of activity in the cryptocurrency market with a focus on Bitcoin, Ethereum, Dogecoin, and WINkLink from January 2020 to December 2022. Market activity measures - logarithmic returns, volume, and transaction number, sampled every 10 seconds, were divided into intraday and intraweek periods and then further decomposed into recurring and noise components via correlation matrix formalism. The key findings include the distinctive market behavior from traditional stock markets due to the nonexistence of trade opening and closing. This was manifest in three enhanced-activity phases aligning with Asian, European, and U.S. trading sessions. An intriguing pattern of activity surge in 15-minute intervals, particularly at full hours, was also noticed, implying the potential role of algorithmic trading. Most notably, recurring bursts of activity in bitcoin and ether were identified to coincide with the release times of significant U.S. macroeconomic reports such as Nonfarm payrolls, Consumer Price Index data, and Federal Reserve statements. The most correlated daily patterns of activity occurred in 2022, possibly reflecting the documented correlations with U.S. stock indices in the same period. Factors that are external to the inner market dynamics are found to be responsible for the repeatable components of the market dynamics, while the internal factors appear to be substantially random, which manifests itself in a good agreement between the empirical eigenvalue distributions in their bulk and the random matrix theory predictions expressed by the Marchenko-Pastur distribution. The findings reported support the growing integration of cryptocurrencies into the global financial markets.
Large language models (LLMs) have shown the potential of revolutionizing natural language processing tasks in diverse domains, sparking great interest in finance. Accessing high-quality financial data is the first challenge for financial LLMs (FinLLMs). While proprietary models like BloombergGPT have taken advantage of their unique data accumulation, such privileged access calls for an open-source alternative to democratize Internet-scale financial data. In this paper, we present an open-source large language model, FinGPT, for the finance sector. Unlike proprietary models, FinGPT takes a data-centric approach, providing researchers and practitioners with accessible and transparent resources to develop their FinLLMs. We highlight the importance of an automatic data curation pipeline and the lightweight low-rank adaptation technique in building FinGPT. Furthermore, we showcase several potential applications as stepping stones for users, such as robo-advising, algorithmic trading, and low-code development. Through collaborative efforts within the open-source AI4Finance community, FinGPT aims to stimulate innovation, democratize FinLLMs, and unlock new opportunities in open finance. Two associated code repos are \url{https://github.com/AI4Finance-Foundation/FinGPT} and \url{https://github.com/AI4Finance-Foundation/FinNLP}
With the fast development of quantitative portfolio optimization in financial engineering, lots of AI-based algorithmic trading strategies have demonstrated promising results, among which reinforcement learning begins to manifest competitive advantages. However, the environment from real financial markets is complex and hard to be fully simulated, considering the observation of abrupt transitions, unpredictable hidden causal factors, heavy tail properties and so on. Thus, in this paper, first, we adopt a heavy-tailed preserving normalizing flows to simulate high-dimensional joint probability of the complex trading environment and develop a model-based reinforcement learning framework to better understand the intrinsic mechanisms of quantitative online trading. Second, we experiment with various stocks from three different financial markets (Dow, NASDAQ and S&P) and show that among these three financial markets, Dow gets the best performance based on various evaluation metrics under our back-testing system. Especially, our proposed method is able to mitigate the impact of unpredictable financial market crises during the COVID-19 pandemic period, resulting in a lower maximum drawdown. Third, we also explore the explanation of our RL algorithm. (1), we utilize the pattern causality method to study the interactive relation among different stocks in the environment. (2), We analyze the dynamic loss and actor loss to ensure the convergence of our strategies. (3), by visualizing high dimensional state transition data comparisons from real and virtual buffers with t-SNE, we uncover some effective patterns of better portfolio optimization strategies. (4), we also utilize eigenvalue analysis to study the convergence properties of the environmen's model.
Volume prediction is one of the fundamental objectives in the Fintech area, which is helpful for many downstream tasks, e.g., algorithmic trading. Previous methods mostly learn a universal model for different stocks. However, this kind of practice omits the specific characteristics of individual stocks by applying the same set of parameters for different stocks. On the other hand, learning different models for each stock would face data sparsity or cold start problems for many stocks with small capitalization. To take advantage of the data scale and the various characteristics of individual stocks, we propose a dual-process meta-learning method that treats the prediction of each stock as one task under the meta-learning framework. Our method can model the common pattern behind different stocks with a meta-learner, while modeling the specific pattern for each stock across time spans with stock-dependent parameters. Furthermore, we propose to mine the pattern of each stock in the form of a latent variable which is then used for learning the parameters for the prediction module. This makes the prediction procedure aware of the data pattern. Extensive experiments on volume predictions show that our method can improve the performance of various baseline models. Further analyses testify the effectiveness of our proposed meta-learning framework.
Price movement prediction has always been one of the traders' concerns in financial market trading. In order to increase their profit, they can analyze the historical data and predict the price movement. The large size of the data and complex relations between them lead us to use algorithmic trading and artificial intelligence. This paper aims to offer an approach using Twin-Delayed DDPG (TD3) and the daily close price in order to achieve a trading strategy in the stock and cryptocurrency markets. Unlike previous studies using a discrete action space reinforcement learning algorithm, the TD3 is continuous, offering both position and the number of trading shares. Both the stock (Amazon) and cryptocurrency (Bitcoin) markets are addressed in this research to evaluate the performance of the proposed algorithm. The achieved strategy using the TD3 is compared with some algorithms using technical analysis, reinforcement learning, stochastic, and deterministic strategies through two standard metrics, Return and Sharpe ratio. The results indicate that employing both position and the number of trading shares can improve the performance of a trading system based on the mentioned metrics.
Within the mathematical finance literature there is a rich catalogue of mathematical models for studying algorithmic trading problems -- such as market-making and optimal execution -- in limit order books. This paper introduces \mbtgym, a Python module that provides a suite of gym environments for training reinforcement learning (RL) agents to solve such model-based trading problems. The module is set up in an extensible way to allow the combination of different aspects of different models. It supports highly efficient implementations of vectorized environments to allow faster training of RL agents. In this paper, we motivate the challenge of using RL to solve such model-based limit order book problems in mathematical finance, we explain the design of our gym environment, and then demonstrate its use in solving standard and non-standard problems from the literature. Finally, we lay out a roadmap for further development of our module, which we provide as an open source repository on GitHub so that it can serve as a focal point for RL research in model-based algorithmic trading.
Designing profitable and reliable trading strategies is challenging in the highly volatile cryptocurrency market. Existing works applied deep reinforcement learning methods and optimistically reported increased profits in backtesting, which may suffer from the false positive issue due to overfitting. In this paper, we propose a practical approach to address backtest overfitting for cryptocurrency trading using deep reinforcement learning. First, we formulate the detection of backtest overfitting as a hypothesis test. Then, we train the DRL agents, estimate the probability of overfitting, and reject the overfitted agents, increasing the chance of good trading performance. Finally, on 10 cryptocurrencies over a testing period from 05/01/2022 to 06/27/2022 (during which the crypto market crashed two times), we show that the less overfitted deep reinforcement learning agents have a higher return than that of more overfitted agents, an equal weight strategy, and the S&P DBM Index (market benchmark), offering confidence in possible deployment to a real market.
This scientific paper propose a novel portfolio optimization model using an improved deep reinforcement learning algorithm. The objective function of the optimization model is the weighted sum of the expectation and value at risk(VaR) of portfolio cumulative return. The proposed algorithm is based on actor-critic architecture, in which the main task of critical network is to learn the distribution of portfolio cumulative return using quantile regression, and actor network outputs the optimal portfolio weight by maximizing the objective function mentioned above. Meanwhile, we exploit a linear transformation function to realize asset short selling. Finally, A multi-process method is used, called Ape-x, to accelerate the speed of deep reinforcement learning training. To validate our proposed approach, we conduct backtesting for two representative portfolios and observe that the proposed model in this work is superior to the benchmark strategies.
In this article, we develop a modular framework for the application of Reinforcement Learning to the problem of Optimal Trade Execution. The framework is designed with flexibility in mind, in order to ease the implementation of different simulation setups. Rather than focusing on agents and optimization methods, we focus on the environment and break down the necessary requirements to simulate an Optimal Trade Execution under a Reinforcement Learning framework such as data pre-processing, construction of observations, action processing, child order execution, simulation of benchmarks, reward calculations etc. We give examples of each component, explore the difficulties their individual implementations \& the interactions between them entail, and discuss the different phenomena that each component induces in the simulation, highlighting the divergences between the simulation and the behavior of a real market. We showcase our modular implementation through a setup that, following a Time-Weighted Average Price (TWAP) order submission schedule, allows the agent to exclusively place limit orders, simulates their execution via iterating over snapshots of the Limit Order Book (LOB), and calculates rewards as the \$ improvement over the price achieved by a TWAP benchmark algorithm following the same schedule. We also develop evaluation procedures that incorporate iterative re-training and evaluation of a given agent over intervals of a training horizon, mimicking how an agent may behave when being continuously retrained as new market data becomes available and emulating the monitoring practices that algorithm providers are bound to perform under current regulatory frameworks.
Optimal execution is a sequential decision-making problem for cost-saving in algorithmic trading. Studies have found that reinforcement learning (RL) can help decide the order-splitting sizes. However, a problem remains unsolved: how to place limit orders at appropriate limit prices? The key challenge lies in the "continuous-discrete duality" of the action space. On the one hand, the continuous action space using percentage changes in prices is preferred for generalization. On the other hand, the trader eventually needs to choose limit prices discretely due to the existence of the tick size, which requires specialization for every single stock with different characteristics (e.g., the liquidity and the price range). So we need continuous control for generalization and discrete control for specialization. To this end, we propose a hybrid RL method to combine the advantages of both of them. We first use a continuous control agent to scope an action subset, then deploy a fine-grained agent to choose a specific limit price. Extensive experiments show that our method has higher sample efficiency and better training stability than existing RL algorithms and significantly outperforms previous learning-based methods for order execution.
This article introduces an intrinsic entropy model that can be used as an indicator to gauge investor interest in a given exchange-traded security, along with the state of the general market corroborated by individual security trade data. Although the syntagma of intrinsic entropy might sound somehow pleonastic, since entropy itself characterizes the fundamentals of a system, we would like to make a clear distinction between entropy models based on the values that a random variable may take and the model that we propose, which employs actual stock exchange trading data. The model we propose for intrinsic entropy does not include any exogenous factor that could influence the level of entropy. The intrinsic entropy signals whether the market is inclined to buy the security or rather to sell it. We further explore the usage of the intrinsic entropy model for algorithmic trading, in order to demonstrate the value of our model in assisting investors in the selection of the intraday stock portfolio, along with timely generated signals to support the buy / sell decision making process. The test results provide empirical evidence that the proposed intrinsic entropy model can be used as an indicator to evaluate the direction and intensity of intraday trading activity of an exchange-traded security. The data used for the test consisted of historical intraday transactions executed on The Bucharest Stock Exchange (BVB).
We study a differential Riccati equation (DRE) with indefinite matrix coefficients, which arises in a wide class of practical problems. We show that the DRE solves an associated control problem, which is key to provide existence and uniqueness of a solution. As an application, we solve two algorithmic trading problems in which the agent adopts a constant absolute risk-aversion (CARA) utility function, and where the optimal strategies use signals and past observations of prices to improve their performance. First, we derive a multi-asset market making strategy in over-the-counter markets, where the market maker uses an external trading venue to hedge risk. Second, we derive an optimal trading strategy that uses prices and signals to learn the drift in the asset prices.
Reinforcement learning (RL) techniques have shown great success in many challenging quantitative trading tasks, such as portfolio management and algorithmic trading. Especially, intraday trading is one of the most profitable and risky tasks because of the intraday behaviors of the financial market that reflect billions of rapidly fluctuating capitals. However, a vast majority of existing RL methods focus on the relatively low frequency trading scenarios (e.g., day-level) and fail to capture the fleeting intraday investment opportunities due to two major challenges: 1) how to effectively train profitable RL agents for intraday investment decision-making, which involves high-dimensional fine-grained action space; 2) how to learn meaningful multi-modality market representation to understand the intraday behaviors of the financial market at tick-level. Motivated by the efficient workflow of professional human intraday traders, we propose DeepScalper, a deep reinforcement learning framework for intraday trading to tackle the above challenges. Specifically, DeepScalper includes four components: 1) a dueling Q-network with action branching to deal with the large action space of intraday trading for efficient RL optimization; 2) a novel reward function with a hindsight bonus to encourage RL agents making trading decisions with a long-term horizon of the entire trading day; 3) an encoder-decoder architecture to learn multi-modality temporal market embedding, which incorporates both macro-level and micro-level market information; 4) a risk-aware auxiliary task to maintain a striking balance between maximizing profit and minimizing risk. Through extensive experiments on real-world market data spanning over three years on six financial futures, we demonstrate that DeepScalper significantly outperforms many state-of-the-art baselines in terms of four financial criteria.
The feasibility of making profitable trades on a single asset on stock exchanges based on patterns identification has long attracted researchers. Reinforcement Learning (RL) and Natural Language Processing have gained notoriety in these single-asset trading tasks, but only a few works have explored their combination. Moreover, some issues are still not addressed, such as extracting market sentiment momentum through the explicit capture of sentiment features that reflect the market condition over time and assessing the consistency and stability of RL results in different situations. Filling this gap, we propose the Sentiment-Aware RL (SentARL) intelligent trading system that improves profit stability by leveraging market mood through an adaptive amount of past sentiment features drawn from textual news. We evaluated SentARL across twenty assets, two transaction costs, and five different periods and initializations to show its consistent effectiveness against baselines. Subsequently, this thorough assessment allowed us to identify the boundary between news coverage and market sentiment regarding the correlation of price-time series above which SentARL's effectiveness is outstanding.
Bitcoin is firmly becoming a mainstream asset in our global society. Its highly volatile nature has traders and speculators flooding into the market to take advantage of its significant price swings in the hope of making money. This work brings an algorithmic trading approach to the Bitcoin market to exploit the variability in its price on a day-to-day basis through the classification of its direction. Building on previous work, in this paper, we utilise both features internal to the Bitcoin network and external features to inform the prediction of various machine learning models. As an empirical test of our models, we evaluate them using a real-world trading strategy on completely unseen data collected throughout the first quarter of 2021. Using only a binary predictor, at the end of our three-month trading period, our models showed an average profit of 86\%, matching the results of the more traditional buy-and-hold strategy. However, after incorporating a risk tolerance score into our trading strategy by utilising the model's prediction confidence scores, our models were 12.5\% more profitable than the simple buy-and-hold strategy. These results indicate the credible potential that machine learning models have in extracting profit from the Bitcoin market and act as a front-runner for further research into real-world Bitcoin trading.
We study recurrent patterns in volatility and volume for major cryptocurrencies, Bitcoin and Ether, using data from two centralized exchanges (Coinbase Pro and Binance) and a decentralized exchange (Uniswap V2). We find systematic patterns in both volatility and volume across day-of-the-week, hour-of-the-day, and within the hour. These patterns have grown stronger over the years and can be related to algorithmic trading and funding times in futures markets. We also document that price formation mainly takes place on the centralized exchanges while price adjustments on the decentralized exchanges can be sluggish.
While a substantial literature on structural break change point analysis exists for univariate time series, research on large panel data models has not been as extensive. In this paper, a novel method for estimating panel models with multiple structural changes is proposed. The breaks are allowed to occur at unknown points in time and may affect the multivariate slope parameters individually. Our method adapts Haar wavelets to the structure of the observed variables in order to detect the change points of the parameters consistently. We also develop methods to address endogenous regressors within our modeling framework. The asymptotic property of our estimator is established. In our application, we examine the impact of algorithmic trading on standard measures of market quality such as liquidity and volatility over a time period that covers the financial meltdown that began in 2007. We are able to detect jumps in regression slope parameters automatically without using ad-hoc subsample selection criteria.
Modern evolvements of the technologies have been leading to a profound influence on the financial market. The introduction of constituents like Exchange-Traded Funds, and the wide-use of advanced technologies such as algorithmic trading, results in a boom of the data which provides more opportunities to reveal deeper insights. However, traditional statistical methods always suffer from the high-dimensional, high-correlation, and time-varying instinct of the financial data. In this dissertation, we focus on developing techniques to stress these difficulties. With the proposed methodologies, we can have more interpretable models, clearer explanations, and better predictions.