This study presents an innovative approach for predicting cryptocurrency time series, specifically focusing on Bitcoin, Ethereum, and Litecoin. The methodology integrates the use of technical indicators, a Performer neural network, and BiLSTM (Bidirectional Long Short-Term Memory) to capture temporal dynamics and extract significant features from raw cryptocurrency data. The application of technical indicators, such facilitates the extraction of intricate patterns, momentum, volatility, and trends. The Performer neural network, employing Fast Attention Via positive Orthogonal Random features (FAVOR+), has demonstrated superior computational efficiency and scalability compared to the traditional Multi-head attention mechanism in Transformer models. Additionally, the integration of BiLSTM in the feedforward network enhances the model's capacity to capture temporal dynamics in the data, processing it in both forward and backward directions. This is particularly advantageous for time series data where past and future data points can influence the current state. The proposed method has been applied to the hourly and daily timeframes of the major cryptocurrencies and its performance has been benchmarked against other methods documented in the literature. The results underscore the potential of the proposed method to outperform existing models, marking a significant progression in the field of cryptocurrency price prediction.
Johann Lussange, Stefano Vrizzi, Stefano Palminteri, Boris Gutkin
Building on a previous foundation work (Lussange et al. 2020), this study introduces a multi-agent reinforcement learning (MARL) model simulating crypto markets, which is calibrated to the Binance's daily closing prices of $153$ cryptocurrencies that were continuously traded between 2018 and 2022. Unlike previous agent-based models (ABM) or multi-agent systems (MAS) which relied on zero-intelligence agents or single autonomous agent methodologies, our approach relies on endowing agents with reinforcement learning (RL) techniques in order to model crypto markets. This integration is designed to emulate, with a bottom-up approach to complexity inference, both individual and collective agents, ensuring robustness in the recent volatile conditions of such markets and during the COVID-19 era. A key feature of our model also lies in the fact that its autonomous agents perform asset price valuation based on two sources of information: the market prices themselves, and the approximation of the crypto assets fundamental values beyond what those market prices are. Our MAS calibration against real market data allows for an accurate emulation of crypto markets microstructure and probing key market behaviors, in both the bearish and bullish regimes of that particular time period.
In the burgeoning realm of cryptocurrency, social media platforms like Twitter have become pivotal in influencing market trends and investor sentiments. In our study, we leverage GPT-4 and a fine-tuned transformer-based BERT model for a multimodal sentiment analysis, focusing on the impact of emoji sentiment on cryptocurrency markets. By translating emojis into quantifiable sentiment data, we correlate these insights with key market indicators like BTC Price and the VCRIX index. Our architecture's analysis of emoji sentiment demonstrated a distinct advantage over FinBERT's pure text sentiment analysis in such predicting power. This approach may be fed into the development of trading strategies aimed at utilizing social media elements to identify and forecast market trends. Crucially, our findings suggest that strategies based on emoji sentiment can facilitate the avoidance of significant market downturns and contribute to the stabilization of returns. This research underscores the practical benefits of integrating advanced AI-driven analyses into financial strategies, offering a nuanced perspective on the interplay between digital communication and market dynamics in an academic context.
An extension of the Hawkes process, the Marked Hawkes process distinguishes itself by featuring variable jump size across each event, in contrast to the constant jump size observed in a Hawkes process without marks. While extensive literature has been dedicated to the non-parametric estimation of both the linear and non-linear Hawkes process, there remains a significant gap in the literature regarding the marked Hawkes process. In response to this, we propose a methodology for estimating the conditional intensity of the marked Hawkes process. We introduce two distinct models: \textit{Shallow Neural Hawkes with marks}- for Hawkes processes with excitatory kernels and \textit{Neural Network for Non-Linear Hawkes with Marks}- for non-linear Hawkes processes. Both these approaches take the past arrival times and their corresponding marks as the input to obtain the arrival intensity. This approach is entirely non-parametric, preserving the interpretability associated with the marked Hawkes process. To validate the efficacy of our method, we subject the method to synthetic datasets with known ground truth. Additionally, we apply our method to model cryptocurrency order book data, demonstrating its applicability to real-world scenarios.
Concentrated liquidity (CL) provisioning is a way how to improve the capital efficiency of Automated Market Makers (AMM). Allowing liquidity providers to use leverage is a step towards even higher capital efficiency. A number of Decentralized Finance (DeFi) protocols implement this technique in conjunction with overcollateralized lending. However, the properties of leveraged CL positions have not been formalized and are poorly understood in practice. This article describes the principles of a leveraged CL provisioning protocol, formally models the notions of margin level, assets, and debt, and proves that within this model, leveraged LP positions possess several properties that make them safe to use.
Decentralized Finance (DeFi) has reshaped the possibilities of reserve banking in the form of the Collateralized Debt Position (CDP). Key to the safety of CDPs is the money supply architecture that enables issued debt to maintain its value. In traditional markets, and with respect to the United States Dollar system, interest rates are set by the Federal Reserve in an attempt to influence the effects of excessive inflation. DeFi enables a more transparent approach that typically relies on interest rates or other debt recovery mechanisms being directly informed by asset price. This research investigates contemporary DeFi money supply and debt management strategies and their limitations. Furthermore, this paper introduces a time-weighted approach to interest rate management that implements a Proportional-Integral-Derivative control system to constantly adapt to market activities and protect the value of issued currency, while addressing observed limitations.
Decentralized Finance (DeFi) money markets have seen explosive growth in recent years, with billions of dollars borrowed in various cryptocurrency assets. Key to the safety of money markets is the implementation of interest rates that determine the cost of borrowing, and govern counterparty exposure and return. In traditional markets, interest rates are set by risk managers, portfolio managers, the Federal Reserve, and a myriad of other sources depending on the market function. DeFi enables an algorithmic approach that typically relies on interest rates being directly dependent on market utilization. The benefit of algorithmic interest rate management is the system's continual response to market behaviors in real time, and thus an inherent ability to mitigate risks on behalf of protocols and users. These interest rate strategies target an optimal utilization based on the protocol's risk threshold, but historically lack the ability to compensate for excessive or diminished utilization over time. This research investigates contemporary DeFi interest rate management strategies and their limitations. Furthermore, this paper introduces a time-weighted approach to interest rate management that implements a Proportional-Integral-Derivative (PID) control system to constantly adapt to market utilization patterns, addressing observed limitations.
Cryptocurrencies have emerged as a novel financial asset garnering significant attention in recent years. A defining characteristic of these digital currencies is their pronounced short-term market volatility, primarily influenced by widespread sentiment polarization, particularly on social media platforms such as Twitter. Recent research has underscored the correlation between sentiment expressed in various networks and the price dynamics of cryptocurrencies. This study delves into the 15-minute impact of informative tweets disseminated through foundation channels on trader behavior, with a focus on potential outcomes related to sentiment polarization. The primary objective is to identify factors that can predict positive price movements and potentially be leveraged through a trading algorithm. To accomplish this objective, we conduct a conditional examination of return and excess return rates within the 15 minutes following tweet publication. The empirical findings reveal statistically significant increases in return rates, particularly within the initial three minutes following tweet publication. Notably, adverse effects resulting from the messages were not observed. Surprisingly, sentiments were found to have no discerni-ble impact on cryptocurrency price movements. Our analysis further identifies that inves-tors are primarily influenced by the quality of tweet content, as reflected in the choice of words and tweet volume. While the basic trading algorithm presented in this study does yield some benefits within the 15-minute timeframe, these benefits are not statistically significant. Nevertheless, it serves as a foundational framework for potential enhance-ments and further investigations.
I study the price dynamics of non-fungible tokens (NFTs) and propose a deep learning framework for dynamic valuation of NFTs. I use data from the Ethereum blockchain and OpenSea to train a deep learning model on historical trades, market trends, and traits/rarity features of Bored Ape Yacht Club NFTs. After hyperparameter tuning, the model is able to predict the price of NFTs with high accuracy. I propose an application framework for this model using zero-knowledge machine learning (zkML) and discuss its potential use cases in the context of decentralized finance (DeFi) applications.
Xihan Xiong, Zhipeng Wang, Tianxiang Cui, William J. Knottenbelt · 5 authors
Technological advancement drives financial innovation, reshaping the traditional finance landscape and redefining user-market interactions. The rise of blockchain and Decentralized Finance (DeFi) underscores this intertwined evolution of technology and finance. While DeFi has introduced exciting opportunities, it has also exposed the ecosystem to new forms of market misconduct. This paper aims to bridge the academic and regulatory gaps by addressing key research questions about market misconduct in DeFi. We begin by discussing how blockchain technology can potentially enable the emergence of novel forms of market misconduct. We then offer a comprehensive definition and taxonomy for understanding DeFi market misconduct. Through comparative analysis and empirical measurements, we examine the novel forms of misconduct in DeFi, shedding light on their characteristics and social impact. Subsequently, we investigate the challenges of building a tailored regulatory framework for DeFi. We identify key areas where existing regulatory frameworks may need enhancement. Finally, we discuss potential approaches that bring DeFi into the regulatory perimeter.
Paweł Szydło, Marcin Wątorek, Jarosław Kwapień, Stanisław Drożdż
A non-fungible token (NFT) market is a new trading invention based on the blockchain technology, which parallels the cryptocurrency market. In the present work, we study capitalization, floor price, the number of transactions, the inter-transaction times, and the transaction volume value of a few selected popular token collections. The results show that the fluctuations of all these quantities are characterized by heavy-tailed probability distribution functions, in most cases well described by the stretched exponentials, with a trace of power-law scaling at times, long-range memory, persistence, and in several cases even the fractal organization of fluctuations, mostly restricted to the larger fluctuations, however. We conclude that the NFT market-even though young and governed by somewhat different mechanisms of trading-shares several statistical properties with the regular financial markets. However, some differences are visible in the specific quantitative indicators.
Jakub Michańków, Łukasz Kwiatkowski, Janusz Morajda
In this paper, we develop a hybrid approach to forecasting the volatility and risk of financial instruments by combining common econometric GARCH time series models with deep learning neural networks. For the latter, we employ Gated Recurrent Unit (GRU) networks, whereas four different specifications are used as the GARCH component: standard GARCH, EGARCH, GJR-GARCH and APARCH. Models are tested using daily logarithmic returns on the S&P 500 index as well as gold price Bitcoin prices, with the three assets representing quite distinct volatility dynamics. As the main volatility estimator, also underlying the target function of our hybrid models, we use the price-range-based Garman-Klass estimator, modified to incorporate the opening and closing prices. Volatility forecasts resulting from the hybrid models are employed to evaluate the assets' risk using the Value-at-Risk (VaR) and Expected Shortfall (ES) at two different tolerance levels of 5% and 1%. Gains from combining the GARCH and GRU approaches are discussed in the contexts of both the volatility and risk forecasts. In general, it can be concluded that the hybrid solutions produce more accurate point volatility forecasts, although it does not necessarily translate into superior VaR and ES forecasts.
This paper proposes a novel approach to hedging portfolios of risky assets when financial markets are affected by financial turmoils. We introduce a completely novel approach to diversification activity not on the level of single assets but on the level of ensemble algorithmic investment strategies (AIS) built based on the prices of these assets. We employ four types of diverse theoretical models (LSTM - Long Short-Term Memory, ARIMA-GARCH - Autoregressive Integrated Moving Average - Generalized Autoregressive Conditional Heteroskedasticity, momentum, and contrarian) to generate price forecasts, which are then used to produce investment signals in single and complex AIS. In such a way, we are able to verify the diversification potential of different types of investment strategies consisting of various assets (energy commodities, precious metals, cryptocurrencies, or soft commodities) in hedging ensemble AIS built for equity indices (S&P 500 index). Empirical data used in this study cover the period between 2004 and 2022. Our main conclusion is that LSTM-based strategies outperform the other models and that the best diversifier for the AIS built for the S&P 500 index is the AIS built for Bitcoin. Finally, we test the LSTM model for a higher frequency of data (1 hour). We conclude that it outperforms the results obtained using daily data.
Maximal Extractable Value (MEV) is value extractable by temporary monopoly power commonly found in decentralized systems. This extraction stems from a lack of user privacy upon transaction submission and the ability of a monopolist validator to reorder, add, and/or censor transactions. There are two main directions to reduce MEV: reduce the flexibility of the miner to reorder transactions by enforcing ordering rules and/or introduce a competitive market for the right to reorder, add, and/or censor transactions. In this work, we unify these approaches via \emph{uncertainty principles}, akin to those found in harmonic analysis and physics. This provides a quantitative trade-off between the freedom to reorder transactions and the complexity of an economic payoff to a user in a decentralized network. This trade off is analogous to the Nyquist-Shannon sampling theorem and demonstrates that sequencing rules in blockchains need to be application specific. Our results suggest that neither so-called fair ordering techniques nor economic mechanisms can individually mitigate MEV for arbitrary payoff functions.
This paper investigates the issue of an adequate loss function in the optimization of machine learning models used in the forecasting of financial time series for the purpose of algorithmic investment strategies (AIS) construction. We propose the Mean Absolute Directional Loss (MADL) function, solving important problems of classical forecast error functions in extracting information from forecasts to create efficient buy/sell signals in algorithmic investment strategies. Finally, based on the data from two different asset classes (cryptocurrencies: Bitcoin and commodities: Crude Oil), we show that the new loss function enables us to select better hyperparameters for the LSTM model and obtain more efficient investment strategies, with regard to risk-adjusted return metrics on the out-of-sample data.
We consider the problem of being a cross-chain wealth management platform with deposits, redemptions and investment assets across multiple networks. We discuss the need for blockchain bridges to facilitates fund flows across platforms. We point out several issues with existing bridges. We develop an algorithm - tailored to overcome current constraints - that dynamically changes the utilization of bridge capacities and hence the amounts to be transferred across networks. We illustrate several scenarios using numerical simulations.
We take inspiration from statistical physics to develop a novel conceptual framework for the analysis of financial markets. We model the order book dynamics as a motion of particles and define the momentum measure of the system as a way to summarise and assess the state of the market. Our approach proves useful in capturing salient financial market phenomena: in particular, it helps detect the market manipulation activities called spoofing and layering. We apply our method to identify pathological order book behaviours during the flash crash of the LUNA cryptocurrency, uncovering widespread instances of spoofing and layering in the market. Furthermore, we establish that our technique outperforms the conventional Z-score-based anomaly detection method in identifying market manipulations across both LUNA and Bitcoin cryptocurrency markets.
Abstract The growing attention on cryptocurrencies has led to increasing research on digital stock markets. Approaches and tools usually applied to characterize standard stocks have been applied to the digital ones. Among these tools is the identification of processes of market fluctuations. Being interesting stochastic processes, the usual statistical methods are appropriate tools for their reconstruction. There, besides chance, the description of a behavioural component shall be present whenever a deterministic pattern is ever found. Markov approaches are at the leading edge of this endeavour. In this paper, Markov chains of orders one to eight are considered as a way to forecast the dynamics of three major cryptocurrencies. It is accomplished using an empirical basis of intra-day returns. Besides forecasting, we investigate the existence of eventual long-memory components in each of those stochastic processes. Results show that predictions obtained from using the empirical probabilities are better than random choices.
Identifying the structural dependence between the cryptocurrencies and predicting market trend are fundamental for effective portfolio management in cryptocurrency trading. In this paper, we present a unified Bayesian framework based on potential field theory and Gaussian Process to characterize the structural dependency of various cryptocurrencies, using historic price information. The following are our significant contributions: (i) Proposed a novel model for cryptocurrency price movements as a trajectory of a dynamical system governed by a time-varying non-linear potential field. (ii) Validated the existence of the non-linear potential function in cryptocurrency market through Lyapunov stability analysis. (iii) Developed a Bayesian framework for inferring the non-linear potential function from observed cryptocurrency prices. (iv) Proposed that attractors and repellers inferred from the potential field are reliable cryptocurrency market indicators, surpassing existing attributes, such as, mean, open price or close price of an observation window, in the literature. (v) Analysis of cryptocurrency market during various Bitcoin crash durations from April 2017 to November 2021, shows that attractors captured the market trend, volatility, and correlation. In addition, attractors aids explainability and visualization. (vi) The structural dependence inferred by the proposed approach was found to be consistent with results obtained using the popular wavelet coherence approach. (vii) The proposed market indicators (attractors and repellers) can be used to improve the prediction performance of state-of-art deep learning price prediction models. As, an example, we show improvement in Litecoin price prediction up to a horizon of 12 days.
Historically, gold and silver have played distinct roles in traditional monetary systems. While gold has primarily been revered as a superior store of value, prompting individuals to hoard it, silver has commonly been used as a medium of exchange. As the financial world evolves, the emergence of cryptocurrencies has introduced a new paradigm of value and exchange. However, the store-of-value characteristic of these digital assets remains largely uncharted. Charlie Lee, the founder of Litecoin, once likened Bitcoin to gold and Litecoin to silver. To validate this analogy, our study employs several metrics, including unspent transaction outputs (UTXO), spent transaction outputs (STXO), Weighted Average Lifespan (WAL), CoinDaysDestroyed (CDD), and public on-chain transaction data. Furthermore, we've devised trading strategies centered around the Price-to-Utility (PU) ratio, offering a fresh perspective on crypto-asset valuation beyond traditional utilities. Our back-testing results not only display trading indicators for both Bitcoin and Litecoin but also substantiate Lee's metaphor, underscoring Bitcoin's superior store-of-value proposition relative to Litecoin. We anticipate that our findings will drive further exploration into the valuation of crypto assets. For enhanced transparency and to promote future research, we've made our datasets available on Harvard Dataverse and shared our Python code on GitHub as open source.
Erik Altman, Jovan Blanuša, Luc von Niederhäusern, Béni Egressy · 6 authors
With the widespread digitization of finance and the increasing popularity of cryptocurrencies, the sophistication of fraud schemes devised by cybercriminals is growing. Money laundering -- the movement of illicit funds to conceal their origins -- can cross bank and national boundaries, producing complex transaction patterns. The UN estimates 2-5\% of global GDP or \$0.8 - \$2.0 trillion dollars are laundered globally each year. Unfortunately, real data to train machine learning models to detect laundering is generally not available, and previous synthetic data generators have had significant shortcomings. A realistic, standardized, publicly-available benchmark is needed for comparing models and for the advancement of the area. To this end, this paper contributes a synthetic financial transaction dataset generator and a set of synthetically generated AML (Anti-Money Laundering) datasets. We have calibrated this agent-based generator to match real transactions as closely as possible and made the datasets public. We describe the generator in detail and demonstrate how the datasets generated can help compare different machine learning models in terms of their AML abilities. In a key way, using synthetic data in these comparisons can be even better than using real data: the ground truth labels are complete, whilst many laundering transactions in real data are never detected.
Abstract The rapid advancement of quantum computing has sparked a considerable increase in research attention to quantum technologies. These advances span fundamental theoretical inquiries into quantum information and the exploration of diverse applications arising from this evolving quantum computing paradigm. The scope of the related research is notably diverse. This paper consolidates and presents quantum computing research related to the financial sector. The finance applications considered in this study include portfolio optimization, fraud detection, and Monte Carlo methods for derivative pricing and risk calculation. In addition, we provide a comprehensive analysis of quantum computing’s applications and effects on blockchain technologies, particularly in relation to cryptocurrencies, which are central to financial technology research. As discussed in this study, quantum computing applications in finance are based on fundamental quantum physics principles and key quantum algorithms. This review aims to bridge the research gap between quantum computing and finance. We adopt a two-fold methodology, involving an analysis of quantum algorithms , followed by a discussion of their applications in specific financial contexts. Our study is based on an extensive review of online academic databases, search tools, online journal repositories, and whitepapers from 1952 to 2023, including CiteSeerX, DBLP, ResearchGate, Semantic Scholar, and scientific conference publications. We present state-of-the-art findings at the intersection of finance and quantum technology and highlight open research questions that will be valuable for industry practitioners and academicians as they shape future research agendas.
Lubdhak Mondal, Udeshya Raj, S Abinandhan, Began Gowsik S · 6 authors
This study investigates the relationship between narratives conveyed through microblogging platforms, namely Twitter, and the value of crypto assets. Our study provides a unique technique to build narratives about cryptocurrency by combining topic modelling of short texts with sentiment analysis. First, we used an unsupervised machine learning algorithm to discover the latent topics within the massive and noisy textual data from Twitter, and then we revealed 4-5 cryptocurrency-related narratives, including financial investment, technological advancement related to crypto, financial and political regulations, crypto assets, and media coverage. In a number of situations, we noticed a strong link between our narratives and crypto prices. Our work connects the most recent innovation in economics, Narrative Economics, to a new area of study that combines topic modelling and sentiment analysis to relate consumer behaviour to narratives.
The Non-Fungible-Token (NFT) market has experienced explosive growth in recent years. According to DappRadar, the total transaction volume on OpenSea, the largest NFT marketplace, reached 34.7 billion dollars in February 2023. However, the NFT market is mostly unregulated and there are significant concerns about money laundering, fraud and wash trading. The lack of industry-wide regulations, and the fact that amateur traders and retail investors comprise a significant fraction of the NFT market, make this market particularly vulnerable to fraudulent activities. Therefore it is essential to investigate and highlight the relevant risks involved in NFT trading. In this paper, we attempted to uncover common fraudulent behaviors such as wash trading that could mislead other traders. Using market data, we designed quantitative features from the network, monetary, and temporal perspectives that were fed into K-means clustering unsupervised learning algorithm to sort traders into groups. Lastly, we discussed the clustering results' significance and how regulations can reduce undesired behaviors. Our work can potentially help regulators narrow down their search space for bad actors in the market as well as provide insights for amateur traders to protect themselves from unforeseen frauds.