QMIX combines per-agent values to create decentralised policies.
problem Training decentralised policies from centralised learning.
method QMIX uses a mixing network to estimate joint action-values as a monotonic combination of per-agent values.
result QMIX significantly outperforms existing methods on the StarCraft Multi-Agent Challenge (SMAC).
Decentralised fund framework allocates capital via tokenised vaults.
problem Traditional asset management's inefficiencies and centralisation.
method Permissionless, multi-strategy capital allocation through on-chain vaults.
result Self-regulating, cooperative optimisation across financial domains.
QubitSwap improves DEX efficiency by reducing impermanent loss and slippage.
problem Challenges in decentralised exchanges, especially impermanent loss and slippage.
method Hybrid approach integrating external oracle price with internal pool dynamics, parameterized by z. result Reduction in impermanent loss and slippage compared to traditional DEX frameworks.
In many real-world settings, a team of agents must coordinate their behaviour while acting in a decentralised way. At the same time, it is often possible to train the agents in a centralised fashion in a simulated or laboratory setting, where global state information is available and communication constraints are lifte…
Optimizes liquidations in decentralized finance to manage credit risk.
problem Managing and liquidating positions in decentralized finance exchanges.
method Formulated as an ergodic optimal control problem, derived closed-form solutions for optimal liquidation strategies.
result Closed-form solutions balance immediate executions with price impacts and long-term rewards.
Decentralised optimisation tasks are important components of multi-agent systems. These tasks can be interpreted as n-player potential games: therefore game-theoretic learning algorithms can be used to solve decentralised optimisation tasks. Fictitious play is the canonical example of these algorithms. Nevertheless fic…
We analyse the learning performance of Distributed Gradient Descent in the context of multi-agent decentralised non-parametric regression with the square loss function when i.i.d. samples are assigned to agents. We show that if agents hold sufficiently many samples with respect to the network size, then Distributed Gra…
OpenAlpha validates decentralized capital strategies using game theory and market aggregation.
problem Decentralized capital management's lack of trust-minimised, adaptive deployment.
method Game-theoretic validation, adversarial auditing, market-based belief aggregation.
result Confidence scores from validation phases inform capital allocation rules.
The article provides formulas to hedge impermanent loss in decentralized markets.
problem Impermanent loss in concentrated liquidity provision in decentralized markets.
method Analytical characterizations and static replication formulas using European calls or puts.
result Static replication formulas accurately hedge impermanent loss.
DFMM automates market making with adaptive pricing and risk management.
problem Challenges in decentralised automated market making (AMMs).
method Data aggregator, order routing, rebalancing, arbitrageurs, protective buffers, algorithmic accounting.
result DFMM optimises inventory risk and ensures market stability.
It is now well known that decentralised optimisation can be formulated as a potential game, and game-theoretical learning algorithms can be used to find an optimum. One of the most common learning techniques in game theory is fictitious play. However fictitious play is founded on an implicit assumption that opponents' …
This study optimizes trading and arbitrage in decentralized finance's CPMs, revealing convexity costs and developing efficient strategies.
problem Optimizing trading and arbitrage in decentralized finance's constant product markets (CPMs).
method Developed models for CPMs in competing centralised exchanges, CPMs, and both venues. Derived computationally efficient strategies.
result Accurately estimated convexity costs in CPMs, which are linear in trade size and nonlinear in liquidity depth and exchange rate.
Cryptofinance generates yield through innovative decentralized finance.
problem Lack of intrinsic yield in crypto-assets.
method Analysis of yield-generating mechanisms in cryptofinance.
result Cryptofinance innovations enable yield generation through various decentralized and centralised mechanisms.
The paper proposes a fair and private decentralized deep learning framework.
problem Ensuring fairness and privacy in collaborative deep learning.
method A reputation system and differential privacy are used. FDPDDL framework is built with two stages: initialisation and update.
result FDPDDL achieves high fairness, comparable accuracy to centralised and distributed frameworks, and better accuracy than standalone.
Study liquidity provision in decentralized exchanges considering risk aversion and replication costs.
problem Economic viability of liquidity provision in decentralized exchanges (DEXs).
method Formulated strategic interactions as a sequential game with risk-averse LP, traders, and arbitrageurs.
result DEX liquidity depth is crucial for risk management, influenced by risk aversion and replication costs.
Distributed learning with random features and gradient descent improves performance and reduces memory usage.
problem Improving generalization in decentralized learning with limited memory.
method Distributed Gradient Descent with Random Features and Implicit Regularization.
result High probability bounds on predictive performance with optimal statistical rates.
Study loan contracts in DLPs using derivatives pricing and neural networks.
problem Optimizing and hedging risks in decentralized lending contracts.
method Derivatives pricing theory, deep neural networks, and statistical arbitrage.
result Developed a method to hedge risks in lending contracts and exploit arbitrage opportunities.
Paper analyzes constant-product market making protocols.
problem Understanding and optimizing constant-product market making.
method Mathematical analysis of trade splitting and fee recompounding.
result Splitting trades does not affect final exchange rate.
Improved Q-learning for multi-agent reinforcement learning by weighting joint action values.
problem QMIX restricts Q-values to monotonic mixtures, limiting complex value functions. method Introduced weighted projection to recover optimal policies, improving performance.
result CW QMIX and OW QMIX outperform baseline QMIX on multi-agent tasks.
Optimal benchmark design varies based on costs in financial manipulation.
problem Manipulation of price benchmarks in finance.
method Analyzes empirical pattern and cost structures to determine optimal benchmark design.
result The optimal benchmark depends on the relative sizes of fixed and variable costs.
Examines multiagent systems for complex learning tasks.
problem Achieving cohesive learning behavior in multiagent networks.
method General formulation for multiagent dynamics and conditions for learning.
result Conditions for achieving cohesive learning behavior in multiagent networks.
Paper presents a risk management framework for blockchain protocols.
problem Blockchain protocol risks affecting DLT and digital assets.
method Developed a comprehensive risk management framework using traditional taxonomy.
result Structured approach to identify, measure, monitor and report blockchain protocol risks.
Optimal dynamic fees found for AMMs to deter arbitrageurs and attract noise traders.
problem Optimizing fees in AMMs to balance against arbitrage and noise trading.
method Approximate closed-form solutions to control problem, study of fee structure.
result Two distinct fee regimes identified: high fees to deter arbitrage, low fees to attract noise traders.
Centralised training with decentralised execution is an important setting for cooperative deep multi-agent reinforcement learning due to communication constraints during execution and computational tractability in training. In this paper, we analyse value-based methods that are known to have superior performance in com…
Study on wealth and trading in PoS blockchain.
problem Decentralization and trading incentives in PoS.
method Analytic and stochastic tools, optimal control theory, mean field model.
result Miners balance PoS mining and trading for optimal strategy.
The paper examines Bitcoin's nature using fractal geometry and finds it highly persistent, affecting predictability and decentralization.
problem Understanding the nature and predictability of Bitcoin prices.
method Statistical analysis of Bitcoin returns using fractal geometry.
result Bitcoin exhibits high persistence in prices, reducing efficiency but increasing predictability.
We propose graph-dependent implicit regularisation strategies for distributed stochastic subgradient descent (Distributed SGD) for convex problems in multi-agent learning. Under the standard assumptions of convexity, Lipschitz continuity, and smoothness, we establish statistical learning rates that retain, up to logari…
The blockchain technology promises to transform finance, money and even governments. However, analyses of blockchain applicability and robustness typically focus on isolated systems whose actors contribute mainly by running the consensus algorithm. Here, we highlight the importance of considering trustless platforms wi…
A new algorithm improves Bayesian federated learning by reducing communication overhead.
problem Bayesian federated learning constraints, including privacy, data ownership, and communication overhead.
method Proposes Quantised Langevin Stochastic Dynamics (QLSD) for Bayesian federated learning, using gradient compression and variance reduction techniques.
result Non-asymptotic and asymptotic convergence guarantees for QLSD and its improved versions.
FTX's failure linked to Terra-Luna collapse and Binance's influence.
problem FTX's collapse due to misuse of native token and reliance on leverage.
method Analyzed on-chain data, studied cryptocurrency dependency structures, and examined public trades.
result FTX's downfall was accelerated by Binance's tweets and public reaction.
This paper studies how social media posts, especially by executives, affect stock prices.
problem Predicting stock market movements using social media data.
method Integrated sentiment analysis of Twitter and Reddit posts with historical stock data using time series models and deep learning.
result Improvements in stock price prediction when social media data, especially executive posts, are included.
EMIX minimizes surprise in multi-agent reinforcement learning.
problem Surprise and approximation bias in multi-agent reinforcement learning.
method Energy-based MIXer (EMIX) for minimizing surprise across multiple agents.
result EMIX demonstrates consistent stable performance in challenging StarCraft II scenarios.
The study examines how backrun auctions can protect traders from price manipulation.
problem Price manipulation by arbitrageurs in batched trading venues.
method Developed a laminated queueing model to study price manipulation and introduced a price manipulation coefficient.
result Bound the price manipulation coefficient and found it approximated by a 'zeta value' with measurable parameters.
Deep reinforcement learning algorithms have recently been used to train multiple interacting agents in a centralised manner whilst keeping their execution decentralised. When the agents can only acquire partial observations and are faced with tasks requiring coordination and synchronisation skills, inter-agent communic…
This work uses reinforcement learning to optimize task scheduling and execution in a dynamic multi-agent warehouse environment.
problem Optimizing task scheduling and execution in a dynamic multi-agent warehouse environment with limited observability.
method Deep reinforcement learning to solve both high-level scheduling and low-level multi-agent execution problems.
result Demonstrates the effectiveness of reinforcement learning in optimizing task scheduling and execution in a dynamic multi-agent environment.
MFIN networks improve crypto trading with multiple features.
problem Selecting and processing multiple features for effective trading.
method End-to-end framework using Multi-Factor Inception Networks (MFINs).
result MFINs learn uncorrelated, higher-Sharpe strategies not captured by traditional factors.
PVI combines federated learning and variational inference for probabilistic model training.
problem Federated learning's lack of probabilistic model uncertainty estimation.
method Partitioned variational inference (PVI) framework for federated probabilistic model training.
result PVI unifies fragmented literature and demonstrates effectiveness in various federated settings.
Paper proposes a GRU model to detect spoofing in retail investors.
problem Spoofing in unregulated markets with retail investors.
method GRU-based detection model using market variables.
result Model performs well in early detection of spoofing attempts.
Algorithmic trading systems on DEXs reject most candidate tokens, but the counterfactual outcome of rejected candidates is rarely measured.
problem Measuring the counterfactual outcome of rejected tokens in algorithmic trading systems on decentralized exchanges.
method Post-Rejection Follow-up Sampling (PRFS) introduces a methodology for collecting and analyzing the data needed to evaluate filter precision against actual market outcomes of rejected candidates.
result PRFS produces the data needed to evaluate filter precision against actual market outcomes of rejected candidates, not against synthetic backtest reconstructions.
Network-based strategy for optimal cryptocurrency portfolios identified.
problem Challenges in predicting cryptocurrency prices in a volatile market.
method Network methods to identify decorrelated cryptocurrencies, Markowitz Portfolio Theory.
result Network-based portfolios outperform benchmarks with high expected returns.
Agent-to-agent finance aims to manage payments and trust for AI agents.
problem Managing financial interactions between autonomous AI agents.
method Develops agent-to-agent finance concept and explores blockchain solutions.
result Agent-to-agent finance can address coordination frictions in financial markets.
RED-2400 is a public benchmark of trading events from a Solana exchange, labeled by algorithmic rejection.
problem Analyzing algorithmically-rejected trading events for insights into market dynamics.
method Public dataset of 6,660 algorithmically-rejected trading events, linked to post-rejection price and liquidity trajectories.
result First window of a planned series of datasets extending the time horizon and enabling regime-stratified analysis.
Paper optimizes liquidity provision in decentralized finance markets.
problem Strategic LPs face predictable losses and concentration risk in CL pools.
method Derive optimal liquidity provision strategy based on fees, PL, and concentration risk.
result Optimal strategy increases fee revenue and profit from marginal rate changes.
Study analyzes risk management in Aave and Compound lending protocols, finding v3 better than v2.
problem Risk management in decentralized lending protocols.
method Cross-version and cross-chain analysis using fixed effects model.
result v3 protocols have better risk management, with stronger impact on L2 blockchains.
G3M impermanent losses are a key issue in decentralized finance, affecting diversification benefits.
problem Impermanent losses in G3M market makers due to negative convexity.
method Established non-arbitrage bounds and analyzed empirical data.
result Median liquidity pools have net nil ROI when Impermanent Losses are considered.
Hour-Aware Adaptive Risk Management for Autonomous Memecoin Trading
problem Autonomous memecoin trading system performance
method Hour-of-day effects, filter precision, fragility
result 40.5% win rate, mean per-trade return of +0.62%, cumulative +117.7%
Unified methods for hedging impermanent loss in decentralized exchanges.
problem Hedging impermanent loss in liquidity provision at decentralized exchanges.
method Static and dynamic approaches using arbitrage-based methods for valuation and risk management.
result Unified valuation and hedging formulas for IL protection claims.
PreBit predicts Bitcoin price movements using social media and financial data.
problem Predicting extreme price movements of Bitcoin due to its volatility and speculative trading.
method Hybrid model combining FinBERT embeddings of Twitter content with candlestick data and technical indicators.
result The hybrid model can predict significant market movements with a profitable trading strategy.