{"paper_id":"2508.02356","title":"Neural Network-Based Algorithmic Trading Systems: Multi-Timeframe Analysis and High-Frequency Execution in Cryptocurrency Markets","authors":["Wěi Zhāng"],"paper_date":"2025-08-04","paper_type":"methodological","paper_url":"https://arxiv.org/pdf/2508.02356v1","page_url":"https://thequant.space/flowcharts/neural-network-based-algorithmic-trading-systems-multi-time/","scores":{"empirical_rigor":7,"math_complexity":6.5,"quadrant":"Holy Grail"},"proposes":{"status":"available","statement":"The paper proposes a neural‑network based algorithmic trading system that combines multi‑timeframe trend analysis with high‑frequency direction prediction networks.","quote":"Our approach combines multi-timeframe trend analysis with high-frequency direction prediction networks, achieving positive risk-adjusted returns through statistical modeling and systematic market exploitation.","summary":"This paper explores neural network-based approaches for algorithmic trading in cryptocurrency markets. Our approach combines multi-timeframe trend analysis with high-frequency direction prediction networks, achieving positive risk-adjusted returns through statistical modeling and systematic market…","strategy":"CNN with soft attention approach"},"assumptions":{"status":"available","items":[{"assumption":"Market information flows hierarchically across time‑frames, with larger‑scale trends constraining shorter‑scale movements.","type":"market","stated":"explicit","quote":"Our core hypothesis rests on the hierarchical nature of market information flow across different time‑frames."},{"assumption":"Transaction costs are 0.05% per trade and are included in back‑test results.","type":"execution","stated":"explicit","quote":"The testing framework simulates realistic trading conditions including 0.05% transaction costs and accounts for variance in slippage to ensure results reflect achievable performance."},{"assumption":"Feature normalization uses predetermined scaling parameters rather than running statistics to avoid data snooping.","type":"data","stated":"explicit","quote":"... with feature normalization performed using predetermined scaling parameters rather than dataset running statistics to prevent data snooping and ensure model stability."},{"assumption":"Direction networks are periodically retrained to remain relevant as market regimes shift.","type":"model","stated":"explicit","quote":"Direction networks undergo periodic retraining as market regime shifts occur, ensuring model relevance and preventing performance degradation over time."},{"assumption":"Risk controls (position limits, correlation limits, drawdown monitoring) operate independently of signal generation.","type":"execution","stated":"explicit","quote":"These controls operate independently of the signal generation process, providing multiple layers of protection against model failures or market disruptions."}]},"datasets":{"status":"available","items":[{"name":"Historical cryptocurrency market data","provider":null,"kind":"market-prices","access":"unknown","period":"2018‑present","universe":"single asset","quote":"The data pipeline integrates three primary information sources into a unified framework for trading signal generation, utilizing historical data spanning from 2018 to present with the largest timeframe being daily intervals."},{"name":"On‑chain analytics","provider":null,"kind":"crypto-onchain","access":"unknown","period":"2018‑present","universe":"single asset","quote":"For single asset, the system integrates multiple data sources including on‑chain analytics capturing network activity and transaction patterns."},{"name":"GDELT sentiment data","provider":"GDELT","kind":"text-news","access":"public","period":"15‑minute frequency","universe":"global news and social media","quote":"... providing 15‑minute frequency sentiment data from global news and social media sources.","known_id":"gdelt"},{"name":"S&P 500 index movements","provider":null,"kind":"macro","access":"unknown","period":null,"universe":"S&P 500","quote":"overall economic indicators such as S&P 500 index movements"}],"known":[{"id":"gdelt","name":"GDELT","provider":"GDELT Project","kind":"text-news","access":"public","quote":"For single asset, the system integrates multiple data sources including on-chain analytics capturing network activity and transaction patterns, traditional market data encompassing price and volume information, over-all economic indicators such as S&P 500 index move-ments, and sentiment data from GDELT datasets Leetaru and Schrodt [2013] providing 15-minute fre-quency sentiment data from global news and social me-dia","mentions":2,"used_in":2}],"sample":{"start":null,"end":null,"market":"cryptocurrency markets","asset_class":"crypto","frequency":"intraday"},"data_resolution":"unclear"},"transaction_costs":{"status":"available","modeled":"yes","headline_net_of_costs":true,"detail":"0.05% transaction costs per trade","quote":"The testing framework simulates realistic trading conditions including 0.05% transaction costs and accounts for variance in slippage to ensure results reflect achievable performance.","source":"front_matter","rule":"R3","candidates":{"front_matter":"yes","evidence_card":"yes","evidence_card_quoted":true},"card_detail":"0.05% transaction costs per trade","card_quote":"The testing framework simulates realistic trading conditions including 0.05% transaction costs and accounts for variance in slippage to ensure results reflect achievable performance.","meaning":"costs are part of the reported headline results"},"results":{"status":"available","evaluation_type":"trading-pnl","findings":[{"finding":"The CNN with soft‑attention architecture achieved the highest profit factor of 1.15, outperforming LSTM and CNN‑only variants.","kind":"performance","quote":"CNN + Soft Attention (Current) 1.15 Optimal performance"},{"finding":"Inference latency per prediction is sub‑50 ms, with total trade execution (including network and exchange latency) typically 100‑300 ms.","kind":"performance","quote":"The inference pipeline achieves sub‑50ms latency per prediction, with complete trade execution typically requiring 100‑300ms including network latency and exchange processing time."},{"finding":"Direction networks are compact (~520,000 parameters) and optimized for fast inference, enabling sub‑second trading decisions.","kind":"performance","quote":"The direction networks are relatively compact with approximately 520,000 parameters, optimizing for both performance and inference speed."},{"finding":"The hierarchical multi‑timeframe approach enables consistent profitability across diverse market regimes.","kind":"performance","quote":"This paper demonstrates that neural network‑based approaches can achieve consistent profitability in cryptocurrency markets through systematic exploitation of cross‑timeframe relationships and microstructural patterns."}],"reported_performance":{"strategy":"CNN with soft attention approach","sharpe":null,"sharpe_benchmark":null,"annual_return_pct":null,"total_return_pct":null,"max_drawdown_pct":null},"out_of_sample":{"method":"holdout","headline_out_of_sample":true}},"independent_evidence":{"status":"none_found","supports":[],"contradicts":[],"qualifies":[],"compared_with":30,"citations":{"cited_by_in_corpus":0,"by_independent_authors":0,"how_cited":{"used_as_baseline":0,"uses_method_or_data":0,"extends":0,"discusses_findings":0,"mentions":0,"unclassified":0},"independent_papers":[],"scope":"Citations between papers in this corpus only; citations from outside it are not counted."}},"implementation":{"status":"available","public":false,"code_url":null,"note":"No repository presented by the authors as their own was found in the paper. Absence here is not proof that no code exists."},"replication":{"status":"none_found","card":null,"note":"No replication of this paper has been run by thequant.space yet."},"extraction":{"text_source":"fulltext","model":"openai/gpt-oss-120b","version":1,"extracted_at":"2026-11-14T17:30:45Z"},"provenance":{"claims":"Everything here is what the authors state in the paper. Nothing was re-run or reproduced by the platform.","quotes":"Assumptions, datasets, cost detail and findings were extracted by a language model. Every quote is verbatim paper text. Where an item has `support`, \"full\" means every figure in it appears in the quoted sentences and \"partial\" means a figure is stated elsewhere in the paper; items with a figure found nowhere in the paper were discarded.","datasets_known":"datasets.known lists well-known vendors and public datasets by a stable id, found by a dictionary over the full text where the sentence says the data is used; it is independent of the model and covers only names in the dictionary.","flags":"transaction_costs.modeled, results.evaluation_type and results.out_of_sample.method are platform classifications (structured-decision models); see https://thequant.space/score-guide/#penalty-flags","comparisons":"supports / contradicts / qualifies are a language model's reading of two papers' quoted findings, restricted to closely related papers that share no author. They are pointers to read, not verified replications, and coverage is partial.","implementation":"code_url is a repository the paper itself presents as its own (rule-scored, ambiguous cases judged by a model), checked for liveness when it was recorded."},"docs":"https://thequant.space/agents/"}