Trading Strategy

Financial market geometry: The tube oscillator ArXiv ID: 2407.08036 “View on arXiv” Authors: Unknown Abstract Based on geometrical considerations, we propose a new oscillator for technical market analysis, the tube oscillator. This oscillator measures the trending behavior of a fixed market instrument based on its past history. It is shown in an empirical analysis of the German DAX and the Forex EUR/USD exchange rate that a simple trading strategy based on this oscillator and fixed threshold leads to consistent positive monthly returns of average magnitude of 2% or more. The oscillator is derived from a broader understanding of the geometric behavior of prices throughout a fixed period, which we term financial market geometry. The remarkable profit results of the presented technique show that 1) prices of financial market instruments have a strong underlying deterministic component which can be detected and quantified with a matching approach and 2) financial market geometry is capable of providing such detectors. ...

CatMemo at the FinLLM Challenge Task: Fine-Tuning Large Language Models using Data Fusion in Financial Applications ArXiv ID: 2407.01953 “View on arXiv” Authors: Unknown Abstract The integration of Large Language Models (LLMs) into financial analysis has garnered significant attention in the NLP community. This paper presents our solution to IJCAI-2024 FinLLM challenge, investigating the capabilities of LLMs within three critical areas of financial tasks: financial classification, financial text summarization, and single stock trading. We adopted Llama3-8B and Mistral-7B as base models, fine-tuning them through Parameter Efficient Fine-Tuning (PEFT) and Low-Rank Adaptation (LoRA) approaches. To enhance model performance, we combine datasets from task 1 and task 2 for data fusion. Our approach aims to tackle these diverse tasks in a comprehensive and integrated manner, showcasing LLMs’ capacity to address diverse and complex financial tasks with improved accuracy and decision-making capabilities. ...

The 10 Reasons Most Machine Learning Funds Fail ArXiv ID: ssrn-3104816 “View on arXiv” Authors: Unknown Abstract The rate of failure in quantitative finance is high, and particularly so in financial machine learning. The few managers who succeed amass a large amount of ass Keywords: Financial Machine Learning, Quantitative Finance, Asset Management, Predictive Analytics, Trading Strategy, Quantitative Finance / Equities Complexity vs Empirical Score Math Complexity: 2.0/10 Empirical Rigor: 1.5/10 Quadrant: Philosophers Why: The paper focuses on high-level methodological pitfalls and organizational paradigms in financial machine learning, with minimal advanced mathematical formalism. It lacks empirical backtests, statistical code, or implementation-heavy data analysis, making it more of a conceptual framework than a backtest-ready study. flowchart TD Q["Research Question:<br>Why do ML funds fail?"] --> D["Data: Financial ML<br>papers & strategies"] D --> M["Methodology: Cross-sectional<br>analysis of failures"] M --> C["Computational Process:<br>Identify recurring pitfalls"] C --> F["Findings: 10 systemic reasons<br>e.g., overfitting, data snooping"] F --> O["Outcome: Risk management<br>framework for ML funds"]