Risk Aware Algorithmic Trading From Social Media EventTriggers With Adaptive Sentiment Confidence and HumanCentered Oversight

Author(s)	Mr. Arnav Vandana Chakole
Country	India
Abstract	Social media now drives stock prices hours before traditional news breaks, yet most algorithmic trading systems reduce complex human emotions fear, anger, excitement, optimism, and confusion to a single positive/negative score and operate as uninterpretable black boxes, creating poor trading decisions and legal risks under regulations like MiFID II. We propose IATS (Interpretable Adaptive Trading System), which extracts a five‑dimensional emotion vector from social posts using FinBERT, uses SHAP to explain and gate every trade in real time, employs a Deep Q‑Network to dynamically adjust confidence thresholds based on market volatility and historical emotion patterns, and logs every decision on a private Hyperledger Fabric blockchain for tamper‑proof regulatory audit. Anchored in peer‑reviewed literature, projected performance shows a Sharpe ratio of 1.45-1.65 , maximum drawdown below 15% significantly outperforming scalar sentiment baselines while a case study of a fake GME bankruptcy hoax demonstrates the human‑in‑the‑loop gate preventing a catastrophic loss. This paper presents a rigorous design study with literature‑grounded projections awaiting full empirical validation.
Keywords	Quant, Ai , Algorithmic Trading , shap
Field	Computer > Artificial Intelligence / Simulation / Virtual Reality
Published In	Volume 7, Issue 3, May-June 2026
Published On	2026-05-11
DOI	https://doi.org/10.63363/aijfr.2026.v07i03.5421

About AIJFR Fees & Payment Current Issue Publication Archive	Submit Research Paper Track Submission Status Publication Guidelines Publication Ethics Peer Review & Plagiarism	Join as a Reviewer Editors & Reviewers Reviewer Referral Program Get Reviewer Membership Certi.	Website/Journal Policies Usage Policy Content Policies Privacy Policy

Contact Us		+91-9898-79-39-59	editor@aijfr.com

Advanced International Journal for Research