Toward a Lifecycle-Aware Decision Support System Using Fundamental Data for Robust Stock Selection and Asset Allocation

As AI-driven decision systems move from research prototypes into continuous industrial operation, the central engineering challenge shifts from accuracy to system-level reliability, accountability, and lifecycle integration. Financial decision support exemplifies this challenge: fundamental analysis encodes analyst domain knowledge yet remains underused due to semi-structured, schema-drifting disclosures. We present a lifecycle-aware decision support system carrying fundamental data end-to-end through five decoupled stages with schema-drift accommodation. The extracted knowledge forms a 42-feature, four-category schema spanning financial ratios, stock type, trading behavior, and macroeconomic conditions. Rather than forecasting price magnitude, the model learns a benchmark-relative, interpretable target: whether a stock outperforms a fixed-income benchmark. Firm-level probabilities aggregate by market capitalization into a market-condition score that a rule-based layer converts into stock/fixed-income/gold allocation recommendations, preserving human control over the switching threshold. Evaluated on the Tehran Stock Exchange (2015–2025) under currency depreciation and inflation surges, the system attains about 70% accuracy at mid- to long-term horizons (3–12 months), nearly 10 percentage points above the best-performing fundamental-analysis baseline, and its allocation strategy, under the specified allocation policy, records higher nominal and real cumulative returns than the market and gold benchmarks. Post hoc SHAP analysis shows that macroeconomic conditions dominate model decisions, highlighting the model’s reliance on macroeconomic information while keeping individual predictions auditable to a human investor. These findings suggest the potential of lifecycle-aware engineering and decision-aligned interpretability for financial decision support under changing macroeconomic conditions.

Authors

Institutions

Publication Details

Journal
AI
Published
2026-11-18
DOI
https://doi.org/10.3390/ai7110489
Primary Topic
Stock Market Forecasting Methods
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Toward a Lifecycle-Aware Decision Support System Using Fundamental Data for Robust Stock Selection and Asset Allocation

Jafar Habibi, MohammadAmin Fazli, Ali Abrishami, AmirAli Jarrahi et al.
1 links
AI
Stock Market Forecasting Methods
article

Toward a Lifecycle-Aware Decision Support System Using Fundamental Data for Robust Stock Selection and Asset Allocation

Jafar Habibi, MohammadAmin Fazli, Ali Abrishami, AmirAli Jarrahi, Dariush Amiri
article en

Abstract

As AI-driven decision systems move from research prototypes into continuous industrial operation, the central engineering challenge shifts from accuracy to system-level reliability, accountability, and lifecycle integration. Financial decision support exemplifies this challenge: fundamental analysis encodes analyst domain knowledge yet remains underused due to semi-structured, schema-drifting disclosures. We present a lifecycle-aware decision support system carrying fundamental data end-to-end through five decoupled stages with schema-drift accommodation. The extracted knowledge forms a 42-feature, four-category schema spanning financial ratios, stock type, trading behavior, and macroeconomic conditions. Rather than forecasting price magnitude, the model learns a benchmark-relative, interpretable target: whether a stock outperforms a fixed-income benchmark. Firm-level probabilities aggregate by market capitalization into a market-condition score that a rule-based layer converts into stock/fixed-income/gold allocation recommendations, preserving human control over the switching threshold. Evaluated on the Tehran Stock Exchange (2015–2025) under currency depreciation and inflation surges, the system attains about 70% accuracy at mid- to long-term horizons (3–12 months), nearly 10 percentage points above the best-performing fundamental-analysis baseline, and its allocation strategy, under the specified allocation policy, records higher nominal and real cumulative returns than the market and gold benchmarks. Post hoc SHAP analysis shows that macroeconomic conditions dominate model decisions, highlighting the model’s reliance on macroeconomic information while keeping individual predictions auditable to a human investor. These findings suggest the potential of lifecycle-aware engineering and decision-aligned interpretability for financial decision support under changing macroeconomic conditions.

AIVol. 7(11)
Sharif University of Technology (IR), Khatam University (IR)
Stock Market Forecasting Methods
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon