Japan Quantitative Strategy: Applying LLMs to Corporate Analysis and Equity Investment Strategies

研报英文原文证据摘录

Japan Quantitative Strategy: Applying LLMs to Corporate Analysis and Equity Investment Strategies

Idea

May 21, 2026 01:10 AM GMT

Morgan Stanley MUFG Securities Co., Ltd.+MJapan Quantitative Strategy | Japan Kanji Ohi, CFA

Quantitative Strategist

Applying LLMs to Corporate Kanji.Ohi@morganstanleymufg.comSho Nakazawa +81 3 6836-8924

Equity Strategist

Sho.Nakazawa@morganstanleymufg.com +81 3 6836-8926

Analysis and Equity Investment Ukyo Haraguchi, CFA

Ukyo.Haraguchi@morganstanleymufg.com +81 3 6836-8925

Strategies Masahiro Tran

Masahiro.Tran@morganstanleymufg.com +81 3 6836-8878

We examined an LLM’s earnings-forecasting and stock selection Kazuya Hayashi

capabilities using financial and market data. Input data were de- EquityKazuya.Hayashi@morganstanleymufg.comStrategist +81 3 6836-8923

identified to limit look-ahead bias internal to the LLM. Although

accuracy lagged consensus forecasts, backtests confirmed a

stock selection effect attributable to the LLM.M

Key Takeaways Exhibit 1 : Quintile Spread Returns Using

We examined the scope for applying LLMs to corporate analysis and equity Forecast Growth Rates (LLM-based, After Risk

strategies using de-identified financial/market data and annual securities reports. Neutralization)

Although consensus estimates were more accurate, LLM forecasts were broadly

neutral on sales but tended to conservatively factor in profit-decline risk.

Forecasts tended to assume a continuation of past revenue trends and mean

reversion in profits, suggesting the LLM captured typical earnings patterns.

Backtests showed limited stock selection effects for raw data, but relatively

strong excess returns for high growth-forecast stocks using neutralized data.