CSI A500 Index Active Quant via Regime Alpha Stock Selection Framework
CSI A500 Index Active Quant via Regime Alpha Stock Selection Framework
CSI A500 Index Active Quant via Regime Alpha Stock Selection Framework
By: BOPU Asset Management (Hong Kong) Limited
By: BOPU Asset Management (Hong Kong) Limited
Trade direction
Long-only
Long-only
Holding period
12–24 months
12–24 months
Upside
12.9% ann. excess vs CSI A500 (Live Sharpe 2.85)
12.9% ann. excess vs CSI A500 (Live Sharpe 2.85)
RECOMMENDED SIZE
5–15%
5–15%
KEY HOOK
Regime Alpha delivering high-Sharpe excess in policy-driven A-share regime
Regime Alpha delivering high-Sharpe excess in policy-driven A-share regime
01 · Executive Summary
01 · Executive Summary
This submission outlines a concentrated, high-conviction single quantitative strategy: CSI A500 Index Active Quant built on our proprietary Regime Alpha framework, purpose-built to address China A-share market structural frictions. Traditional static factor models consistently struggle in A-shares due to policy-driven sector rotation, severe factor crowding, and heavy retail trading noise. Our three-layer Regime Alpha architecture—cross-sectional universe segmentation, regime-dependent model gating, and dynamic forecast horizon tuning—resolves these flaws by matching optimized alpha signals to the prevailing macro market state in real time.
Launched live in April 2025, the strategy delivers consistent, risk-efficient excess returns across market cycles. Verified live performance metrics include a 12.89% annualized excess return versus CSI A500, an excess Sharpe Ratio of 2.85, and a capped maximum weekly excess drawdown of -2.89%. Through H1 2026, year-to-date excess return hit 10.03% with an elevated excess Sharpe of 4.45, demonstrating improving risk-adjusted returns amid evolving market regimes.
This submission outlines a concentrated, high-conviction single quantitative strategy: CSI A500 Index Active Quant built on our proprietary Regime Alpha framework, purpose-built to address China A-share market structural frictions. Traditional static factor models consistently struggle in A-shares due to policy-driven sector rotation, severe factor crowding, and heavy retail trading noise. Our three-layer Regime Alpha architecture—cross-sectional universe segmentation, regime-dependent model gating, and dynamic forecast horizon tuning—resolves these flaws by matching optimized alpha signals to the prevailing macro market state in real time.
Launched live in April 2025, the strategy delivers consistent, risk-efficient excess returns across market cycles. Verified live performance metrics include a 12.89% annualized excess return versus CSI A500, an excess Sharpe Ratio of 2.85, and a capped maximum weekly excess drawdown of -2.89%. Through H1 2026, year-to-date excess return hit 10.03% with an elevated excess Sharpe of 4.45, demonstrating improving risk-adjusted returns amid evolving market regimes.
Interview
02 · Investment Thesis
02 · Investment Thesis
2.1 Structural Edge: Why Regime Alpha Works for A-Shares
Regime Alpha is a dynamic stock selection system that identifies discrete, medium-to-long macro-driven market states and dynamically allocates tailored alpha signals, factor pools, and model parameters to each regime. Unlike short-term sentiment-focused contextual alpha models, our framework centers on discrete, macro-anchored regime segmentation with stable, clearly defined boundaries driven by national policy, industrial cycles, and institutional/retail investor structural behavior—rather than fleeting trading themes or minor style shifts. This long-cycle, policy-centric design makes it uniquely suited to A-shares’ policy-dominated market mechanics and delivers more predictable, sustainable alpha.
Market observers commonly cite three core headwinds for dynamic regime strategies in China’s equity market: rapid policy rotation, unstable factor correlations driven by crowding, and extreme retail trading noise. Our framework reframes these features as exploitable structural opportunities:
Persistent underlying macro regimes mask short-term volatility. While single stocks and sectors rotate rapidly, A-share markets operate within multi-quarter regimes defined by top-down policy priorities. These stable underlying states can be systematically modeled and monetized regardless of short-term thematic churn.
Policy transmission lengthens regime longevity. A-shares act as a primary transmission channel for national industrial and macroeconomic policy. Policy stimulus or regulatory guidance creates sustained multi-month market regimes, providing ample time windows for regime-aligned factor models to capture alpha before state shifts.
Regime-specific factor heterogeneity creates constant arbitrage space. No single factor outperforms across all market environments, but every regime hosts a subset of high-efficacy alpha signals. A broad diversified alpha library paired with precise regime identification mitigates single-factor crowding and delivers steady excess returns cycle-over-cycle.
2.2 Edge Persistence Over 12–24 Months
As quantitative A-share strategies multiply and factor crowding intensifies, our full-stack localized Regime Alpha system creates a replicable competitive moat unavailable to generic off-the-shelf quant tools. The three-layer proprietary architecture cannot be rapidly copied by competitors:
Cross-sectional pool segmentation: Separate regression models are built for distinct stock universes (e.g., large-cap index constituents vs. mid/small-cap fringe names) alongside dedicated sector models for innovation and tech industries, capturing universe-specific return drivers ignored by unified single-pool models.
Regime gating logic: Model sub-components are automatically activated or suppressed based on real-time regime classification, eliminating performance drag from irrelevant factors and lowering exposure to crowded factor trades.
Dynamic forecast horizon adaptation: Factor holding periods and prediction windows shift to match the speed of the current market regime, aligning signal timing with market turnover rhythms and boosting predictive hit rates.
2.1 Structural Edge: Why Regime Alpha Works for A-Shares
Regime Alpha is a dynamic stock selection system that identifies discrete, medium-to-long macro-driven market states and dynamically allocates tailored alpha signals, factor pools, and model parameters to each regime. Unlike short-term sentiment-focused contextual alpha models, our framework centers on discrete, macro-anchored regime segmentation with stable, clearly defined boundaries driven by national policy, industrial cycles, and institutional/retail investor structural behavior—rather than fleeting trading themes or minor style shifts. This long-cycle, policy-centric design makes it uniquely suited to A-shares’ policy-dominated market mechanics and delivers more predictable, sustainable alpha.
Market observers commonly cite three core headwinds for dynamic regime strategies in China’s equity market: rapid policy rotation, unstable factor correlations driven by crowding, and extreme retail trading noise. Our framework reframes these features as exploitable structural opportunities:
Persistent underlying macro regimes mask short-term volatility. While single stocks and sectors rotate rapidly, A-share markets operate within multi-quarter regimes defined by top-down policy priorities. These stable underlying states can be systematically modeled and monetized regardless of short-term thematic churn.
Policy transmission lengthens regime longevity. A-shares act as a primary transmission channel for national industrial and macroeconomic policy. Policy stimulus or regulatory guidance creates sustained multi-month market regimes, providing ample time windows for regime-aligned factor models to capture alpha before state shifts.
Regime-specific factor heterogeneity creates constant arbitrage space. No single factor outperforms across all market environments, but every regime hosts a subset of high-efficacy alpha signals. A broad diversified alpha library paired with precise regime identification mitigates single-factor crowding and delivers steady excess returns cycle-over-cycle.
2.2 Edge Persistence Over 12–24 Months
As quantitative A-share strategies multiply and factor crowding intensifies, our full-stack localized Regime Alpha system creates a replicable competitive moat unavailable to generic off-the-shelf quant tools. The three-layer proprietary architecture cannot be rapidly copied by competitors:
Cross-sectional pool segmentation: Separate regression models are built for distinct stock universes (e.g., large-cap index constituents vs. mid/small-cap fringe names) alongside dedicated sector models for innovation and tech industries, capturing universe-specific return drivers ignored by unified single-pool models.
Regime gating logic: Model sub-components are automatically activated or suppressed based on real-time regime classification, eliminating performance drag from irrelevant factors and lowering exposure to crowded factor trades.
Dynamic forecast horizon adaptation: Factor holding periods and prediction windows shift to match the speed of the current market regime, aligning signal timing with market turnover rhythms and boosting predictive hit rates.
03 · Valuation & Catalysts
03 · Valuation & Catalysts
3.1 Dual Performance Drivers
Excess returns stem from two complementary, layered return engines that stabilize performance through bull, bear, and sideways markets:
Regime-matched alpha (marginal performance enhancer): Accurate regime labeling and targeted factor weighting drive incremental outperformance that differentiates this strategy from standard static-factor index active quant products. This is the core proprietary return source of the Regime Alpha framework.
Idiosyncratic stock selection alpha (baseline return foundation): Within each segmented universe, customized fundamental and price-volume models capture individual stock mispricing, establishing consistent full-cycle baseline excess returns and setting the strategy’s long-term return ceiling.
3.2 Favorable & Challenging Market Regimes
High-Performance Regimes & Catalysts
Markets anchored by clear dominant policy logic: When a coherent long-cycle regime (e.g., tech innovation policy growth leadership) is established, regime classification accuracy peaks and factor matching efficiency maximizes stock-selection alpha capture.
Concentrated policy implementation windows: Clustered industrial policy rollouts reinforce regime stability and extend trend duration, lengthening profitable positioning windows for regime-aligned portfolios.
Quarterly/annual earnings reporting cycles: Mass synchronized fundamental data releases accelerate price convergence to intrinsic value, amplifying single-stock alpha while validating the prevailing macro regime narrative.
Range-bound markets with wide cross-stock return dispersion: Elevated individual stock return spreads expand idiosyncratic alpha opportunities, while stable regime conditions support consistent model predictive power.
Constrained Performance Regimes
Unstructured, ambiguous choppy markets: Environments lacking a clear core policy or growth driver generate blurred regime signals, reducing factor matching efficiency and compressing excess return potential.
Abrupt unanticipated regime transitions: Sudden policy reversals or macro tail events trigger fast state shifts. Temporary model recognition lag creates short-term factor misalignment and modest excess drawdowns.
3.1 Dual Performance Drivers
Excess returns stem from two complementary, layered return engines that stabilize performance through bull, bear, and sideways markets:
Regime-matched alpha (marginal performance enhancer): Accurate regime labeling and targeted factor weighting drive incremental outperformance that differentiates this strategy from standard static-factor index active quant products. This is the core proprietary return source of the Regime Alpha framework.
Idiosyncratic stock selection alpha (baseline return foundation): Within each segmented universe, customized fundamental and price-volume models capture individual stock mispricing, establishing consistent full-cycle baseline excess returns and setting the strategy’s long-term return ceiling.
3.2 Favorable & Challenging Market Regimes
High-Performance Regimes & Catalysts
Markets anchored by clear dominant policy logic: When a coherent long-cycle regime (e.g., tech innovation policy growth leadership) is established, regime classification accuracy peaks and factor matching efficiency maximizes stock-selection alpha capture.
Concentrated policy implementation windows: Clustered industrial policy rollouts reinforce regime stability and extend trend duration, lengthening profitable positioning windows for regime-aligned portfolios.
Quarterly/annual earnings reporting cycles: Mass synchronized fundamental data releases accelerate price convergence to intrinsic value, amplifying single-stock alpha while validating the prevailing macro regime narrative.
Range-bound markets with wide cross-stock return dispersion: Elevated individual stock return spreads expand idiosyncratic alpha opportunities, while stable regime conditions support consistent model predictive power.
Constrained Performance Regimes
Unstructured, ambiguous choppy markets: Environments lacking a clear core policy or growth driver generate blurred regime signals, reducing factor matching efficiency and compressing excess return potential.
Abrupt unanticipated regime transitions: Sudden policy reversals or macro tail events trigger fast state shifts. Temporary model recognition lag creates short-term factor misalignment and modest excess drawdowns.
04 · Risks & Mitigation
04 · Risks & Mitigation
This dedicated risk management system addresses unique vulnerabilities inherent to regime-based active quant investing for CSI A500.
4.1 Regime Misidentification & Recognition Lag Risk
Risk: Market state shifts may outpace model detection speed, or misclassification of the dominant macro regime leads to misaligned factor weights, suppressing near-term excess returns.
Mitigation:
Multi-signal cross-verification regime engine integrating policy announcements, institutional fund flows, corporate fundamentals, and price-volume metrics to eliminate single-dimension signal bias.
All regime definitions and matching rules undergo multi-cycle historical back testing and strict out-of-sample validation before live deployment.
4.2 Regime Switch Drawdown Risk
Risk: During rapid regime transitions, expiring factors lose efficacy while new regime factor calibrations lag, creating sequential excess drawdown pressure.
Mitigation:
Gradual transition bands within the regime gating mechanism; model weights shift incrementally rather than via full overnight rebalancing to avoid trading shock volatility.
Historical extreme regime shift scenario stress testing to calibrate conservative risk budget limits and embed permanent downside performance buffers.
This dedicated risk management system addresses unique vulnerabilities inherent to regime-based active quant investing for CSI A500.
4.1 Regime Misidentification & Recognition Lag Risk
Risk: Market state shifts may outpace model detection speed, or misclassification of the dominant macro regime leads to misaligned factor weights, suppressing near-term excess returns.
Mitigation:
Multi-signal cross-verification regime engine integrating policy announcements, institutional fund flows, corporate fundamentals, and price-volume metrics to eliminate single-dimension signal bias.
All regime definitions and matching rules undergo multi-cycle historical back testing and strict out-of-sample validation before live deployment.
4.2 Regime Switch Drawdown Risk
Risk: During rapid regime transitions, expiring factors lose efficacy while new regime factor calibrations lag, creating sequential excess drawdown pressure.
Mitigation:
Gradual transition bands within the regime gating mechanism; model weights shift incrementally rather than via full overnight rebalancing to avoid trading shock volatility.
Historical extreme regime shift scenario stress testing to calibrate conservative risk budget limits and embed permanent downside performance buffers.
05 · Position Sizing & Portfolio Fit
05 · Position Sizing & Portfolio Fit
5.1 Core Position Sizing Rules
The strategy follows standard index active quant construction principles, maintaining CSI A500 benchmark beta exposure while generating excess returns through disciplined active stock selection:
Non-negotiable pre-trade hard limits enforcing industry neutrality and style neutrality versus the benchmark to control tracking error drift.
Regime-aware dynamic risk budgeting: Within the approved total active risk budget, cross-universe and factor weights adjust continuously per real-time regime signals.
5.2 Institutional Portfolio Allocation Role
Within multi-strategy institutional portfolios, this CSI A500 Regime Alpha tool functions as a core strategic equity allocation building block. It delivers steady broad-market mid-cap A-share beta exposure alongside persistent, low-correlated regime-enhanced alpha. Its return stream exhibits low correlation with market-neutral equity, CTA, and global macro strategies, enabling robust overall portfolio diversification and lifting long-term aggregate portfolio Sharpe ratios when combined with alternative allocations.
5.1 Core Position Sizing Rules
The strategy follows standard index active quant construction principles, maintaining CSI A500 benchmark beta exposure while generating excess returns through disciplined active stock selection:
Non-negotiable pre-trade hard limits enforcing industry neutrality and style neutrality versus the benchmark to control tracking error drift.
Regime-aware dynamic risk budgeting: Within the approved total active risk budget, cross-universe and factor weights adjust continuously per real-time regime signals.
5.2 Institutional Portfolio Allocation Role
Within multi-strategy institutional portfolios, this CSI A500 Regime Alpha tool functions as a core strategic equity allocation building block. It delivers steady broad-market mid-cap A-share beta exposure alongside persistent, low-correlated regime-enhanced alpha. Its return stream exhibits low correlation with market-neutral equity, CTA, and global macro strategies, enabling robust overall portfolio diversification and lifting long-term aggregate portfolio Sharpe ratios when combined with alternative allocations.
06 · Manager Skill Highlight
06 · Manager Skill Highlight
6.1 Team Background & Local Specialization
The core quantitative investment team includes PhD graduates and top-tier talent from MIT, Stanford, and Tsinghua University, alongside multiple global mathematics and computer science competition gold medalists. The lead portfolio manager carries 18+ years of global quantitative equity experience, previously managing over USD 40 billion AUM at leading international quant firms, with dual expertise building global quant frameworks and adapting them to China’s unique A-share microstructure.
6.2 Proprietary Research & Trading Infrastructure
The strategy is supported by end-to-end self-built data, modeling, and execution systems exclusive to our A-share Regime Alpha workflow:
Custom A-share regime labeling and real-time recognition engine processing along with alpha optimizer.
Fully automated low-latency execution infrastructure paired with integrated pre-trade, intraday, and post-trade risk control systems, guaranteeing standardized, repeatable strategy operation with minimal manual intervention
6.1 Team Background & Local Specialization
The core quantitative investment team includes PhD graduates and top-tier talent from MIT, Stanford, and Tsinghua University, alongside multiple global mathematics and computer science competition gold medalists. The lead portfolio manager carries 18+ years of global quantitative equity experience, previously managing over USD 40 billion AUM at leading international quant firms, with dual expertise building global quant frameworks and adapting them to China’s unique A-share microstructure.
6.2 Proprietary Research & Trading Infrastructure
The strategy is supported by end-to-end self-built data, modeling, and execution systems exclusive to our A-share Regime Alpha workflow:
Custom A-share regime labeling and real-time recognition engine processing along with alpha optimizer.
Fully automated low-latency execution infrastructure paired with integrated pre-trade, intraday, and post-trade risk control systems, guaranteeing standardized, repeatable strategy operation with minimal manual intervention
Optional Exhibit —————
Optional Exhibit —————
The strategy launched live April 14, 2025, with verified cross-cycle performance through June 2026, confirming practical outperformance beyond theoretical backtest outputs. Key realized live metrics:
Annualized excess return vs CSI A500: 12.89%
Excess Sharpe Ratio: 2.85
Excess return volatility: 4.53%
Maximum weekly excess drawdown: -2.89%
January–June 2026 YTD excess return reached 10.03%, with excess Sharpe improved to 4.45, reflecting strengthened risk-adjusted returns amid the current tech innovation regime.
Two critical regime shifts within the live track record validate the framework’s adaptability:
July 2025: The regime engine correctly identified a sustained rotation from value to growth leadership, triggering targeted factor gating adjustments that captured incremental excess returns through the style transition period.
Early 2026 onward: The dedicated tech & innovation segmented model, built specifically for the current policy-driven growth regime, has consistently captured alpha from innovation-themed mid-cap constituents within the CSI A500 universe.
The strategy has delivered stable outperformance across rising bull markets, range-bound consolidation phases, and moderate market pullbacks, proving robust regime adaptability and repeatable stock-selection alpha generation in live trading conditions.
The strategy launched live April 14, 2025, with verified cross-cycle performance through June 2026, confirming practical outperformance beyond theoretical backtest outputs. Key realized live metrics:
Annualized excess return vs CSI A500: 12.89%
Excess Sharpe Ratio: 2.85
Excess return volatility: 4.53%
Maximum weekly excess drawdown: -2.89%
January–June 2026 YTD excess return reached 10.03%, with excess Sharpe improved to 4.45, reflecting strengthened risk-adjusted returns amid the current tech innovation regime.
Two critical regime shifts within the live track record validate the framework’s adaptability:
July 2025: The regime engine correctly identified a sustained rotation from value to growth leadership, triggering targeted factor gating adjustments that captured incremental excess returns through the style transition period.
Early 2026 onward: The dedicated tech & innovation segmented model, built specifically for the current policy-driven growth regime, has consistently captured alpha from innovation-themed mid-cap constituents within the CSI A500 universe.
The strategy has delivered stable outperformance across rising bull markets, range-bound consolidation phases, and moderate market pullbacks, proving robust regime adaptability and repeatable stock-selection alpha generation in live trading conditions.
Conflicts Disclosure
Conflicts Disclosure
This submission details a focused, high-conviction single index active quant strategy centered on our proprietary Regime Alpha stock selection framework. Unlike opaque black-box quant models reliant solely on polished backtest curves, our strategy’s alpha originates from explainable, enduring structural A-share market inefficiencies, backed by standardized systematic workflows, multi-layered regime-specific risk management, and cross-cycle verified live trading performance.
This submission details a focused, high-conviction single index active quant strategy centered on our proprietary Regime Alpha stock selection framework. Unlike opaque black-box quant models reliant solely on polished backtest curves, our strategy’s alpha originates from explainable, enduring structural A-share market inefficiencies, backed by standardized systematic workflows, multi-layered regime-specific risk management, and cross-cycle verified live trading performance.
About Bopu & Linchun sun
About Bopu & Linchun sun
Hong Kong Bopu Asset Management Limited (hereinafter referred to as “Hong Kong Bopu Asset” or the “Company”) was established on 23 July 2024. It is an indirectly controlled affiliated company of Shenzhen Bopu Technology Co., Ltd. (hereinafter referred to as “Shenzhen Bopu Technology”). Founded in 2012, Shenzhen Bopu Technology is a technology-driven quantitative private fund management firm with AUM exceeding RMB 10 billion. It focuses on multi-strategy investment and quantitative research across both domestic and international markets.
On 26 March 2025, Hong Kong Bopu Asset officially obtained the Type 9 (Asset Management) regulated activity license from the Hong Kong Securities and Futures Commission (SFC), marking the Company’s formal qualification to conduct regulated asset management business in Hong Kong. Leveraging the quantitative investment research capabilities and global market experience accumulated by Shenzhen Bopu Technology over many years, Hong Kong Bopu Asset is committed to providing differentiated asset management products and services centered on quantitative investment for professional investors in Hong Kong and globally.
Since its establishment, the Company’s business has developed rapidly. As of the end of June 2026, the assets under management (AUM) of Hong Kong Bopu Asset had exceeded USD 47 million.
Hong Kong Bopu Asset Management Limited (hereinafter referred to as “Hong Kong Bopu Asset” or the “Company”) was established on 23 July 2024. It is an indirectly controlled affiliated company of Shenzhen Bopu Technology Co., Ltd. (hereinafter referred to as “Shenzhen Bopu Technology”). Founded in 2012, Shenzhen Bopu Technology is a technology-driven quantitative private fund management firm with AUM exceeding RMB 10 billion. It focuses on multi-strategy investment and quantitative research across both domestic and international markets.
On 26 March 2025, Hong Kong Bopu Asset officially obtained the Type 9 (Asset Management) regulated activity license from the Hong Kong Securities and Futures Commission (SFC), marking the Company’s formal qualification to conduct regulated asset management business in Hong Kong. Leveraging the quantitative investment research capabilities and global market experience accumulated by Shenzhen Bopu Technology over many years, Hong Kong Bopu Asset is committed to providing differentiated asset management products and services centered on quantitative investment for professional investors in Hong Kong and globally.
Since its establishment, the Company’s business has developed rapidly. As of the end of June 2026, the assets under management (AUM) of Hong Kong Bopu Asset had exceeded USD 47 million.
Linchun Sun currently serves as Deputy General Manager and Licensed Representative of Hong Kong Bopu Asset. He previously worked on database development with the Oracle Berkeley DB team. He joined Bopu in 2012 and has 14 years of experience in the asset management industry, with deep expertise and extensive practical experience in quantitative system design, product design, and multi-asset allocation.
Linchun Sun currently serves as Deputy General Manager and Licensed Representative of Hong Kong Bopu Asset. He previously worked on database development with the Oracle Berkeley DB team. He joined Bopu in 2012 and has 14 years of experience in the asset management industry, with deep expertise and extensive practical experience in quantitative system design, product design, and multi-asset allocation.
COPFA Alpha Showcase 2026
For Professional investors only
This website and the articles contained herein are prepared solely for purpose of educational and academic discussion among members of COPFA and the broader industry community. The content in this section is intended solely for readers of industry community who shall be Professional Investors (as defined under the Securities and Futures Ordinance of Hong Kong and its subsidiary legislation). If you are not a Professional Investors, please do not proceed. The China Overseas Private Funds Association (“COPFA”) is a non-profit industry association. The COPFA Alpha Showcase is an industry initiative to showcase high-quality investment thinking. COPFA does not provide investment advice and does not endorse, recommend or guarantee any specific investment idea, strategy, product, or manager. The views and opinions expressed in the articles are solely those of the individual contributing managers and do not represent the views of COPFA. The articles are provided for illustrative and educational purposes only and do not constitute investment, legal, tax or professional advice, nor do they constitute an offer to sell, or a solicitation to buy any securities or financial instruments in any jurisdiction. Where the contributing managers are licensed or regulated, such licensing status does not imply any verification or endorsement of the content by COPFA or the regulatory authorities. COPFA makes no representation or warranties, express or implied, as to the accuracy, completeness, timeliness, or reliability of the information contained herein. Past performance is not indicative of future results. Readers must not rely on any information or viewpoints contained herein to make any investment or business decisions. To the fullest extent permitted by law, COPFA, its directors, officers, employees, and the contributing authors/managers hereby expressly disclaim any and all liability for any direct, indirect, incidental, or consequential loss or damage arising from or in connection with the use of or reliance on this website or its contents. For professional investors only. Not for public distribution.
Copyright © COPFA. All rights reserved. No part of this website or information herein may be reproduced, distributed, or transmitted in any form without the prior written permission of COPFA.