DOCSEARCHbyAlgolia
Submit Docs

Qiankun Zhu Portfolio

website·Indexed 29 days ago

https://qiankun.co.uk

A collection of financial research projects and case studies by Qiankun (Kenny) Zhu, covering equity valuation (CSCO), systematic alpha research with WorldQuant BRAIN, global equity portfolio construction, LNG SPA real-option valuation, and corporate ESG. Includes technical writing on Energy Trading and Risk Management (ETRM) systems.

Tokens
3.6K
Snippets
8
Records
25
Agent score
50%

What's inside Qiankun Zhu

  1. Create high-Sharpe alpha expressions using fastexpr

    Use fastexpr to combine financial metrics with ranking operators to generate alpha signals. Effective strategies often combine fundamental ratios (like Operating Income or Free Cash Flow relative to equity) with time-series ranks and industry grouping to neutralize sector bias.
    /* High Sharpe Strategy: Combining Operating Income and Estimated EPS */
0.5 * group_rank(ts_rank(operating_income / equity, 126), subindustry)
+ 0.5 * group_rank(ts_rank(est_eps / close, 126), industry)

/* Diversifier: Using Free Cash Flow and Price Reverse to lower correlation */
0.6 * group_rank(ts_rank(free_cash_flow_reported_value / equity, 126), industry)
+ 0.4 * group_rank(-ts_delta(close, 5), industry)

/* Options Hybrid: Combining Operating Income with Put/Call Open Interest */
0.5 * group_rank(ts_rank(operating_income / equity, 126), subindustry)
+ 0.5 * group_rank(-ts_rank(pcr_oi_270, 120), industry)

  2. Technical stack for equity portfolio optimization

    The portfolio construction and risk management pipeline is built using a config-driven Python architecture. It leverages scientific computing libraries for factor modeling and a convex optimization library for portfolio weighting.
    # Core dependencies
pip install pandas numpy scipy cvxpy scikit-learn

  3. Understand ETRM and Commodity Management (CM) systems

    Energy Trading and Risk Management (ETRM) software manages the end-to-end lifecycle of trading energy commodities. It is a specialized application of Commodity Trading and Risk Management (CTRM). The industry is evolving toward 'Commodity Management' (CM), which extends traditional trading and risk functionality to include supply chain logistics, inventory management, workforce management, and ERP-like accounting integration to support physical producers and consumers.

  4. Modern ETRM architectural trends

    ETRM architecture is shifting from monolithic on-premise installations to cloud-native SaaS and microservices. This evolution enables:

    • API-driven integration: Allowing data extraction for digital transformation and enterprise tool integration.
    • Agile implementation: Microservices allow vendors to deploy independent modules (e.g., accounting or market connections) without extensive custom coding.
    • Reduced Cost: SaaS models lower the high upfront licensing and implementation costs associated with legacy systems.

  5. Value LNG SPA contracts using Real-Option Valuation

    A long-term LNG Sale and Purchase Agreement (SPA) should be valued as a strip of real options rather than a static annuity. A static NPV approach ignores extrinsic value derived from the right to choose the destination and the right to cancel a cargo. For a 20-year US Gulf Coast FOB LNG SPA, this extrinsic value can represent a significant portion (e.g., 43%) of the contract's full mark.

  6. Cisco Systems (CSCO) Intrinsic Valuation Summary

    As of May 28, 2026, Cisco Systems (CSCO) is analyzed as materially overvalued. The base-case intrinsic value is estimated at $48.56 per share, compared to a market price of approximately $120.16. This valuation is derived from a free-cash-flow-to-firm (FCFF) discounted cash flow model, cross-checked with peer trading multiples and a 50,000-trial Monte Carlo simulation.

  7. Understand the Cisco Equity Valuation Python engine architecture

    The valuation is powered by a Python package (model/) where assumptions.py serves as the single source of truth for all tunable inputs. The engine is modularized to handle specific financial valuation tasks:

    <table> <thead> <tr><th>Module</th><th>Role</th></tr> </thead> <tbody> <tr><td>assumptions.py</td><td>Centralized tunable inputs; eliminates magic numbers.</td></tr> <tr><td>historicals.py</td><td>5-year 10-K data cleaning, segment crosswalks, and inorganic isolation.</td></tr> <tr><td>wacc.py</td><td>CAPM Ke, Hamada re-levering, tax glide, and market-value weights.</td></tr> <tr><td>forecast.py</td><td>10-year FCFF engine generating bull/base/bear arrays.</td></tr> <tr><td>dcf.py</td><td>Discounting, Gordon terminal value, and EV to equity bridge.</td></tr> <tr><td>multiples.py</td><td>Peer engine for cash-netted EV/EBITDA and GAAP P/E.</td></tr> <tr><td>sensitivity.py</td><td>5x5 grids, scenario tables, and reverse-DCF root-finders (using brentq).</td></tr> <tr><td>probabilistic.py</td><td>Scenario weighting and 50k Monte Carlo simulations with Gaussian copula.</td></tr> <tr><td>figures.py</td><td>Automatic generation of report exhibits.</td></tr> <tr><td>export.py + writers/</td><td>Generation of a 15-sheet Excel workbook.</td></tr> <tr><td>verify.py</td><td>Formula validation asserting Excel output equals Python engine output.</td></tr> </tbody> </table>

  8. Implement ETRM systems using best practices

    To ensure a successful ETRM implementation, follow these guidelines:

    • Executive Commitment: Ensure senior management dedicates top-tier internal resources rather than treating the project as a side task.
    • Data Cleansing: Perform data cleansing before migration to avoid importing obsolete counterparties or locations.
    • Phased Rollout: Avoid 'big bang' implementations. Use a phased approach or prototyping with live data to identify issues early.
    • Replacement Strategy: When replacing legacy systems, run the new system in parallel with the old one to validate financial figures before final cutover.

  9. Implement a Global Equity Portfolio Construction pipeline

    This system manages a £10M long-only global equity portfolio benchmarked against the MSCI World. The construction process follows a five-step pipeline: 1. Mandate definition (TE budget, ESG floor, DM/EM mix); 2. Scoring via 13-factor DM/EM composites (sector-neutral); 3. Selection based on liquidity, free float, vetoes, and greedy caps; 4. Optimization using factor covariance and MV/CVaR/robust methods via CVXPY; 5. Risk analysis covering Tracking Error (TE), VaR suite, stress testing, and FX.

  10. Perform Reverse DCF for Cisco Systems

    A reverse DCF analysis determines the growth assumptions required to justify a market price of $120.16. To reach this price, Cisco would require either:

    1. Perpetual forward revenue growth of approximately 17.8% (3.5x the base CAGR).
    2. A WACC of 5.79% (314 basis points below the calculated bottom-up rate).

  11. Implement a systematic alpha research pipeline for WorldQuant BRAIN

    To build a successful ACTIVE book in the WorldQuant BRAIN Challenge, follow a repeatable mining pipeline:

    1. Signal Generation: Create candidate signals using a large field universe (e.g., 4,300 fields) and baseline operators like group_rank and ts_rank.
    2. Quality Gating: Filter signals through specific performance metrics: Sharpe ratio, Fitness, Turnover (TO), Drawdown (DD), and Concentration.
    3. Correlation Filtering: Reject 'clone' strategies that have a daily-return correlation $\ge 0.7$ against the existing ACTIVE book to avoid self-correlation failures.
    4. Submission: Submit signals that pass all gates to move from Bronze to Gold tier status.

  12. Configure mandate constraints for Global Equity Portfolio

    The portfolio is governed by specific institutional-grade constraints to balance risk and return. These include limits on single name exposure, sector concentration, and tracking error, as well as liquidity and ESG requirements.
    {
  "constraints": {
    "single_name_max": "7%",
    "sector_max": "35%",
    "sub_industry_max": "15%",
    "tracking_error": {
      "process_cap": "8%",
      "mandate_cap": "15%"
    },
    "liquidity": "Minimum market cap & ADV",
    "esg": "Minimum Bloomberg ESG score per holding"
  }
}