INVESTMENT MANAGEMENT GAME (IMAG)

Lesson 6: Performance Measurement & the Feedback Loop

IMAG — Investment Management Game

Learning Outcomes

By the end of this lesson you will be able to:

  • Calculate and interpret risk-adjusted performance metrics
  • Decompose portfolio return into asset allocation and security selection effects
  • Conduct a structured post-mortem of your investment decisions
  • Close the feedback loop: update your IPS and views for the next round

Teaching chapters: Performance Measurement & Attribution · Feedback Loop

Performance Measurement

Return Calculation

  • Holding period return: \((P_t - P_{t-1} + D_t) / P_{t-1}\)
  • Time-weighted return (TWR): eliminates impact of external cash flows — the industry standard
  • Money-weighted return (MWR / IRR): reflects timing of investor cash flows — useful for client reporting
  • Distinguish: gross of fees vs. net of fees

→ Content: worked return calculation using IMAG Results screen.

Risk-Adjusted Performance Metrics

Metric Formula What it tells you
Sharpe ratio \((R_p - R_f) / \sigma_p\) Return per unit of total risk
Sortino ratio \((R_p - R_f) / \sigma_{downside}\) Return per unit of downside risk
Information ratio \((R_p - R_b) / \text{TE}\) Active return per unit of active risk
Calmar ratio Annualised return / Max drawdown Return per unit of drawdown
Maximum drawdown Peak-to-trough loss Worst-case loss

→ Content: your team’s metrics from Cesim Results.

Benchmark Comparison

  • Did you outperform or underperform the benchmark?
  • Decompose: how much came from asset allocation vs. security selection vs. interaction?
  • Distinguish: good decision vs. lucky outcome (attribution analysis)

→ Content: benchmark return vs. portfolio return chart.

Performance Attribution

The Brinson-Hood-Beebower Model

Three effects explain the difference between portfolio and benchmark return:

\[R_p - R_b = \underbrace{AA}_{\text{Allocation}} + \underbrace{SS}_{\text{Selection}} + \underbrace{IA}_{\text{Interaction}}\]

→ Content: attribution formula derivation; intuitive explanation.

Asset Allocation Effect

\[AA_i = (w_{p,i} - w_{b,i}) \times (R_{b,i} - R_b)\]

  • Did your overweights / underweights add value?
  • Positive if you overweighted asset classes that outperformed the benchmark
  • This is the TAA contribution (Lesson 5 decisions)

→ Content: allocation effect table for each IMAG asset class.

Security Selection Effect

\[SS_i = w_{b,i} \times (R_{p,i} - R_{b,i})\]

  • Did your chosen instruments outperform the asset class benchmark?
  • Positive if your picked securities beat their benchmark within each class
  • This is the security selection contribution (Lesson 5 decisions)

→ Content: selection effect table; best and worst contributors.

Reading the Attribution Report

→ Content: annotated screenshot of Cesim attribution output; interpretation guide.

Note

A negative information ratio over multiple periods is a signal to reduce active risk — not increase conviction.

ESG Performance

  • How does your portfolio ESG score compare to the benchmark?
  • Did ESG integration help or hurt financial returns in this period?
  • Track ESG score evolution over rounds

→ Content: ESG score chart; financial vs. ESG performance scatter plot.

The Feedback Loop

Why a Structured Post-Mortem?

  • The goal is not to explain away underperformance but to improve the process
  • Separate: good process / bad outcome from bad process / good outcome
  • Distinguish luck from skill — especially important in a short-horizon simulation

→ Content: 2×2 process/outcome matrix.

Post-Mortem Framework

For each major decision this round, answer:

  1. What was the decision? (specific weight or security)
  2. What was the rationale? (signal, conviction level)
  3. What was the outcome? (attribution contribution)
  4. Was the process sound? (would you make the same decision again?)
  5. What will you do differently?

→ Content: decision log template; worked example.

Common Behavioural Pitfalls

Bias Symptom Antidote
Overconfidence Excess tracking error Respect risk limits
Anchoring Holding losers too long Pre-commit to stop-loss rules
Herding Following consensus Document independent views before group discussion
Recency bias Over-weighting last quarter Use long-run CMA as anchor
Confirmation bias Seeking data that confirms view Assign devil’s advocate role

Updating Your IPS & Views

  • Were any IPS constraints binding? Should they be revised?
  • Did your CMA forecasts hold? What needs updating?
  • Are your TAA signals still valid for the next round?
  • Carry lessons learned into the next round’s decision process

→ Content: IPS review checklist; CMA update worksheet.

Key Takeaways

  • Attribution is a diagnostic tool — use it to improve, not to justify
  • Separate process quality from outcome quality
  • The feedback loop is what turns a simulation into genuine learning

Simulation Task — Lesson 6

Final Deliverables

Performance Report:

Post-Mortem:

Further Reading

  • Chapter: Performance Measurement & Attribution
  • Chapter: Feedback Loop