Estimated study time: 60 minutes
Content:
Portfolio construction at CFA Level 2 applies the mean-variance optimization (MVO) framework of Markowitz to practical portfolio management, including the capital market line, the efficient frontier, and the role of risk-free assets. The efficient frontier represents all portfolios that maximize expected return for a given level of risk (standard deviation) or minimize risk for a given expected return, when only risky assets are considered. Adding a risk-free asset creates the Capital Market Line (CML): E(Rp) = rf + [(E(Rm) - rf) / sigma_m] * sigma_p, where the slope is the Sharpe ratio of the market portfolio. All investors, regardless of risk preferences, hold a combination of the risk-free asset and the tangency portfolio (which has the highest Sharpe ratio on the efficient frontier). Risk preferences determine the proportion allocated to each.
The Capital Asset Pricing Model (CAPM) extends this by identifying the market portfolio as the tangency portfolio when all investors face identical expectations and the same opportunity set. CAPM: E(Ri) = rf + beta_i * [E(Rm) - rf], where beta_i = Cov(Ri, Rm) / Var(Rm). Alpha (Jensen's alpha) = Actual Return - CAPM Expected Return = Ri - [rf + beta*(Rm - rf)]. Positive alpha indicates the security outperformed its risk-adjusted expected return; negative alpha indicates underperformance. At Level 2, candidates must evaluate active investment strategies by computing information ratio, Sharpe ratio, and alpha, and understand how the Fundamental Law of Active Management connects an analyst's forecasting skill (IC) to portfolio performance: IR = IC * sqrt(breadth), where IC is the information coefficient (correlation between forecasts and outcomes) and breadth is the number of independent investment decisions per year.
Multi-factor models extend CAPM by attributing security returns to multiple risk factors beyond market beta. The Fama-French three-factor model: E(Ri) = rf + b1*(Rm-rf) + b2*SMB + b3*HML, where SMB (Small Minus Big) captures the size premium and HML (High Minus Low book-to-price) captures the value premium. The Carhart four-factor model adds momentum (MOM). Arbitrage Pricing Theory (APT) provides the theoretical foundation for multi-factor models, stating that asset expected returns are linearly related to their exposures (factor loadings) to k systematic risk factors. Multi-factor models are used to decompose portfolio returns into factor exposures and residual alpha, to construct risk-factor-targeted portfolios, and to explain the cross-section of expected stock returns beyond market beta.
Active portfolio management strategy selection at Level 2 involves distinguishing between systematic (rules-based) and discretionary approaches, and evaluating the sources of active return. Tracking error (active risk) = standard deviation of (portfolio return - benchmark return). The information ratio (IR) = active return / tracking error = alpha / active risk. IR is the efficiency metric for active management: a higher IR for the same level of tracking error implies better risk-adjusted active performance. The Fundamental Law: IR = IC * sqrt(N) * TC, where TC is the transfer coefficient (how well the portfolio implements the forecasts, constrained by factors like long-only constraints and turnover limits). Long-only constraints significantly reduce the transfer coefficient relative to a fully unconstrained long-short strategy.
Environmental, Social, and Governance (ESG) integration in portfolio construction creates specific portfolio construction considerations at Level 2. Negative screening reduces the investable universe, potentially increasing tracking error relative to unscreened benchmarks. Best-in-class ESG selection identifies leading ESG performers within each sector, maintaining sector neutrality while tilting toward better ESG profiles within sectors. ESG factor models quantify the financial impact of ESG scores on expected returns. The empirical relationship between ESG scores and returns is debated: some studies find outperformance (ESG as a risk mitigation factor), others find underperformance (reduced diversification), and the relationship may vary by market cycle. At Level 2, candidates evaluate these tradeoffs and how ESG constraints interact with traditional portfolio optimization.
Key Terms:
Quiz Questions:
Q1. Portfolio A has an expected return of 12% and a standard deviation of 15%. Portfolio B has an expected return of 10% and a standard deviation of 12%. The risk-free rate is 3%. Which portfolio is more efficient (higher Sharpe ratio)?
A) Portfolio A: Sharpe = (12%-3%)/15% = 0.60; Portfolio B: Sharpe = (10%-3%)/12% = 0.583; Portfolio A is more efficient. B) Portfolio B: Sharpe = 0.583 > Portfolio A Sharpe = 0.60 — wrong, A is higher. C) Portfolio A: Sharpe = 0.60; Portfolio B: Sharpe = 0.583; Portfolio A has a higher Sharpe ratio and is on the CML closer to the tangency portfolio. D) Portfolio B is more efficient because it has lower absolute risk.
Answer: A — Sharpe Ratio = (E(Rp) - rf) / sigma_p. Portfolio A: (12%-3%)/15% = 9%/15% = 0.60. Portfolio B: (10%-3%)/12% = 7%/12% = 0.583. Portfolio A has a higher Sharpe ratio (0.60 vs. 0.583) and is therefore more efficient — it offers more expected return per unit of total risk. When comparing portfolios on the CML, the portfolio with higher Sharpe ratio is preferred by all investors and would dominate Portfolio B.
---
Q2. A portfolio manager runs a domestic equity fund with the following performance versus the benchmark: Portfolio return = 14%; Benchmark return = 11%; Portfolio beta = 1.10; Market return = 11%; Risk-free rate = 3%. Calculate Jensen's alpha and interpret the result.
A) Alpha = Portfolio return - [rf + beta*(Rm - rf)] = 14% - [3% + 1.10*(11%-3%)] = 14% - [3% + 8.8%] = 14% - 11.8% = 2.2%. The manager generated 2.2% risk-adjusted excess return above what CAPM predicts for the portfolio's beta. B) Alpha = Portfolio return - Benchmark return = 14% - 11% = 3% (active return, not alpha). C) Alpha = (Portfolio return - rf) / beta - (Rm - rf) = (14%-3%)/1.10 - 8% = 10% - 8% = 2%. D) Alpha is undefined because market return equals benchmark return.
Answer: A — Jensen's alpha = Rp - [rf + beta*(Rm - rf)] = 14% - [3% + 1.10*(11%-3%)] = 14% - [3% + 8.8%] = 14% - 11.8% = 2.2%. The portfolio's beta of 1.10 means CAPM expects it to return 11.8% given the market return of 11%. The portfolio actually returned 14%, outperforming by 2.2% after adjusting for beta risk. This positive alpha suggests genuine skill or favorable security selection beyond what market exposure alone would have generated.
---
Q3. An equity analyst makes independent investment recommendations on 100 stocks per year. Her historical information coefficient (IC) — the correlation between her forecasts and realized returns — is 0.08. Assuming a transfer coefficient (TC) of 0.70 (partial implementation due to long-only constraints), what is her expected information ratio?
A) IR = IC * sqrt(N) * TC = 0.08 * sqrt(100) * 0.70 = 0.08 * 10 * 0.70 = 0.56. B) IR = IC * N = 0.08 * 100 = 8.0. C) IR = IC * sqrt(N) = 0.08 * 10 = 0.80 (without TC adjustment). D) IR = IC / N = 0.08 / 100 = 0.0008.
Answer: A — Fundamental Law of Active Management: IR = IC * sqrt(Breadth) * TC = 0.08 * sqrt(100) * 0.70 = 0.08 * 10 * 0.70 = 0.56. The TC of 0.70 reflects the constraint that a long-only fund cannot fully implement all forecasts (short positions are unavailable). Without the TC constraint (TC = 1.0), IR would be 0.80. The fundamental law shows that even modest skill (IC = 0.08) applied broadly (100 decisions/year) can generate meaningful active returns, and that implementation constraints (TC < 1) meaningfully reduce realized IR.
---
Q4. A portfolio is regressed on three Fama-French factors and produces the following: alpha = 0.5%/month, market beta = 1.05, SMB loading = 0.30, HML loading = -0.25. The monthly SMB premium is 0.2% and HML premium is 0.3%. The market premium is 0.6%/month. What is the portfolio's expected monthly return according to the Fama-French model?
A) E(Rp) = rf + beta*(Rm-rf) + b_SMB*SMB + b_HML*HML = rf + 1.05*0.6% + 0.30*0.2% + (-0.25)*0.3% = rf + 0.63% + 0.06% - 0.075% = rf + 0.615%. B) E(Rp) = 0.5% + 1.05*0.6% + 0.30*0.2% - 0.25*0.3% = 0.5% + 0.63% + 0.06% - 0.075% = 1.115%. C) E(Rp) = 1.05*0.6% = 0.63% (ignoring size and value factors). D) E(Rp) = rf + 1.05*0.6% = rf + 0.63% (CAPM, ignoring additional factors).
Answer: B — The Fama-French model: E(Rp) = alpha + rf + b_M*(Rm-rf) + b_SMB*SMB + b_HML*HML. Including alpha as part of the expected return attribution: expected monthly return = 0.5% (alpha) + 1.05*0.6% (market) + 0.30*0.2% (size tilt) - 0.25*0.3% (value, anti-value tilt) = 0.5% + 0.63% + 0.06% - 0.075% = 1.115% per month. The negative HML loading indicates the portfolio tilts away from value (toward growth) stocks.
---
Q5. An institutional investor implements a best-in-class ESG strategy for its domestic equity portfolio: each sector is filled with the top-ranked ESG companies within that sector. Compared to an unconstrained portfolio using the same starting universe, which of the following best describes the expected construction impact?
A) The best-in-class approach eliminates all ESG risk at zero additional cost. B) The best-in-class approach maintains sector weights similar to the benchmark, reducing the sector-level tracking error associated with negative screening; however, within-sector concentration (only holding top ESG names) increases idiosyncratic risk and may reduce diversification relative to an unconstrained portfolio. C) Best-in-class ESG investing always outperforms unconstrained portfolios. D) The best-in-class approach has no impact on the portfolio's expected risk and return because sector weights are unchanged.
Answer: B — The best-in-class approach's key advantage is sector neutrality (avoiding the sector bets inherent in blanket exclusions of energy or utilities, for example). This reduces factor-level tracking error relative to the benchmark. The constraint — holding only top ESG names within each sector — reduces within-sector diversification (the lowest ESG-ranked names may have low correlation to the high-ESG names, providing diversification benefit that is now excluded). Net effect: sector-level tracking error decreases, but individual security selection becomes more concentrated. The empirical performance impact is ambiguous and varies by ESG scoring methodology, market cycle, and sector composition.
---