Behavioral Finance, Portfolio Theory & Construction
Behavioral Finance: Biases, Prospect Theory and Client Application
Two families of biases
Cognitive errors stem from faulty reasoning and are broadly correctable with education; emotional biases stem from feelings and are usually accommodated or adapted to. Belief-perseverance biases include conservatism, confirmation, representativeness, illusion of control and hindsight. Information-processing biases include anchoring, mental accounting, framing, availability and recency. Emotional biases include loss aversion, overconfidence, self-control, status quo, endowment, regret aversion and affinity.
Prospect theory
Value is judged against a reference point. The value function is concave for gains, convex for losses, and steeper for losses (loss aversion, roughly 2:1). Probability weighting overweights small probabilities and underweights moderate ones. Portfolio consequences include the disposition effect (selling winners, holding losers), herding, home bias, overtrading and naive extrapolation of recent returns.
Market implications
Overreaction and underreaction help explain momentum and reversal patterns; the adaptive markets hypothesis holds that efficiency varies with conditions and competition. In practice, advisers use goals-based framing, automatic enrolment and pre-commitment to counter biases.
Traps
Confusing loss aversion with risk aversion; labelling overconfidence a cognitive error; assuming every bias is corrected the same way (educate cognitive, accommodate emotional); forgetting prospect theory concerns changes in wealth, not absolute levels.
Worked example
A coin flip pays +$150 or -$100 with equal probability; expected value is +$25. With a loss-aversion coefficient of 2, perceived value is roughly 0.5(150) minus 2 times 0.5(100) = 75 minus 100 = -25, so a loss-averse client rejects a positive expected-value gamble.
Portfolio Theory and Construction: Efficient Frontiers to the IPS
Markowitz and CAPM
Portfolio expected return is the weighted average of asset returns; portfolio risk is not, because it depends on covariances. Diversification reduces risk toward systematic risk only. The efficient frontier of risky assets combined with the risk-free rate gives the capital allocation line, and the tangency portfolio maximizes the Sharpe ratio. CAPM prices systematic risk: E(R) = Rf + beta times the market risk premium, where beta = covariance with the market divided by market variance.
Construction in practice
The investment policy statement records the return objective, risk tolerance (ability and willingness) and constraints: time horizon, tax, liquidity, legal and unique. Strategic asset allocation dominates long-run variation in returns. Rebalance by calendar or by threshold bands. Multi-factor models (size, value, momentum, quality) extend CAPM. Consider taxes, trading costs and currency hedging policy.
Traps
Treating high-risk assets as high-return by definition; measuring an asset's risk in isolation rather than its contribution to portfolio variance; believing low correlation removes all risk (only unsystematic risk is diversified away); using willingness alone as risk tolerance; assuming beta is stable and that alpha reflects skill rather than noise.
Worked example
Rf = 3%, market risk premium = 5%, beta = 1.2, so required return = 3 + 1.2(5) = 9%. A portfolio of 60% asset A (beta 1.0) and 40% asset B (beta 1.5) has beta = 0.6 + 0.6 = 1.2. Rule of thumb: adding a lower-return asset can still raise the Sharpe ratio if its correlation with the existing portfolio is low enough.