IFoA SP5 key terms

A framework that projects assets and liabilities together under consistent economic and demographic assumptions, so that a strategy can be judged by whether it funds obligations as they fall due, not by return alone.

A portfolio construction method that selects weights trading expected return against variance of return, producing an efficient frontier. It relies on inputs of expected returns, volatilities and correlations that are estimated with error.

The set of portfolios offering the highest expected return for each level of risk, or the lowest risk for each expected return. Portfolios below it are dominated and should not be held by a rational investor.

The danger that optimisers treat estimated inputs as if they were true values. Small errors in expected returns are magnified into extreme weights, so optimised portfolios often underperform naive ones out of sample.

A Bayesian approach that starts from market equilibrium implied returns and blends in an investor's views with stated confidence. The result is a set of expected returns that is more stable than raw historical averages.

A technique that pulls noisy sample statistics such as historical means or covariance matrices towards a structured target, reducing estimation error and improving the stability of portfolio weights derived from them.

A table giving the variance of each asset's return and the pairwise covariances between assets. It drives diversification benefits in portfolio construction and must be positive semi-definite to be internally consistent.

A standardised measure between minus one and one of how two series move together. In portfolio construction it determines diversification benefit; correlations often rise in crises, weakening that benefit exactly when it is needed.

Spreading exposure across assets, sectors or factors whose returns are imperfectly correlated, so that portfolio risk falls below the weighted average of individual risks without a proportionate fall in expected return.

Excess expected return over the risk-free rate divided by return volatility. It measures reward per unit of total risk and is used to compare portfolios and to assess whether active risk-taking has been rewarded.

Active return relative to a benchmark divided by tracking error. It measures the consistency of a manager's skill in deviating from the benchmark and is central to judging active mandates.

The standard deviation of the difference between portfolio and benchmark returns. It quantifies the size of active bets and constrains how far a manager may deviate from the benchmark.

The proportion of portfolio holdings that differ from benchmark weights. It measures the extent of stock-specific deviation, complementing tracking error, which also captures factor and timing differences.

An allocation of a total risk allowance across managers, sectors or strategies, so that each unit of risk can be monitored and attributed. It recognises that risk, not capital, is the scarce resource.

The amount of total portfolio volatility attributable to a holding, equal to its weight times its marginal contribution to risk. It reveals concentrations that weight-based measures hide.

A loss threshold that is expected to be exceeded with a stated small probability over a stated horizon. It says nothing about the size of losses beyond that threshold.

The average loss conditional on losses exceeding the value at risk threshold. It captures tail severity, is sub-additive, and is therefore preferred to VaR for aggregating risk across positions.

Evaluating a portfolio under severe but plausible scenarios, such as a rate shock or credit crisis, to reveal vulnerabilities that statistical models calibrated on benign history would miss.

Projecting portfolio and liability outcomes under a coherent set of specified economic conditions, often combining several risk factors moving together, to test resilience and inform contingency planning.

Simulating many future paths of economic variables and asset returns from an assumed model, producing a distribution of outcomes rather than a single deterministic path.

Setting model parameters to reflect actual expected returns, volatilities and correlations, including risk premia, so simulated outcomes represent plausible real experience. It differs from risk-neutral calibration used for pricing.

Pricing derivatives by discounting expected payoffs under a probability measure adjusted so that all assets earn the risk-free rate. It is appropriate for pricing but not for projecting realistic long-term returns.

The principle that two portfolios with identical future payoffs must cost the same today, otherwise a riskless profit could be made. It underpins derivative pricing and the construction of replicating portfolios.

A combination of traded assets whose cash flows mimic those of a liability or derivative. It is used to price the obligation and to hedge it by holding the replicating assets.

Structuring a bond portfolio so that its value is protected against small parallel shifts in interest rates, typically by matching duration and present value with liabilities, and often extended to convexity matching.

A measure of a bond's price sensitivity to a small change in yield, expressed in years. Macaulay duration is the weighted average term of cash flows; modified duration gives the percentage price change per unit yield change.

The second-order curvature of a bond's price-yield relationship. Positive convexity means prices rise more for a yield fall than they fall for an equal yield rise, and it improves the accuracy of duration-based hedging.

The excess yield on a corporate or sovereign bond over a comparable risk-free bond, compensating for default, downgrade and liquidity risk. It widens when default expectations or risk aversion rise.

The likelihood that an issuer fails to meet a payment obligation within a stated period. It combines with loss given default and exposure to give expected credit loss.

The share of exposure not recovered if a borrower defaults, equal to one minus the recovery rate. It depends on seniority, collateral and the legal regime.

The risk that a position cannot be sold quickly at a price close to its fair value, or that funding cannot be raised when needed. It is usually mild in normal markets and severe in stressed ones.

The difference between the price at which a dealer will buy and the price at which it will sell. It is a direct transaction cost and a practical measure of market liquidity.

The risk of extreme outcomes in the far tails of a return distribution, occurring more often than a normal distribution implies. It is poorly captured by volatility and better addressed by scenario and expected shortfall analysis.

The observed tendency of financial return distributions to have more extreme observations than a normal distribution predicts. It understates the frequency of large losses if normality is wrongly assumed.

The danger arising from large exposure to a single issuer, sector, factor or counterparty, so that one adverse event causes disproportionate loss. It is reduced by limits and by diversification.

The explicit and implicit costs of trading, including commissions, taxes, bid-ask spread and market impact. They reduce realised returns and must be modelled when evaluating turnover-heavy strategies.