Set out the main inputs, outputs and cautions when using analytical tools to run fixed-income portfolios

Portfolio analytics systems take in each position's terms and cash flows, market prices, yield and spread curves, volatility and prepayment assumptions for option-embedded bonds, and scenario definitions. Outputs include effective duration, key rate durations, convexity, spread duration and DTS, expected return decompositions, tracking error against the benchmark and scenario profit and loss. Managers use them to test strategies before trading and to monitor risk afterwards. Cautions apply: outputs depend heavily on model assumptions, such as prepayment behaviour and rate volatility, which are hardest to verify when markets are stressed; stale or evaluated prices for illiquid bonds can understate risk; and historical correlations may break down. The trap is treating model outputs as facts rather than estimates to be questioned.