Code-first walkthroughs of the API, with real request/response pairs you can reproduce — and the conventions and pitfalls spelled out. Fixed income and risk first; that is what this desk is for.
Fixed incomeClean/dirty price, accrued interest, YTM, duration, convexity and BPV from one POST — flat yield in, the full analytics sheet out.
Fixed incomeThe full analytics sheet of a fixed-rate bond in one POST — clean/dirty, accrued, YTM — and why day counts and settlement dates are where spreadsheets go wrong.
Fixed incomeDirty = clean + accrued: the same bond priced on a coupon date and mid-period, side by side, with the act/act accrual arithmetic shown.
Fixed incomeWhat the bpv field means, how it ties to modified duration, and how to scale it to a real position without losing precision to rounding.
CurvesMarket quotes in, discount factors and continuously-compounded zero rates at every pillar out — conventions and interpolation written down.
CurvesDF(t) = exp(−z·t) and back — the two views of the same curve, verified in pure Python against the API's own pillars.
CurvesThe desk workflow in two calls: bootstrap the swap curve, feed the pillars into the bond pricer, read off the curve-implied YTM.
RiskHistorical and parametric VaR side by side, portfolio mode with a positions × returns matrix, and a reproducible numpy example with byte-identical output.
RiskWhy ES is always the bigger number, why Basel FRTB standardized on it, and why the parametric-normal column deserves suspicion on fat-tailed data.
RiskEmpirical percentile vs fitted normal — a seeded example where the parametric number comes out higher, and why the gap is a distribution diagnostic, not a rule.
RiskMatrix mode: T×N returns plus market values in, VaR and Expected Shortfall in currency out — correlations captured by P&L aggregation, no correlation matrix input.
OptionsEuropean and American pricing with full greeks, implied volatility, and the unit conventions (theta/day, vega per point) that usually bite first.
OptionsThe volatility that reproduces an observed option price, by numerical inversion — with the no-arbitrage bounds and the 422s that guard them.
Choosing your stackAn honest comparison of self-hosted QuantLib, self-hosted ORE and a hosted API — by cost, coverage, setup and who should pick which. Including when not to use Quant-Box.