TheQuantBateman

01

INTUITION

Separate the quote from the quantity inferred from it.

A zero rate compresses one maturity’s discount factor; a forward rate describes the marginal price of money between two dates. The forward curve is fixed by today’s discount curve under no-arbitrage, while its interpretation as an expectation requires an additional model and risk-premium view.

01

Average zero rates and marginal forward rates contain different information.

02

A smooth zero curve can imply unstable forwards if the interpolation is poorly chosen.

03

Forward equals expected future short rate only under a specified measure and after accounting for convexity or term premium.

02

WHY MARKETS CARE

Start from executable inputs and conventions.

Forwards drive FRA and swap projection, carry and roll-down, policy-path inference, and the hedge map between adjacent curve tenors.

INSTRUMENTS

zero-coupon bonds

FRAs

OIS forwards

swaps and futures

QUOTE CONVENTION

This lesson displays annualised continuous forwards unless labelled simple. Market contracts use their specified accrual basis and compounding.

03

MATHEMATICS

Transform quotes without losing units or arbitrage constraints.

Formula · Short derivation

Interval continuous forward

f(0;T_1,T_2)=\frac{\ln P(0,T_1)-\ln P(0,T_2)}{T_2-T_1}

The constant continuous rate implied for investing between two future dates.

Open in Analytics

Formula · Short derivation

Simple forward

L(0;T_1,T_2)=\frac{P(0,T_1)/P(0,T_2)-1}{\delta(T_1,T_2)}

The money-market-style forward compatible with an accrual-period payoff.

Open in Analytics

Short derivation

Locking a future borrowing rate with zero-coupon claims

Compare investing to T₂ directly with investing to T₁ and rolling under a rate fixed today.

01
Buy a T₂ unit payoff
Pay P(0,T₂) today to receive one unit at T₂.
02
Normalize at the forward start
A claim costing P(0,T₂)/P(0,T₁) at T₁ has the same time-zero value after discounting to today.
$P(0,T_1)\frac{P(0,T_2)}{P(0,T_1)}=P(0,T_2)$
03
Invert the accumulation factor
The simple forward accumulation from T₁ to T₂ is the reciprocal of that forward-start discount price.
$1+\delta L=\frac{P(0,T_1)}{P(0,T_2)}$
04
Take the local limit
With continuous compounding, shrink the interval to obtain the instantaneous forward as the derivative of log discount.
$f(0,T)=-\partial_T\ln P(0,T)$
Numerical checkIntegrating f from 0 to T must reconstruct -ln P(0,T).

Forward rates are no-arbitrage coordinates of today’s curve; forecast interpretation is a separate economic claim.

Inputs

z(T): continuous zero rate
f(0;T_1,T_2): interval forward
f(0,T): instantaneous forward
δ(T_1,T_2): accrual fraction

Assumptions and limits

Forward curves amplify node noise and interpolation artifacts.
Credit-sensitive term rates are not interchangeable with overnight forwards.
Risk premia and convexity separate forwards from expectations.

Formula · Short derivation

Instantaneous forward

f(0,T)=-\partial_T\ln P(0,T)

The local slope of log discount factors; interpolation determines its numerical behaviour.

Open in Analytics

05

MODEL / PRICING

Invert, fit, and reprice the market instruments.

METHOD

Construct forwards from one internally consistent discount representation and expose both interval and instantaneous diagnostics.

CALIBRATION

Forwards inherit the fitted curve and interpolation. Validate quote repricing and inspect forwards between every pair of adjacent pillars.

Implementation with current QuantLib

Query zeroRate and forwardRate from the same YieldTermStructure instance with explicit compounding, frequency and day count. Never compare outputs without aligning those arguments.

API authority: upstream QuantLib reference pinned in the source registry.

06PYTHON IMPLEMENTATIONOpen the implementation and checks.

ARCHITECTURE

Parse dated market inputs and conventions at the boundary.
Build deterministic curve objects in the framework-free quant layer.
Return PV, repricing residuals and sensitivities together.
Test inversion, par conditions, monotonic dates and invalid domains.

PYTHON 3 · NUMPY / SCIPY

Zero-to-forward consistency

Compute interval forwards and reconstruct the terminal discount factor.

REUSABLE EXAMPLE

01from __future__ import annotations

02

03import math

04

05def forward(p1: float, p2: float, dt: float) -> float:

06 if min(p1, p2, dt) <= 0:

07 raise ValueError("positive discounts and interval required")

08 return math.log(p1 / p2) / dt

09

10p1 = math.exp(-0.025 * 2.0)

11p2 = math.exp(-0.034 * 5.0)

12f = forward(p1, p2, 3.0)

13reconstructed = p1 * math.exp(-f * 3.0)

14assert abs(reconstructed - p2) < 1e-12

15print(f"2y5y continuous forward={f:.4%}")

07

INTERACTIVE LAB

Move the quote and inspect every linked representation.

AVERAGE ZERO VS MARGINAL FORWARD

Zero/forward curve laboratory

Switch normal, inverted and humped regimes; compare averages, marginal forwards and discount-factor reconstruction.

SYNTHETIC · CONTROLLED SCENARIOS

SCENARIO INTENSITY70%

2Y zero2.72%

10Y zero3.62%

Max forward4.19%

Annualised rate by Maturity (years)

Normal: Rising zeros and forwards.

zero rate
segment forward

Maturity (years): 0.1Y. zero rate: 2.229%. segment forward: 2.229%.

Use Left/Right or Up/Down arrows to inspect values; Home and End jump to the bounds.

08

FRONT OFFICE

Where the model meets the book.

ON THE DESK

“A forward is what today’s curve locks—not an unqualified prediction of where fixing will print.”

VISIBLE INPUTS

discount pillars

interpolation rule

compounding

day-count basis

forward interval

CALIBRATION

Forwards inherit the fitted curve and interpolation. Validate quote repricing and inspect forwards between every pair of adjacent pillars.

RISK

slope and butterfly risk

interpolation risk

forward-start basis

convexity

DAILY WORKFLOW

normalize discount factors
derive zero view
derive forwards
inspect jumps
reconstruct and reconcile

Production failure modes

mixed rate conventions
negative year fraction
node duplicates
forward spikes hidden by zero-rate plot

09MACRO CONNECTIONOpen the transmission channel.

MACRO CONNECTION

Forward curves as priced policy paths

Central-bank expectations move the front end, while term premia, inflation uncertainty and supply shape longer forwards.

01Macro releasetransmits

changes expected policy timing

02Discount nodestransmits

reprice by maturity

03Forward curvetransmits

localises the move

04Trade expressionoutput

selects level, slope or butterfly

10COMMON PITFALLSOpen the failure checklist.

01

Calling the forward curve a pure forecast.

02

Comparing simple and continuous forwards directly.

03

Using zero-rate interpolation without inspecting implied forwards.

04

Dropping the accrual basis from a term-rate payoff.

11SOURCES / FURTHER READINGOpen sources and continue the track.

research

Interest-rate products, term structures and short-rate lectures

Research map for the rates progression and numerical experiments; all platform prose and code are original.

Source: Financial Engineering: Interest Rates & xVA
Author: L. A. Grzelak
Ref: main

OPEN ORIGINAL SOURCE ↗

research

Stochastic processes, Monte Carlo and model-calibration lectures

Mathematical cross-reference for stochastic dynamics and implementation checks.

Source: Computational Finance Course
Author: L. A. Grzelak
Ref: main

OPEN ORIGINAL SOURCE ↗

implementation reference

Current term structures, indexes, rate helpers, instruments, engines and tests

Implementation authority for production abstractions; Academy derives the mathematics before introducing library objects.

Source: QuantLib upstream
Author: QuantLib contributors
Ref: v1.42.1

OPEN ORIGINAL SOURCE ↗

Locking a future borrowing rate with zero-coupon claims

Buy a T₂ unit payoff

Normalize at the forward start

Invert the accumulation factor

Take the local limit

Forward curves as priced policy paths

Interest-rate products, term structures and short-rate lectures

Stochastic processes, Monte Carlo and model-calibration lectures

Current term structures, indexes, rate helpers, instruments, engines and tests