Kumu / Finance & Securities

International Finance & Securities

The FX Hedging project runs on AI + GitHub throughout. Your firm is owed millions in a foreign currency, a year from now. Frame the exposure, design the hedge model on paper, have an AI build it from your spec, load live market data, then validate the whole chain by hand before making the call a CFO would sign off on.

Upper undergrad · 6 stages Design → Build → Validate

Stage 0 — get GitHub working, before the project

Every stage of this project is handed in by committing it to a public repository of your own, so the repository comes first: before the brief, before the spec, before any Excel. One sitting, once, and the rest of the term is about finance instead of tooling.

It comes in two parts: the account, then the workspace.

1

Part 1 — Get a GitHub account never used GitHub? start here

Create the account and learn the three verbs: add → commit → push. If you already have a GitHub account, use it and go straight to part 2. Any email works for your GitHub account; for the free GitHub Education benefits, add your school address under Settings › Emails and apply, which is optional and never graded.

account + mechanics
2

Part 2 — Set up your workspace

Stand up the public repo every stage of this project (and every project after it) lands in. Built to the standard, named for you, and finished before the brief starts. Already have one from another course? Bring it up to the standard instead.

the workspace

Part 1 in full: the Git mechanics walkthrough

Start Here → Git mechanics walks the whole workflow once: local versus remote, add → commit → push in the desktop app and on the command line, editing files straight on github.com, and fixes for the mistakes everyone makes the first time. Read it before part 2. The stage page assumes you can already save work to a repo, and does not re-teach the mechanics.

The learning path — stages 1 to 5

The spec comes before any Excel exists. You design the workbook (named ranges, tabs, formula logic, validation checks), and then an AI builds it from your document. The build tests your design; live market data (stage 4) tests the structure; the stage-5 validation tests everything. An AI builds the workbook; you specify, audit, populate, and verify by hand (why the work is split this way).

1

Executive Brief

Frame the FX receivable exposure for your CFO: what it is, why it's risky, and three hedge families with quick pros/cons. It is the project's engagement brief: a hypothesis committed before any modeling.

.md brief
2

Model Specification

Design the workbook before it exists: named-range contract, tab architecture, formula logic, validation rules. In stage 3, an AI builds exactly what you wrote.

.md spec
3

AI-Assisted Build + Audit

Generate the workbook from your own spec (any AI tool, or by hand), then audit the output ruthlessly: ≥3 documented findings.

.xlsx + audit
4

Market Data + Population

Replace placeholders with live, sourced, timestamped market data, and find out whether your model survives contact with reality.

data memo
5

Validate & Decision Memo

The capstone: a fresh LLM runs your spec + data memo cold; you reconcile, hand-verify, and ship the CFO decision memo.

validation + memo
0 of 6 stages complete

What every stage produces

Everything lands in one public portfolio repo, organized by capability and engagement rather than by course, so it keeps working after the semester ends. Stage 0 builds the standard skeleton; the project adds no folders to it. The spec and workbook live in capabilities/fx-hedging/, as in the Micro & Macro Economics course (an earlier version of the project used docs/specs/ and models/builds/).

StageWhat you produceWhere it lives
1 · BriefThe exposure framed for the CFO, and the plan for Stages 2–5docs/briefs/
2 · SpecThe workbook designed on paper: named ranges, tabs, formula logic, checkscapabilities/fx-hedging/
3 · BuildThe workbook an AI built from your spec, and your audit of itcapabilities/fx-hedging/ · analysis/
4 · DataLive, sourced, timestamped inputs, and the re-committed workbookdata/ · capabilities/fx-hedging/
5 · ValidateThe hand-verified reconciliation, and the decision memo the CFO signsanalysis/ · docs/decisions/

Specs define; memos answer. Every graded file is named YYYY-MM-DD-{scenario}-{type}, and each stage page states its own exact path.

Course tools

Worth opening before Stage 1, and returning to during the project.

Interactive lab

FX Hedging Lab

Forward, money-market, and put-option hedges, the covered-interest-parity check, and the ±5% sensitivity grid with payoff chart (the workbook you'll spec at Stage 2, made live). It's also your Stage 4 check-figure oracle.

Required before Stage 3

AI Tools Lab

Required pre-reading before your Stage 3 build. The field guide to coding LLMs: chat vs "code" tools, web vs desktop vs CLI, and why giving the AI your repo beats pasting into a chat box.

Setup

GitHub & your portfolio repo

The repository every stage lands in, and how it is organized: by capability and engagement rather than by course. The Git mechanics page covers local versus remote, committing, pushing, and how work is submitted.

AI tutor

Kumu knows this course

Ask the ✳ tutor to explain interest rate parity with your scenario's numbers, find the ambiguity in your spec before an AI builds the wrong thing from it, or check your read of a payoff chart, all of it logged for your prompt log.

The scenario

You're the treasury analyst. Your firm expects a large foreign-currency receivable (e.g., EUR 4.5M–20M, one year out, depending on your assigned scenario). If the euro weakens before settlement, your USD proceeds shrink, and the CFO wants to know what to do about it. Four tools are on the table:

StrategyWhat it doesThe trade-off
Forward contractLocks today's forward rate for the settlement dateTotal certainty, zero upside
Money-market hedgeBorrow FC now → convert at spot → invest USD (a synthetic forward)Should ≈ the forward (interest rate parity); uses balance-sheet capacity
Put optionBuys a floor under your USD proceeds; keeps the upsideThe premium (insurance isn't free)
No hedgeTake whatever the spot rate is at settlementFull exposure, both directions

Try all four live in the FX Hedging Lab, with the same named ranges (FC_AMT, S0_in, F0_in…), same sensitivity range as your model. At stage 4 the lab becomes your check-figure oracle: enter your live inputs, and if the lab's outputs disagree with your workbook, one of them is wrong; find out which before you submit.

House rules