International Finance

The Financing Decision: Where and How to Borrow

Main issues

• When does the currency of borrowing matter?

• The financing irrelevance result: under CIP, currency choice is irrelevant

• Four frictions that break irrelevance: basis, taxes, natural hedging, market access

• Instruments for the financing decision: FRAs, interest rate swaps, cross-currency swaps

• The Eurocurrency market and the LIBOR \(\to\) SOFR transition

• Putting it together: the Reverse Yankee trade

The Financing Puzzle

The second firm decision

We now move from risk management (Lectures 6–8) to the financing decision.

The question: Where should the firm borrow, and in what currency?

• The firm needs funding in its home currency (say USD)

• It can borrow in USD directly

• Or it can borrow in EUR, JPY, or any other currency

Does the choice matter?

This parallels Lecture 6 exactly. There, we asked “should the firm hedge?” and started with the MM benchmark showing hedging is irrelevant in frictionless markets. Here, we ask “does the borrowing currency matter?” and start with CIP showing currency choice is irrelevant. The intellectual structure is identical.

The irrelevance result

A US firm needs USD 10M for one year.

Route 1: Borrow USD directly

• Borrow USD 10M at \(r^{USD} = 5\%\)
• Repay USD 10.5M in one year

Route 2: Borrow EUR and swap to USD

• Borrow EUR at \(r^{EUR} = 3\%\), convert to USD at spot \(S_0\)
• Buy EUR forward at \(F_{0,1}\) to repay the EUR loan

Under CIP: \(F_{0,1} = S_0 \cdot \frac{1 + r^{USD}}{1 + r^{EUR}}\)

This is the financing analog of Lecture 6’s “hedged vs. unhedged” example. We use simple numbers to make the arithmetic transparent.

The math: both routes cost the same

Route 2 in detail:

• Borrow \(\frac{\text{USD } 10\text{M}}{S_0}\) EUR, convert to USD 10M at spot

• Repay: \(\frac{\text{USD } 10\text{M}}{S_0} \times (1 + r^{EUR}) \times F_{0,1}\)

Substitute the CIP forward rate:

\[\frac{10\text{M}}{S_0} \times (1 + r^{EUR}) \times S_0 \times \frac{1 + r^{USD}}{1 + r^{EUR}} = 10\text{M} \times (1 + r^{USD})\]

All-in cost of Route 2 = \(r^{USD}\). Same as Route 1.

The EUR rate, the spot rate, and the forward rate all cancel out.

Financing irrelevance

This is the financing analog of MM for hedging. Under CIP, the currency you borrow in is irrelevant — the all-in cost is always your home-currency rate. This means you cannot “save” by borrowing in a low-rate currency. The forward rate exactly offsets the interest rate advantage. We showed this for CIP in Lecture 4 — now we see its direct implication for the financing decision.

The financing puzzle

If borrowing currency doesn’t matter, why do firms care?

Why do US corporates issue EUR bonds? Why do Japanese firms borrow in USD?

The answer is the same as in Lecture 6: frictions.

Just as Modigliani–Miller irrelevance for hedging breaks down with distress costs, taxes, and agency problems…

…financing irrelevance breaks down when CIP doesn’t hold exactly, or when other frictions create wedges.

Four Frictions That Break Irrelevance

Friction 1: The cross-currency basis

We saw in Lecture 4 that CIP deviations are small but persistent.

The cross-currency basis measures the wedge:

\[\text{basis} = \text{implied rate from FX swap} - \text{direct borrowing rate}\]

• A negative basis means synthetic USD is more expensive than direct USD borrowing

• A positive basis means synthetic USD is cheaper

When basis \(\neq\) 0, Route 1 \(\neq\) Route 2.

This is the most important friction. The basis is typically negative for USD (strong structural demand for USD funding globally), which means borrowing in EUR and swapping to USD can actually be cheaper than borrowing in USD directly — depending on credit spread differentials. This is what drives the Reverse Yankee trade.

The basis in practice

Basis proxy = difference in 3M interbank term premia (US minus foreign), in basis points. Source: FRED (OECD interbank rates).

This is the same proxy construction as Lecture 4. The key observation: the basis has been persistently negative since the GFC, especially for JPY/USD. This reflects global demand for USD funding and post-crisis regulatory constraints (Basel III leverage ratio, net stable funding ratio) that limit bank arbitrage. The negative EUR/USD basis since ~2014 is what opened the Reverse Yankee window.

Friction 2: Tax asymmetries

Even when CIP holds, the tax treatment of the two routes may differ.

• Interest payments on direct borrowing are tax-deductible in the borrower’s jurisdiction

• FX gains/losses on foreign-currency debt may be taxed differently (income vs. capital gains)

• The forward contract in Route 2 may generate gains/losses taxed at a different rate or in a different period

Example: A firm borrows in a low-rate currency. The currency appreciates. The FX loss on debt repayment is a capital loss — but the interest savings were ordinary income.

Tax asymmetries can make one route systematically cheaper after tax, even when CIP holds before tax.

Friction 3: Natural hedging

If the firm has revenues in EUR, borrowing in EUR creates a natural hedge.

• EUR revenues service EUR debt directly

• No need for forward contracts or swaps

• Reduces transaction costs, basis risk, and counterparty exposure

This is an operational benefit, not a financial arbitrage.

The logic: Match the currency of your liabilities to the currency of your assets.

• A European subsidiary generating EUR revenue should be funded in EUR

• A US exporter with JPY receivables might borrow in JPY

Natural hedging connects the financing decision back to the risk management decision from Lectures 6-7. By matching currencies, the firm reduces its need for derivative hedging. This is especially valuable for operating exposure, which is hard to hedge with financial instruments.

Friction 4: Market access and comparative advantage

Firms may face different credit spreads in different markets.

Why?

• Information advantage: Local banks know local firms better

• Regulatory barriers: Some bond markets restrict foreign issuers

• Investor base: US market is deepest; EUR market has grown rapidly

• Name recognition: Known issuers get tighter spreads at home

If each firm borrows where it has an advantage and swaps into its desired currency, both can gain. This is the comparative advantage argument for swaps.

Summary: why financing currency matters

Friction	Mechanism
Cross-currency basis	CIP deviations create cost wedges
Tax asymmetries	Different tax treatment of interest vs. FX gains
Natural hedging	Currency matching reduces hedging needs
Market access	Comparative advantage in credit markets

Under pure CIP with no frictions: currency choice is irrelevant.

With real-world frictions: currency choice affects firm value.

This parallels the hedging summary from Lecture 6. The CF vs. DR column uses the same diagnostic we’ve been applying since Lecture 1. Most financing frictions are cash flow effects — they change the all-in cost of funding. Market access is the exception: if a firm can access cheaper funding through comparative advantage, this reduces its cost of capital (a discount rate effect).

Instruments for the Financing Decision

The corporate treasurer’s toolkit

The financing decision requires instruments that transform the currency and interest rate profile of debt.

Three key instruments:

1. Forward Rate Agreements (FRAs): Lock in a future borrowing rate

2. Interest Rate Swaps (IRS): Convert fixed to floating (or vice versa)

3. Cross-Currency Swaps (CCS): Convert the currency of debt

Each instrument solves a specific corporate problem.

Interest rate conventions

Money market rates are quoted as simple (annualized) interest rates:

\[\text{Interest earned} = r \times \tau \times N\]

where \(r\) is the annualized rate, \(\tau\) is the year fraction, and \(N\) is the notional.

Year fractions depend on the day-count convention:

Convention	\(\tau\)	Used for
ACT/360	Actual days / 360	USD, EUR, JPY money markets
ACT/365	Actual days / 365	GBP money markets

Example: Invest USD 10M for 90 days at \(r = 5\%\) (ACT/360):

\[\text{Interest} = 0.05 \times \frac{90}{360} \times 10\text{M} = \text{USD } 125{,}000\]

Tenor notation

Tenor = the length of the borrowing/deposit period.

• “3M rate” = an annualized rate for a 3-month borrowing period

• “6M rate” = an annualized rate for a 6-month borrowing period

FRA notation: A 3x6 FRA means:

• The borrowing period starts in 3 months

• The borrowing period ends in 6 months

• So the underlying tenor is 3 months (from month 3 to month 6)

Similarly: a 6x12 FRA locks in a 6-month rate starting in 6 months.

From Forward-Forwards to FRAs

The problem: “We need to borrow in 6 months. Can we lock in the rate today?”

The old solution: Forward-Forward (FF) contracts

• Actual deposit/loan made at future date \(t_1\), repaid at \(t_2\)

• On-balance-sheet: the deposit is real, affecting leverage ratios

The modern solution: Forward Rate Agreement (FRA)

• Same economic function, but notional only

• Off-balance-sheet: no actual deposit, just a cash settlement

• Settled at \(t_1\) (not \(t_2\)), at the present value of the interest difference

FRAs replaced FFs because they don’t inflate the balance sheet.

FRA mechanics

FRA settlement

At settlement date \(t_1\), the FRA pays (or receives):

\[\text{Settlement} = \frac{(r_{market} - r_{FRA}) \times \tau \times N}{1 + r_{market} \times \tau}\]

where \(\tau = (t_2 - t_1)\) as a year fraction, \(N\) = notional.

• If \(r_{market} > r_{FRA}\): FRA buyer receives (locked in a lower rate)

• If \(r_{market} < r_{FRA}\): FRA buyer pays (market rate fell below locked rate)

Discounting: Settlement is at \(t_1\), not \(t_2\), so the payment is the PV of the interest difference.

The settlement formula may look complex, but the logic is simple: the FRA compensates you for the difference between the rate you locked in and the rate that actually prevails. The discounting to t1 is because settlement happens at the start of the notional period, not the end.

FRA pricing: no-arbitrage

The FRA rate is determined by the yield curve via no-arbitrage.

If the 6-month rate is \(r_6 = 4\%\) and the 12-month rate is \(r_{12} = 5\%\):

\[1 + r_{12} = (1 + r_6 \times 0.5) \times (1 + r_{FRA} \times 0.5)\]

\[r_{FRA} = 2 \times \left[\frac{1 + r_{12}}{1 + r_6 \times 0.5} - 1\right] \approx 5.9\%\]

The 6x12 FRA rate is the forward rate implied by the term structure. FRAs now reference SOFR (not LIBOR).

The no-arbitrage logic is identical to what we used for FX forwards in Lecture 4. There, CIP linked spot, forward, and interest rate differential across currencies. Here, the same logic links short rates, long rates, and forward rates within a single currency. It’s all the same principle: no-arbitrage pins down the forward price.

Interest rate swaps

The problem: “We have floating-rate debt but want fixed-rate certainty.”

An interest rate swap exchanges:

• Fixed-rate payments (the “swap rate”) for

• Floating-rate payments (SOFR, reset each period)

No principal is exchanged — only the net interest difference.

IRS mechanics

Each period, only the net cash flow is exchanged:

\[\text{Net payment} = (\text{Fixed rate} - \text{SOFR}_t) \times \tau \times N\]

• If Fixed > SOFR: payer of fixed pays the difference

• If Fixed < SOFR: payer of fixed receives the difference

The swap rate is set so the swap has zero value at inception.

A swap is a portfolio of FRAs: Each period’s settlement equals an FRA at the corresponding forward rate.

IRS valuation over time

At inception: value = 0 (by construction).

As interest rates change, the swap gains or loses value:

\[V_{swap} = PV(\text{remaining floating leg}) - PV(\text{remaining fixed leg})\]

• If rates rise: fixed payer gains (locked in a lower rate; floating payments received increase)

• If rates fall: fixed payer loses (locked in a higher rate; floating payments received decrease)

The mark-to-market value matters for:

• Counterparty risk management (collateral calls)

• Accounting (hedge effectiveness testing)

• Unwinding the swap before maturity

Cross-currency swaps

The problem: “We can borrow cheaply in USD, but we need EUR funding.”

A cross-currency swap (CCS) exchanges:

1. Principals in two currencies (at inception and maturity)

2. Periodic coupons in two currencies (during the life)

Key difference from IRS: Principal IS exchanged, because the currencies differ.

The principals are exchanged at the same spot rate at both inception and maturity — so there is no FX risk on the principal exchange itself.

CCS cash flows

CCS: where the basis enters

In a CCS, the two coupon rates are swap rates in each currency adjusted for the cross-currency basis.

\[\text{All-in cost of synthetic EUR via CCS} = r^{EUR}_{swap} + \text{basis}\]

• When basis < 0 (typical for EUR/USD): obtaining EUR synthetically is cheaper than direct EUR borrowing

• When basis > 0: obtaining EUR synthetically is more expensive

This is why the basis is the primary friction in the financing decision.

The basis enters the CCS pricing directly. When an investor or corporate uses a CCS to swap USD funding into EUR, the negative basis means they effectively receive EUR funding at a rate below the “fair” EUR swap rate. This is the mirror image of what we discussed in Lecture 4 — where the negative basis made it expensive to obtain USD synthetically. In a CCS, it means EUR is cheap to obtain synthetically from USD.

CCS valuation over time

At inception: Value = 0 (principals have equal market value; coupon streams are set to offset).

Over time, value changes with:

• Spot exchange rate (affects the relative value of the two principal amounts)

• Interest rates in both currencies (affects the PV of remaining coupons)

• Basis (affects the relative value of the two legs)

Example: A firm swaps EUR 8M at 3% for USD 10M at 5% (spot = 0.80).

Two years later: spot = 0.75, EUR rates fall to 2%, USD rates fall to 4%.

• USD leg PV \(\approx\) USD 10.3M (discounted remaining flows)

• EUR leg PV \(\approx\) EUR 8.2M = USD 10.9M at new spot

• Swap value \(\approx\) USD –0.6M (negative for the firm)

The valuation follows exactly the same logic as forward contract valuation from Lecture 4. The swap is worth zero at inception. As rates and FX move, one side gains and the other loses. The CCS has more moving parts than an IRS because both currencies’ rates and the spot rate all matter.

The IBM–World Bank swap (1981)

The birth of the modern swap market.

• IBM had existing debt in CHF and DEM, but wanted USD exposure
  • Strong credit in European markets; had borrowed there at attractive rates

• World Bank wanted to borrow in CHF and DEM for its lending operations
  • Had already borrowed heavily in those markets; investors wanted diversification

The solution:

• World Bank issued new USD bonds

• IBM and World Bank swapped their payment obligations

• IBM paid USD coupons (on World Bank’s bonds); World Bank paid CHF/DEM coupons (on IBM’s bonds)

Both gained: Each borrowed where they had a comparative advantage, then swapped into their desired currency.

This is the founding case of the cross-currency swap market. Salomon Brothers intermediated. The logic is pure comparative advantage: IBM had exhausted its USD borrowing appetite but had cheap access to CHF/DEM; World Bank had the reverse problem. By swapping, both ended up with their desired currency exposure at a lower cost than direct borrowing. This was 1981 — before swaps were standardized. The market grew from zero to trillions within a decade.

The Eurocurrency Market and the LIBOR \(\to\) SOFR Transition

The Eurocurrency market

Definition: Deposits and loans denominated in a currency outside its home country.

• Eurodollars: USD deposits in London (or anywhere outside the US)

• Euroyen: JPY deposits outside Japan

• “Euro” in Eurocurrency \(\neq\) the EUR currency

Why it exists:

• No reserve requirements (cheaper for banks)

• No deposit insurance (lower regulatory cost)

• Large transactions between high-grade counterparties

• Historical origins: Cold War — Soviet USD holdings kept in London to avoid US seizure

The Eurocurrency market is where interbank rates are determined.

From LIBOR to SOFR

LIBOR (London Interbank Offered Rate) was the dominant benchmark for decades.

• Panel of banks submitted estimates of their borrowing costs

• Published daily for multiple currencies and tenors

• Referenced in trillions of dollars of contracts (swaps, FRAs, loans, mortgages)

What went wrong:

• The underlying market (unsecured interbank lending) shrank after 2008

• Banks were increasingly guessing, not reporting actual transactions

• Manipulation scandal (2012): traders colluded to move LIBOR for profit

Result: LIBOR was no longer credible.

The LIBOR scandal is one of the largest financial frauds in history. Traders at major banks (Barclays, UBS, Deutsche Bank, and others) were found to have manipulated submissions to benefit their derivatives positions. The fines totaled over $9 billion. But the deeper problem was structural: the interbank market had dried up, so LIBOR was being set by judgment rather than transactions.

SOFR and its equivalents

SOFR (Secured Overnight Financing Rate) replaced USD LIBOR.

Feature	LIBOR	SOFR
Basis	Survey (expert judgment)	Transactions (>USD 1T daily)
Security	Unsecured (interbank credit)	Secured (Treasury repo)
Tenor	Overnight to 12 months	Overnight only
Credit component	Yes (bank credit risk)	No (risk-free-ish)
Manipulation risk	High (small panel)	Low (massive volume)

Transition: LIBOR ended 2021–2024. Replaced by SONIA (GBP), STR (EUR), TONA (JPY).

What changed for corporate treasurers

• Floating-rate debt now references compounded SOFR (not LIBOR)

• FRAs and swaps reference SOFR — mechanics unchanged, benchmark different

• No credit component in SOFR: In stress, LIBOR spiked (bank credit risk); SOFR may fall (flight to Treasuries)

• Term structure: LIBOR had built-in term rates (3M, 6M). SOFR is overnight — term rates must be constructed (CME Term SOFR)

For this course: The instruments we’ve discussed (FRAs, IRS, CCS) work identically with SOFR. The economics are unchanged; only the reference rate is different.

Putting It Together: The Reverse Yankee Trade

Why are US firms borrowing in EUR?

In 2025, US corporates have issued a record EUR 42 billion in EUR-denominated bonds (“Reverse Yankee” issuance).

Why would a US firm with USD costs borrow in EUR?

The setup: A US firm needs USD 1 billion for 5 years.

Three routes:

1. Issue USD bonds directly

2. Issue EUR bonds + enter a cross-currency swap to USD

3. Issue EUR bonds + roll FX forwards quarterly

Route 1: Direct USD issuance

• Issue USD 1B, 5-year bond

• Pay USD Treasury rate + credit spread

• All-in cost: say 5.0% + 1.20% = 6.20%

Simple. No FX risk. But is it cheapest?

Route 2: EUR bond + cross-currency swap

• Issue EUR bond at Bund rate + credit spread: say 2.8% + 0.85% = 3.65%

• Enter 5-year CCS: receive EUR coupons, pay USD coupons

• The CCS rate adjusts for the interest rate differential and the basis

• With basis around –7 to –15 bp, the all-in USD cost is:

\[\underbrace{r^{USD}_{swap}}_{\text{from CIP}} + \underbrace{\text{EUR credit spread}}_{\text{tighter than USD}} + \underbrace{\text{basis}}_{\text{negative} \to \text{benefit}} \approx 6.07\%\]

Saving: about 13 bp vs. direct USD issuance. On USD 1B, that’s USD 1.3M per year for 5 years.

The comparison

The savings come from two sources: (1) EUR credit spreads are tighter than USD credit spreads for investment-grade issuers — the European credit market is less deep but more competitive for high-quality names; (2) the negative cross-currency basis means the swap “gives back” some basis points. Route 3 (rolling forwards) captures similar economics but adds rollover risk — each quarter, the firm must re-enter FX forwards at potentially different basis levels.

Why does this opportunity exist?

If this is a “free lunch,” why doesn’t arbitrage eliminate it?

It’s not free:

• Balance sheet constraints: Banks that would arbitrage the basis face leverage ratio limits (Basel III)

• Counterparty risk: 5-year CCS creates significant bilateral exposure

• Roll risk (Route 3): Forward rates and basis can move against you at each rollover

• Credit spread risk: EUR spread advantage can reverse

The Reverse Yankee window opened post-2014 as:

1. EUR rates fell toward (and below) zero — ECB rate cuts and quantitative easing

2. The basis widened (more demand for USD funding, fewer arbitrageurs)

3. The US investment-grade credit market became relatively more expensive

2025 is a record year because all three conditions persist.

This connects directly to Lecture 4’s discussion of why CIP deviations persist. The same forces that keep the basis nonzero — bank balance sheet constraints, regulatory costs, structural USD demand — are what make the Reverse Yankee trade profitable. If banks could arbitrage freely, the basis would close and the opportunity would vanish.

Market context

Metric	Value	Source
FX daily turnover	USD 9.5T	BIS 2025
CCS notional (2024)	>USD 7T	Clarus 2024
CCS volume growth	+21% notional	Clarus 2024
Reverse Yankee (2025)	EUR 42B (record)	ECB/ING 2025
EUR/USD 3M basis	around –7 bp	CME 2024
JP insurer hedge ratio	60% to 40%	BIS 2025

The swap and basis markets are among the largest and most active in global finance.

Internal Financing: A Brief Note

Moving cash within the multinational

Multinational firms also have internal mechanisms for moving funds across borders:

• Intercompany loans: Parent lends to subsidiary (or vice versa)

• Transfer pricing: Prices on intra-firm goods/services shift profits between jurisdictions

• Royalties and management fees: Payments for IP or services between affiliates

• Dividend policy: Timing and size of subsidiary-to-parent dividends

These tools serve tax optimization and political risk management, not pure financing.

Full treatment is beyond this course, but corporate treasurers should know these mechanisms exist alongside external instruments.

Summary and Connections

Key takeaways

1. Financing irrelevance: Under CIP in frictionless markets, borrowing currency doesn’t matter. This is the financing analog of Modigliani–Miller for hedging.

2. Four frictions break irrelevance: Cross-currency basis (primary), tax asymmetries, natural hedging, and market access.

3. Instruments transform debt: FRAs lock in future rates, interest rate swaps convert fixed/floating, cross-currency swaps convert currencies. Each solves a specific corporate problem.

4. LIBOR is dead; SOFR lives: The reference rate changed, but the economics didn’t. All instruments now reference transaction-based overnight rates.

5. The Reverse Yankee trade demonstrates all four frictions at work: basis advantage, spread differentials, and market access create real savings for US corporates borrowing in EUR.

Connections to the course

• Lecture 4 (CIP/Basis): The basis we introduced there now drives the financing decision. Same data, new application.

• Lecture 6 (Why Hedge?): The MM-style irrelevance argument is identical in structure. Frictions break irrelevance; the question is always which frictions matter most.

• Lecture 7 (Exposure): Natural hedging connects financing to exposure management. Currency matching is simultaneously a financing choice and a risk management strategy.

• Looking ahead: The financing decision feeds directly into cross-border valuation (Lecture 10). The cost of capital depends on how the firm funds itself — and the basis affects that cost.

The key intellectual point is that the financing decision uses exactly the same analytical framework as the risk management decision: start with the frictionless benchmark (CIP irrelevance / MM irrelevance), identify the frictions that break it, and then make the optimal choice given those frictions. This is the course’s operating system at work.