What makes one currency a funder and another a target?
Not every low-yield currency gets borrowed, and not every high-yield currency gets bought. The roles are assigned by a short and fairly stable list of properties — and the list explains why the same handful of names keeps appearing.
What a funding currency needs
- A low policy rate, which is the obvious one.
- A deep, liquid money market, so that very large amounts can be borrowed and repaid at a tight price on any day.
- Free convertibility, with no capital controls to trap the position.
- No tendency to rally when risk assets fall, because a funding currency that spikes in a sell-off moves against the borrowed leg at exactly the worst moment. This one is routinely underweighted — and the classic funders fail it, which is precisely why the previous lesson's unwind runs the way it does.
The classic funders are the yen and the Swiss franc. But the role is a function of rates, not of nationality: the euro was widely used as a funding currency through the ECB's negative-deposit-rate years, and the dollar served the same purpose for much of 2009 to 2015. Change the differential and the roles rotate.
What a target currency needs
A high rate, obviously — but also enough market depth to absorb size, and enough convertibility to get out. This is why some of the world's highest policy rates attract little carry positioning: the currency is non-deliverable, or the exit is legally uncertain. Where a currency is not freely deliverable, exposure is often taken through non-deliverable forwards (NDFs), which settle a cash difference in dollars rather than exchanging the currency itself.
Historically the developed-market targets have been the Australian and New Zealand dollars; the emerging-market ones include the Brazilian real, Mexican peso, South African rand and Turkish lira.
The differential is compensation, not a gift
This is the discipline that keeps the whole subject honest. A large nominal interest rate is usually paying for something.
Worked example. Country A offers 12.00% with expected inflation of 9.5%. Country B offers 4.50% with expected inflation of 2.5%.
Real rate A: 12.00 − 9.5 = 2.5%
Real rate B: 4.50 − 2.5 = 2.0%
The headline gap is 7.50 points. The real gap is half a point. Most of that eye-catching differential is compensation for money that is losing purchasing power — and, as Unit 2's relative purchasing power parity will argue, an inflation differential is precisely what tends to show up in the exchange rate over long horizons.
On top of inflation, a high rate may be compensating for devaluation risk, capital-control risk, default risk, or simply the difficulty of getting out in a hurry. None of that is free money; all of it is a risk premium with a name.
Carry per unit of risk
Practitioners therefore rarely look at the differential alone. They compare it to the volatility of the pair — carry divided by annualised volatility. A 4.0 point differential on a pair with 8% annualised volatility is a very different proposition from the same 4.0 points on a pair with 25% volatility, even though the interest income is identical. The measure is descriptive, not predictive, and it says nothing at all about the crash risk in the previous lesson, which is exactly the risk that volatility measures understate.
Try it now
- The policy rates of three central banks, the Fed, the ECB and the Bank of England, are below, latest day each. Rank them from lowest to highest, and say what happens to your ranking if you swap the end of the Fed's band you used.
- The lowest and the highest meet in EUR/USD. Switch the chart to Monthly, look at the last two years, and note its volatility informally: how large are the typical monthly moves? Measure three of them.
- For the high-rate economy, subtract inflation from the policy rate. The US consumer-price release of September 2026, from the economic calendar, is below. How much of the headline differential survives?