SmartCarry trades built the Kiwi dollar’s reputation as a high-yield darling of global forex markets. The principle is straightforward — borrow where rates are cheap, invest where they are not — but the NZD/JPY chart from 2007 to 2008 tells the full story: three years of steady escalator profits followed by an elevator drop that erased the lot in eight weeks.
We break down the mechanics, the maths, and the position sizing that separates traders who collect carry from those who collect margin calls.
The Mechanics of a Carry Trade
Borrowing Cheap, Parking Expensive
A carry trade is, at its core, an arbitrage on the price of money. You borrow in a currency where interest rates are low, convert that capital into a currency where rates are higher, and collect the difference. The concept is simple enough to explain in a sentence. The execution is where most traders either make consistent returns or learn an expensive lesson about risk asymmetry.
Take NZD/JPY as the textbook example. For most of the past two decades, the Reserve Bank of New Zealand has maintained an Official Cash Rate significantly above the Bank of Japan, which spent years at or near zero. A trader going long NZD/JPY is effectively borrowing yen at close to nothing and parking the proceeds in NZ dollars earning a materially higher rate.
In practice, this shows up in your trading account as a daily swap credit. On a standard 100,000-unit NZD/JPY position, a 4% annualised rate differential translates to roughly NZD 11 per day — credited to your account at the daily rollover, typically around 5am NZT. That is not going to fund your retirement on its own, but across a portfolio of positions held for weeks or months, it accumulates. The appeal of carry is not any single day but the compounding effect of collecting that differential while the exchange rate cooperates.
Why the NZD Has Been a Carry Favourite for Decades
New Zealand has sat at the high-yield end of the G10 spectrum for most of its modern monetary history, and the reasons are structural rather than accidental.
The RBNZ was one of the first central banks globally to adopt explicit inflation targeting, and it has consistently set rates to keep CPI within its 1-3% target band. For a small, open economy with a floating currency, that has typically meant higher rates than larger economies with deeper capital markets and lower inflation volatility. NZ imports most of its manufactured goods and a fair portion of its fuel, which creates persistent inflationary pressure that the RBNZ manages with interest rate settings that would look aggressive by European or Japanese standards.
Add to that a chronic current account deficit — New Zealand has run a deficit in most years since the 1970s — and the structural picture becomes clearer. Higher rates serve a dual purpose: controlling domestic inflation and attracting the foreign capital inflows needed to fund the gap between what the country earns and what it spends abroad.
The Kiwi dollar, in effect, has to pay a premium to attract the capital it needs. This is why NZD has been a perennial favourite on the long side of carry trades. It is not a temporary policy choice but a consequence of NZ economic structure.
Swap Rates, Rollover, and the Numbers That Actually Matter
The interest rate differential between two central banks sets the theoretical framework for carry. The swap rate your broker actually pays you is a different number, and the gap between the two deserves your attention.
Brokers calculate swap rates using the interbank forward points for each currency pair, then apply their own markup. That markup varies significantly — we have seen differences of 30-40% between the best and worst swap rates offered by NZ-accessible brokers on the same pair. If you are running a carry strategy, the broker you choose is not a minor detail. It directly affects your daily income.
Wednesday is triple-swap day on most platforms, accounting for the weekend settlement delay. Your Wednesday rollover credit (or debit) covers Saturday and Sunday as well, which means a NZD/JPY carry position earns three days of swap in a single overnight hold. Some traders specifically enter carry positions on Tuesday evening to capture the triple rollover, though the market prices this in more efficiently than it once did.
Before entering any carry trade, pull up your broker swap rates and do the arithmetic. Calculate the daily credit in your account currency, multiply by 30 for a monthly figure, and compare that to the average daily range of the pair. If one bad day can erase a month of carry income, your position sizing needs to reflect that reality.
When Carry Trades Print Money
The Goldilocks Environment: Low Vol and Positive Sentiment
Carry trades are not all-weather strategies. They thrive in specific conditions, and recognising those conditions before you commit capital is half the battle.
The ideal carry environment combines low volatility, stable or improving risk appetite, and a monetary policy divergence that markets expect to persist. When the VIX sits below 15 and global equity markets are grinding higher, carry trades tend to perform well because the funding currency (typically JPY or CHF) stays weak while the high-yielder benefits from capital inflows seeking return.
The 2004 to 2007 period was the golden age of NZD carry. The RBNZ pushed the OCR to 8.25% while the BOJ held at effectively zero. NZD/JPY climbed from the low 60s to above 97 over three years, delivering both capital gains and daily swap income. Traders who held through that period collected carry and price appreciation simultaneously — the dream scenario.
A similar, if more modest, window opened in 2017-2018 when the NZD offered a comfortable yield premium over JPY and EUR in a low-volatility environment. These periods share common features: central bank policy is predictable, geopolitical risk is contained, and global growth is positive. Recognising these conditions in real time is more art than science, but the indicators are observable if you know where to look.
Reading the Yield Differential Before You Enter
Before entering a carry trade, the yield differential needs to justify the exchange rate risk you are taking on. This is not a philosophical question — it is a specific calculation.
Start with the current and expected policy rates. The RBNZ publishes its own OCR track projections, and OIS (Overnight Index Swap) markets price in the expected path for both the funding and target currencies. If the RBNZ is projected to cut rates over the next twelve months while the BOJ is expected to hold, the carry differential is narrowing, and your trade is swimming against the current.
The break-even calculation makes the risk concrete. Take your annual carry income in pips — if the swap rate gives you 5 pips per day, that is roughly 1,825 pips per year on a standard lot. Now compare that to the historical annual range of the pair. NZD/JPY has an average annual range of approximately 1,500-2,000 pips in normal years, and considerably more in crisis years. Your annual carry income barely covers a single directional move across the normal range.
This is the fundamental tension of carry trading: the income is real but modest relative to the price risk. The break-even drawdown — the number of pips the pair can move against you before your carry profit turns negative — is the single most important number to calculate before you enter.
NZD Pairs Worth Watching for Carry
Not all NZD pairs offer attractive carry, and the landscape shifts as central banks adjust their rate settings.
NZD/JPY remains the classic carry pair for Kiwi traders. Japan has maintained ultra-low rates for decades, and even the BOJ incremental tightening in recent years has left a substantial gap with NZ rates. The pair is liquid, widely available, and well-understood. Its primary risk is that it correlates heavily with global risk sentiment — when equity markets sell off, NZD/JPY tends to fall hard as yen strengthens.
NZD/CHF offers a European alternative with similar carry dynamics. Swiss rates have historically been low, and the pair provides some geographic diversification away from Asian session risk. It is less liquid than NZD/JPY, which means wider spreads and occasionally less favourable swap rates.
NZD/USD carry depends entirely on the Fed-RBNZ rate gap, which fluctuates more than the BOJ gap. When the Fed runs rates well below the RBNZ — as it did from 2010 to 2014 — NZD/USD carry is attractive. When rates converge or the Fed overtakes, the carry disappears or reverses.
NZD/AUD is rarely a carry trade. The two economies share enough structural similarities that their central banks tend to move in rough parallel. Rate differentials exist but are usually too narrow to justify the position risk.
The Unwind: When the Elevator Door Opens
2008: Six Years of Carry Erased in Eight Weeks
The carry trade has a well-known asymmetry that traders describe as “escalator up, elevator down.” The 2008 Global Financial Crisis provided the definitive illustration.
From 2002 to mid-2007, NZD/JPY climbed steadily from the low 50s to above 97, delivering years of swap income alongside capital appreciation. Carry traders were collecting daily credits and watching their unrealised profits grow. It was, by any measure, a spectacular run.
Then Lehman Brothers collapsed, and the elevator doors opened. Between July and October 2008, NZD/JPY fell from 97 to below 50 — a decline of more than 48% in roughly twelve weeks. Five years of carry income was erased in two months. The speed was the killer. Carry unwinds are self-reinforcing: as the high-yield currency falls, leveraged carry positions hit stop losses or margin calls, forcing liquidation that pushes the currency lower still, triggering more stops. The liquidity that existed on the way up evaporated on the way down.
Traders who had been collecting a few pips of swap per day watched hundreds of pips of capital evaporate per hour. The maths is brutally simple: at 5 pips per day of carry income, a 2,000-pip adverse move erases 400 trading days of accumulated carry. Roughly eighteen months of patience, undone before lunch.
March 2020: The COVID Crash as a Carry Stress Test
The March 2020 COVID sell-off provided a modern stress test for carry trade resilience, and the comparison with 2008 is instructive.
NZD/JPY fell from around 73 in mid-February to below 60 in mid-March — a 17% decline in roughly four weeks. The VIX spiked above 80, liquidity contracted, and the familiar carry unwind dynamics reappeared. Anyone holding NZD carry positions without hedging or reduced sizing experienced significant drawdowns.
The critical difference from 2008 was the recovery timeline. Where the GFC unwind took years to repair, the COVID unwind largely reversed within three to four months. NZD/JPY was back above 70 by August 2020. Central bank intervention was faster and larger — the Fed moved to near-zero rates and launched massive QE within days, the RBNZ cut the OCR to 0.25% — and the market treated the event as a liquidity shock rather than a solvency crisis.
For carry traders, the lessons are nuanced. The initial unwind was almost as vicious as 2008 in terms of daily drawdowns. If you were overleveraged or running tight stops, you were forced out before the recovery. If you had conservative sizing and could weather the drawdown, the carry trade recovered within months rather than years.
The distinction matters for strategy design: surviving a carry unwind is primarily a position-sizing problem, not a prediction problem. You cannot reliably predict when the unwind will come, but you can ensure your account survives it.
Sizing Carry Positions That Survive the Shock
The Break-Even Drawdown Calculation
The break-even drawdown is the number that separates carry traders who understand their risk from those who merely hope things work out.
Suppose you are long one standard lot of NZD/JPY and collecting a net swap of 5 pips per day after your broker takes its cut. Over a month, that is approximately 150 pips of accumulated carry income. Over a quarter, roughly 450 pips. Over a year, around 1,825 pips — in theory.
Now consider that NZD/JPY moved more than 1,300 pips during the relatively calm period of January to April 2024. In a single quarter of unremarkable price action, the pair moved almost three times your quarterly carry income. During the COVID unwind, NZD/JPY fell approximately 1,300 pips in four weeks — nearly your entire annual carry income, gone in a month.
The break-even calculation forces you to confront this: your carry income per day is measured in single-digit pips, while adverse price moves are measured in hundreds. A 300-pip move against your position, which can happen in a single session during risk-off events, wipes out two months of carry. A 600-pip move, which is unremarkable in a trend, wipes out four months.
This is not an argument against carry trading. It is an argument for treating carry income as a secondary benefit of a position you would hold anyway based on technical or fundamental conviction, rather than the primary reason for the trade.
Position Sizing for Asymmetric Risk
Standard position sizing models assume roughly symmetrical risk — that gains and losses are distributed in a way that resembles a bell curve. Carry trade returns violate this assumption comprehensively.
Carry returns are positively skewed in frequency (you collect small daily income most of the time) but negatively skewed in magnitude (the occasional loss dwarfs the accumulated gains). This is the textbook definition of negative convexity, and it means that standard 1-2% risk-per-trade models systematically underestimate the actual risk of a carry position.
A practical framework starts with maximum position size as a percentage of account equity. For carry-specific positions, we would argue for a ceiling of 0.5% risk per position — half of what you might allocate to a standard directional trade. The reasoning is straightforward: you are likely to hold carry positions for longer than directional trades, which means more time exposed to the tail risk of a sudden unwind.
Scaling in also suits carry trades better than all-at-once entries. Open the position at quarter size when conditions look favourable, add another quarter if the trade moves in your direction and volatility remains low, and maintain the option of a full position only when both the carry and the technical picture are aligned. This approach means your average entry price improves if the trade works, and your maximum exposure is limited if it does not.
The overarching principle: size every carry position on the assumption that the worst historical drawdown will happen tomorrow.
Building a Carry Strategy That Respects the Cycle
Using the VIX as a Carry Trade Traffic Light
If you want a single indicator that captures carry trade risk better than anything else we have found, it is the VIX. Not because it predicts unwinds precisely, but because it tracks the one thing carry trades cannot survive: sustained volatility.
A simple traffic light framework works surprisingly well:
– Below 15 (green zone): carry conditions are favourable. Full position sizing, wide stops, let the swap accumulate
– 15 to 25 (amber zone): conditions are deteriorating. Reduce position size by half, tighten stops, avoid adding new carry exposure
– Above 25 (red zone): exit or hedge aggressively
Applying this framework retrospectively to 2008 would have triggered an amber warning in mid-2007 when the VIX started climbing through 15, and a red exit signal in early 2008 — well before the worst of the NZD/JPY collapse. In 2020, the VIX breached 25 in late February, giving carry traders nearly two weeks of warning before the worst of the March sell-off.
No filter is perfect, and the VIX has periods where it stays elevated without a carry unwind materialising. But the cost of false signals — exiting carry positions that would have been fine — is modest compared to the cost of being fully exposed during a genuine unwind.
Hedging the Carry: Options and Partial Closes
Exiting an entire carry position every time risk indicators flash amber is one approach. It is also the least capital-efficient one, because re-entering after a false alarm means paying the spread again and potentially missing days of carry income during the rebalancing.
Buying out-of-the-money put options on NZD provides catastrophe insurance. An NZD/USD put 300-400 pips below the current price costs relatively little when volatility is low (which is precisely when you should be buying it) and provides a hard floor on your carry trade losses. The premium reduces your net carry income, so the decision is whether the insurance cost makes the trade unattractive. When the option premium exceeds 30-40% of your expected carry income, the trade is telling you that the market prices the risk higher than the reward.
Partial closes offer a simpler alternative. Take a third of your carry position off the table when a key technical level is reached or when the VIX crosses into amber territory. This locks in some accumulated carry income while maintaining exposure to further gains. The psychological benefit is significant: traders who have taken partial profits handle drawdowns on the remaining position with far more discipline than those riding a full position into deteriorating conditions.
Correlated pair hedging — shorting AUD/JPY to hedge a long NZD/JPY, for example — reduces carry exposure without closing the position. The correlation is not perfect, which means the hedge leaks in both directions, but it dampens the worst of the unwind risk.
The RBNZ Cycle and NZD Carry Windows
The RBNZ rate cycle is the single largest determinant of whether NZD carry is worth pursuing, and reading it correctly gives you a structural edge over traders who simply react to the current rate differential.
Carry trades are most attractive when the RBNZ is at or near the peak of a tightening cycle. At that point, rates are high, the market has priced in the hikes, and the NZD is typically well-supported by inflows. The carry differential is at its widest, and the risk of further appreciation provides a tailwind. This was the situation in 2007 and again in late 2023 after the RBNZ pushed the OCR to 5.50%.
The danger zone is the transition from peak to cutting cycle. Once the RBNZ signals that cuts are coming, the carry differential begins to narrow, and the NZD typically weakens in anticipation. Traders who hold carry positions into an easing cycle are fighting two headwinds: declining swap income and adverse exchange rate movement.
The funding currency cycle matters equally. NZD/JPY carry was exceptional for decades partly because the BOJ showed no appetite for raising rates. The BOJ incremental tightening since 2024 has narrowed the gap, making the carry calculation less straightforward than it was during the zero-rate era.
Positioning ahead of the cycle — entering carry when the RBNZ is hiking and exiting when the language shifts dovish — captures the majority of the carry return while avoiding the worst of the cyclical drawdowns. The RBNZ Monetary Policy Statements provide the roadmap. Read them.
Carry trading the Kiwi dollar is not a passive strategy, regardless of what the daily swap credit might suggest. It demands active attention to the RBNZ cycle, honest arithmetic about break-even drawdowns, and the discipline to cut exposure when volatility shifts from background noise to signal.
The yield differential is the reason to look at the trade. Position sizing is the reason you survive it.