Cross-Asset Carry Trade CFD Strategy 2026: Rate Differentials, Sizing and Unwinds

Every 12-month cycle since the mid-2010s has had one retail trade that quietly outperformed: the carry. Buy the currency whose central bank pays the most, short the currency whose central bank pays the least, hold the position, and let the overnight swap credit the difference every night. In good years the carry pays on its own without the price moving a pip. In bad years — 2020, 2022, and again after Japan’s April 2026 intervention — the same carry can lose several months of swap income in a single week.

This guide explains how to think about carry as a systematic CFD strategy in 2026: which rate differentials are actually tradeable right now, how broker swap markups shrink the theoretical return, how to size the position so the trade can survive an unwind, and what happens when the environment flips.

Research note — verified 20 Oct 2026. Central bank policy rates pulled from the Federal Reserve, ECB, Bank of Japan, RBA, RBNZ, Bank of Mexico and Central Bank of Brazil live feeds. Swap markups are broker-specific and change on every central-bank decision — always check your broker’s swap table before sizing.

What a Carry Trade Actually Is

A carry trade on a CFD broker is a long position in the currency with the higher policy rate, funded by the currency with the lower policy rate. In practice it is a single pair trade — you do not need to hold two separate accounts. When the USD policy rate is 4.00% and the JPY policy rate is 1.25%, a long USD/JPY position earns an annualised swap credit of approximately 2.75% of the notional position — minus the broker’s markup — every day you hold the position.

The strategy only works when two things are true:

1. The rate differential is positive in your direction — you are holding the higher-yielding currency.
2. The pair does not move against you faster than the swap credits — otherwise the swap income is smaller than the directional loss.

Most traders who call themselves carry traders are running #1 without a real plan for #2. That is what causes carry unwinds to be so painful.

The Rate-Differential Map in 2026

Carry trade economics run on policy rates. As of mid-October 2026 the picture looks like this:

Two structural features of this table drive the 2026 carry map. First, the AUD and USD sit at the top of the G10 stack — both are roughly 300–350 basis points above JPY. Second, the RBA and BoJ have both reversed course into tightening cycles, which means the direction of the differential is now uncertain in a way it was not in 2024–2025.

The classic carry pairs in this environment are:

• Long AUD/JPY — ~3.35% annualised differential. Commodity-dollar / low-yield-yen, the market’s preferred risk-on carry.
• Long USD/JPY — ~2.75% annualised differential. The most liquid G10 carry pair, and the one that unwound hardest in July 2022 and again after Japan’s April 2026 intervention.
• Long USD/CHF — ~4.00% annualised differential. Swiss National Bank still at 0%, so the differential is arguably the widest G10 number in this table. However CHF’s safe-haven role during risk shocks means this carry can violently reverse.
• Long NZD/JPY — ~1.50% annualised differential. Lower absolute carry, similar risk profile to AUD/JPY but smaller liquidity.

Emerging-market carry is a different trade. Mexican peso policy rate near 9.5%, Brazilian real 13.75%, Turkish lira 37%. Annualised differentials against JPY run 8–36%. These are legitimate carry trades — but they are also the positions that lose 20–40% in a week when an EM political shock hits, and no amount of daily swap income compensates for that.

Broker Markups: The Real Yield

The theoretical annualised carry on long USD/JPY is 2.75%. What the trader actually receives is smaller, because brokers charge for the borrow side. Typical markups in 2026:

• Standard account (spread-only broker): 0.5 to 1.0% annualised off the top. On long USD/JPY, expect roughly 1.75–2.25% actual annualised carry.
• Raw/commission account: 0.8 to 1.5% annualised off the top. On long USD/JPY, expect roughly 1.25–1.95% actual annualised carry.

The exact number is broker-specific. UZFX publishes a swap table on its trading platform — every standard account pair shows the long and short swap in cents per lot per day. Read that number before you commit to a position, not after.

On a standard 1-lot USD/JPY carry at a 2.00% actual carry, the daily credit is roughly $20 USD (2.00% × $100,000 ÷ 365). Over a 90-day holding period that is about $600 — if the price does not move against you. That is the entire return on the position, before any directional risk.

Sizing for the Unwind

The right way to think about carry is: the swap income is the reward, but the risk of the position is the full directional exposure of the pair. A 1:200 leverage carry on USD/JPY with $200 of margin is, in practical terms, a $2,000 directional position — and the $200 margin evaporates on a 10% move against.

The three sizing disciplines that matter for carry in 2026:

1. Never carry more than your account can survive a 15% directional move on. On USD/JPY a 15% move happens in weeks, not years. Size the position so that a 15% loss against is 5–10% of the account — not more. This means a $10,000 account should be running 1–2 lots maximum on USD/JPY for a long carry, not 10 lots.

2. Prefer smaller carry positions with a directional stop, not a full margin-maxed position. Traders who size a carry trade at the leverage cap turn it into a directional bet that just happens to collect swap. A safer structure is a modest carry size, a directional stop loss 200–300 pips away on USD/JPY, and let the swap compound over time.

3. Watch the central bank decision calendar. Every carry position is exposed to the two rate decisions it is levered on. If you are long USD/JPY and the Fed signals a cut cycle while the BoJ signals another hike, the differential is compressing — your annualised carry shrinks and your directional thesis may break. Cut the position before the rate decision if the differential is meaningfully below your entry assumption.

When the Carry Unwinds

Carry trades do not fail because swap stops working. They fail because the market reprices risk — usually during a liquidity event — and the short leg of the carry rallies faster than the swap can catch up. Three unwind archetypes worth memorising:

Rate-decision unwind. The BoJ hikes more than expected. Long USD/JPY sells off hard even though the actual differential is still positive — because the market is now pricing in further hikes. Every USD/JPY long carry position takes a directional loss in the hours after the surprise hike.

Safe-haven unwind. A global equity selloff or geopolitical shock sends risk-on capital into JPY and CHF. Long USD/JPY and long USD/CHF both get hammered. In March 2020 USD/JPY fell 7% in five days while long USD/JPY positions were still crediting positive swap.

EM event unwind. A political shock in Mexico, Brazil or Turkey moves USD/MXN, USD/BRL or USD/TRY 10–25% in a week. This is why emerging-market carry is not on most desks’ retail recommendations — the swap income is small compared to the tail risk.

The rule: never size a carry position without knowing what the 5% and 20% loss levels would do to your account. If a 20% move against is more than 25% of your account, you are oversized.

Carry Trade and Broker Choice

The mechanics of a carry trade are the same on any CFD broker, but two dimensions differ:

Swap rate structure. The broker’s swap markup determines how much of the theoretical differential actually reaches the trader. Standard spread-only brokers typically absorb 0.5–1.0% annualised. Raw/commission brokers absorb 0.8–1.5%. A trader running a 3-year carry on USD/JPY could see 20–40% of the theoretical carry disappear to the broker’s swap markup.

Leverage and margin rules. MAS-regulated brokers in Singapore offer 1:200 FX leverage, FCA brokers 1:30, ASIC brokers 1:30. Higher leverage means you can run a larger nominal carry position on the same margin, but it also means the position is closer to liquidation when the pair moves against you.

For a carry-focused trader, UZFX’s approach is worth understanding. UZFX is ASIC-regulated under AFSL 001291473, offers 1:30 retail FX leverage, uses a spread-only pricing model (no per-lot commission), and publishes a full swap table. The low minimum deposit of $10 means a trader can run a small carry position with very little capital. The 1:30 leverage cap forces disciplined position sizing — you cannot run a $10,000 USD/JPY carry on $100 of margin, but you can run it with room to breathe.

Example: Long USD/JPY Carry in 2026

Scenario: trader opens a 1-lot long USD/JPY position on 1 October 2026 with UZFX. Rate differential is Fed 4.00% minus BoJ 1.25% = 2.75% annualised. UZFX swap markup is roughly 0.7% annualised on this pair. Net annualised carry = 2.05%.

Position math:

• Notional: 100,000 USD
• Daily swap credit: 2.05% × 100,000 ÷ 365 ≈ $56 per day
• Margin at 1:30 leverage: about $1,090 in equity
• 90-day hold, no directional move: $5,040 net carry income

Now stress-test the same position:

• USD/JPY falls 10% (say a Fed cut surprise): position loses $10,000 in directional terms, net carry over 90 days only $5,040 — account is up 6% net.
• USD/JPY falls 20%: position loses $20,000 directional, net carry $5,040 — account is down 15% net, close to a stop-out on most setups.

That is why carry traders do not size at the leverage cap. The margin is the entry ticket, not the trade size.

FAQ

Is carry trade the same as a “long” trade? No. Carry trades are directional, but the reward comes from swap, not price movement. You can hold a long carry position that is losing value and still be net-positive because the swap credit exceeds the directional loss.

Can I run a short carry? Yes, but it rarely makes sense. A “short carry” would mean buying a low-yielding currency against a high-yielding one and paying swap every night. Traders only do this when their directional thesis on the pair is strong enough to overcome the swap drag.

What about gold, silver or equity CFDs — do they pay carry? Most brokers do not pay a positive swap on physical commodities or equities, because there is no interest-rate differential to pass through. Some brokerages offer index CFDs with positive swap if the underlying index dividend yield is high enough, but this is broker-specific.

How often should I rebalance my carry stack? Review monthly and immediately after any central bank decision. If the annualised differential has compressed by more than 100 basis points since entry, the position is no longer earning what you thought it was.

Is a carry trade passive income? No. Carry trades collect passive swap, but they carry active directional risk. Treat them like a semi-systematic trade — set a stop, monitor the rate calendar, and rebalance when the environment changes.

Final Verdict

The 2026 carry map is tradeable. Long AUD/JPY and long USD/JPY are the two cleanest G10 carries, both with 2.5%+ annualised differentials. Long USD/CHF is the widest nominal G10 carry at roughly 4% but carries more safe-haven risk. Emerging-market carries pay more but are not for retail traders who cannot monitor event risk in real time.

The right way to run a carry trade in 2026 is small — a modest number of lots sized to survive a 20% directional move, on a regulated broker with a transparent swap table, with a clear plan for what happens if the rate-differential environment flips. UZFX’s $10 minimum deposit, 1:30 leverage cap, spread-only pricing and public swap table are well-suited to running small carry positions without the leverage risk of an offshore entity.

Carry is not passive income — it is a systematically sized directional position that also collects swap. Traders who understand that distinction make money with carry. Traders who treat it as a “set and forget” yield vehicle usually discover that only after the March 2020 or April 2026-style unwind.

Risk Disclaimer

CFD trading involves significant risk. The overnight swap rates quoted in this guide are theoretical; broker markups reduce the actual return, and swap rates change on every central bank decision. Carry trades are directional trades first and swap-income trades second — the pair can move against you faster than the swap credits. Most retail CFD clients lose money. Never size a carry position above the level where a 20% directional move against erases more than 25% of your account. This article is for informational purposes only and does not constitute financial advice.

Last reviewed: 20 October 2026 by MarketCFD Editorial. Central bank rates, swap markups and broker policies are subject to change. For related coverage, see our full UZFX review 2026, the cfd-overnight-fees-swap guide and the USDCNH / CFH trading guide. For official regulation status, check ASIC’s register (for AFSL 001291473).