The Carry Trade, How Interest Rate Differences Move Currencies

[Fairman]

The Carry Trade, How Interest Rate Differences Move Currencies

A carry trade involves borrowing in a low-interest currency and investing in a higher-interest one, aiming to earn the interest rate difference. In forex, traders can capture this through positive swap, the daily interest credited for holding certain positions.

Classic example: borrowing in Japanese yen, historically low interest, to buy a higher-yielding currency such as the Australian or New Zealand dollar.

Why it matters: carry trades influence currency flows. When risk appetite is high, carry trades tend to grow, supporting high-yield currencies. When fear rises, carry trades may be unwound quickly, causing sharp moves.

Key points:

1. Swap rates. Brokers list swap for long and short positions. Some pairs pay positive swap in one direction.

2. Risk of reversal. Carry gains can be small compared with price swings. A sharp adverse move can erase months of interest.

3. Sensitivity to sentiment. Unwinding often happens suddenly during risk-off periods.

4. Central bank changes. Shifts in interest rate expectations can change the whole trade.

How to use this knowledge:

- Be aware of swap on positions you hold for days or weeks
- Watch for carry unwind signals, like rapid yen strength
- Don't hold a trade just for positive swap if the price trend is against you

Treat carry as a bonus, not a reason to ignore risk.

Interest is nice. Capital protection is nicer.

[LondonNewsTrader]

The unwind risk is worth a concrete example. In early August 2024 a combination of a Bank of Japan hike and weak US jobs data triggered a sharp unwind of yen-funded positions. USDJPY fell from around 162 in early July to about 142 in early August, a little over 12 percent in roughly a month, and much of it in a few sessions. Months of positive swap on long yen-funded positions disappeared in days.

The practical detail for retail traders is that the swap you see is not the full interest differential. Brokers keep a margin on both sides, so the positive swap on the long leg is usually smaller than the rate gap suggests, while the negative side is charged in full. And triple swap is charged once a week, usually on Wednesday night, to cover the weekend.

So the swap is a small, slow income and the price risk is large and fast, which is your point exactly.

[Fairman]

One caveat I should have stressed: carry looks great until risk sentiment flips. Positive swap pays you a little every day, but a risk-off week can wipe out months of that interest in a few sessions. Carry works best when volatility is low and stays low. A quick way to put it in perspective is to compare the daily swap with the pair's average daily range. If AUDJPY pays you the equivalent of half a pip a day and moves 80 pips on a normal day, one bad session is worth several months of interest. That doesn't make carry useless, it just means the swap is a small bonus on a trade you'd want to hold anyway.

[Fairman]

A practical check before holding for carry: look up the actual swap your broker pays. Some brokers' positive swaps are tiny after their markup, and sometimes both sides are negative. The theory only helps if your broker actually credits you. In MT4 or MT5, right click the symbol in Market Watch and open the specification to see the long and short swap. Remember the triple swap day too, usually Wednesday night, when three days are charged or credited to cover the weekend. Compare the yearly swap income on your position size with a normal daily move against you. Often one bad day wipes out months of carry.

[Fairman]

How I'd sanity-check a carry idea: look at the trend on the daily too. Carry with the trend behind it is the comfortable version. Carry against a strong trend means you're collecting pennies while price walks the other way. The yen carry unwind in August 2024 showed how fast it can go wrong, with USDJPY falling more than ten big figures in a few weeks and wiping out far more than a year of swap. I'd also watch volatility. Carry trades tend to work in calm markets and suffer when volatility jumps, so a rising VIX is a reason to cut size on carry positions.

[Shadow Trader]

The half a pip a day against an 80 pip range is worth stretching over a longer hold to see how slowly the bonus builds. Credited every night, with the Wednesday triple covering the weekend, half a pip a day comes to 3.5 pips a week.

Over 26 weeks that is 91 pips, so half a year of carry is worth a little more than one normal day of range. The made up chart below plots that line against a normal 80 pip day and a 160 pip risk off day.

img_12814_r1.png (70.09 KiB) Viewed 27 times

A 160 pip day against the position, only twice the normal range, takes back nearly 46 weeks of swap. That is close to a year of carry gone in a single session.

One way to compare pairs is to divide the daily swap by the average daily range. Half a pip over 80 pips is about 0.6 percent, and anything that small is decoration on a trade that has to work on price.

The ratio also moves when central banks shift, since the swap can shrink after a single meeting while the range widens on the news. Do you recheck the swap after each rate decision, or only when opening the trade?

[Shadow Trader]

Another way to size up a carry trade is to compare the yearly interest gap with the pair's volatility over the same period. It turns the swap versus range idea into one ratio that works on any holding period.

Say a pair pays a 3.5 percent rate gap a year and has annualised volatility of 10 percent. That is a carry to risk ratio of 0.35.

Monthly, the carry is about 0.29 percent, since 3.5 divided by 12 is 0.29. A one standard deviation month is 10 divided by the square root of 12, about 2.9 percent, so an ordinary bad month wipes out roughly ten months of interest.

That is before the broker markup, which as LondonNewsTrader pointed out comes off the credited side. If the markup takes 1 point of the 3.5, the ratio drops to 0.25 and the same bad month costs about 14 months of carry.

The ratio also falls when volatility jumps, which is exactly when unwinds tend to happen. Does anyone track carry to volatility across pairs, or just pick the highest swap?