Carry positions often look stable for long stretches because gains accrue slowly through yield pickup, spread income, or roll mechanics. That surface stability can hide a more fragile structure underneath. A carry trade depends not just on collecting income, but on calm price behavior, reliable financing, manageable leverage, and the ability to keep the position open without disruptive changes in funding or volatility.
That is why carry unwinds tend to be asymmetric. The income side builds gradually, but the loss side can arrive quickly through repricing. A position may earn carry for weeks or months, yet a short period of wider spreads, higher volatility, tighter financing, or forced deleveraging can overwhelm that accumulated income. The weakness is not that carry never works. It is that the mechanism of earning and the mechanism of losing operate on very different clocks.
Why carry becomes vulnerable
Carry is most vulnerable when the assumptions that make the position holdable begin to deteriorate. The trade may still look sensible in theory, but the practical conditions supporting it become less stable. Financing may become more expensive, rollover may become less certain, volatility may rise enough to enlarge mark-to-market swings, or liquidity may weaken enough that exiting becomes harder and more costly.
Leverage makes that vulnerability sharper. Modest adverse moves that would be tolerable in an unlevered position can become much harder to absorb when collateral usage, funding terms, and risk limits are tight. In structures such as duration carry, that problem is especially visible because the income stream can look steady while the valuation side remains exposed to changes in yield levels, curve behavior, and financing conditions.
Maturity mismatch adds another layer of fragility. Many carry structures depend on funding or hedging that resets faster than the investment thesis itself. That means time is not neutral. A position can appear sound if held to horizon, yet still become vulnerable because the liabilities, collateral terms, or refinancing assumptions supporting it are repriced before the carry thesis has time to play out.
Main triggers of a carry unwind
One common trigger is deterioration in financing conditions. If borrowing costs rise, lenders become more selective, or balance-sheet capacity becomes scarcer, the income advantage behind the trade narrows. The position does not need to fail dramatically on the asset side for this to matter. The liability side alone can erode the economics enough to make continuation unattractive or impossible.
Another trigger is spread compression. Carry depends on a return margin that compensates for risk, leverage, and liquidity usage. When that margin shrinks, the position may still be profitable in a narrow sense, but no longer profitable enough relative to the capital, volatility tolerance, and financing dependence required to keep it on. In that case the rationale weakens before the trade fully breaks.
Volatility repricing is another important catalyst. Carry tends to perform best when interim price movement remains contained. Once volatility rises, the same structure starts to consume more margin, more risk budget, and more institutional tolerance. The problem is not only larger price swings. It is that the cost of surviving those swings rises at the same time that the return buffer from carry often becomes less compelling.
Crowding can turn this deterioration into a faster unwind. When many investors are relying on similar funding conditions, similar spread relationships, or similar assumptions about stability, exits become synchronized. What begins as individual reassessment turns into collective position reduction. Liquidity thins, price moves become larger, and those moves create further pressure on remaining holders.
How the unwind becomes self-reinforcing
Carry unwinds often strengthen themselves through feedback loops. Initial losses reduce collateral value or increase risk usage. That leads to tighter financing terms, reduced balance-sheet tolerance, or internal pressure to cut exposure. Those reductions push prices further, which creates fresh losses and renews the pressure to exit. At that stage, the trade is no longer being judged only on expected carry. It is being judged on whether it can still be financed, margined, and held at all.
The process can also spread across markets because the common link is often balance-sheet dependence rather than the asset itself. Positions in rates, credit, foreign exchange, or basis trades may all come under pressure if they share the same funding ecosystem or rely on the same dealer intermediation capacity. Correlation rises not because the assets are suddenly identical, but because the conditions supporting their ownership weaken at the same time.
That is why a carry unwind can start as a structural problem before it becomes a visible market event. In its early phase, the trade may simply feel less comfortable to hold. Financing is a bit tighter, volatility a bit higher, liquidity a bit thinner, and the earned spread a bit less attractive. The visible break usually comes later, when those smaller changes stop being absorbable and begin to interact.
Carry unwind vs normal carry drawdown
Not every loss in a carry position is an unwind. A normal drawdown can happen even when financing remains available, leverage remains serviceable, and the broader spread logic is still intact. In that case the trade is under pressure, but it is still being carried as intended.
An unwind is different because the issue is no longer just poor performance. The conditions required to continue holding the position begin to erode. The holder is pushed toward reducing exposure not simply because the position is down, but because volatility, liquidity, funding, or crowding have made the structure itself less stable. The defining feature is therefore not loss size alone, but the loss of holdability.
That distinction matters because carry fragility is usually structural before it becomes dramatic. By the time price action makes the unwind obvious, the deeper problem is often that the assumptions behind stable carry exposure have already weakened for some time.
FAQ
Why can carry trades earn steadily for a long time and then reverse so fast?
Because the income accumulates slowly, while losses arrive through repricing. Carry is collected over time, but mark-to-market adjustments happen immediately when volatility rises, spreads move, or financing conditions tighten.
Does every rise in volatility cause a carry unwind?
No. A volatility increase becomes part of an unwind when it materially changes the ability to finance, margin, or tolerate the position. Temporary turbulence can create drawdowns without forcing an exit if the trade remains structurally holdable.
Is a carry unwind mainly about the asset side or the funding side?
It can begin with either, but funding often plays the decisive role. A trade may still make sense on paper while becoming unattractive or unsustainable because borrowing costs rise, leverage becomes harder to maintain, or rollover conditions worsen.
Why do carry unwinds often spread beyond one market?
Because many carry structures share the same balance-sheet constraints. If funding conditions tighten or dealer capacity shrinks, pressure can spill across rates, credit, foreign exchange, and other relative-value positions that depend on similar financing assumptions.