Convexity of Carry

What Is Convexity of Carry?

Convexity of Carry refers to the second-order sensitivity of a trade's carry return to movements in the underlying variable — whether that is an interest rate, credit spread, currency forward premium, or commodity basis. While conventional carry analysis focuses on the linear income earned by holding a position, convexity of carry captures the degree to which that income stream itself changes as market conditions evolve. A position with positive carry convexity benefits from large moves in either direction while still collecting carry in stable environments; negative carry convexity means carry income erodes precisely when volatility rises and the position most needs protection. The concept sits at the intersection of carry trade mechanics and duration/convexity analysis familiar from fixed income markets, but applies broadly across asset classes including FX, credit, and rates.

Why It Matters for Traders

Many systematic carry strategies appear attractive on a Sharpe ratio basis in calm regimes but embed deeply negative convexity that only surfaces during stress episodes. A classic example is short volatility risk premium strategies layered with carry: the premium collected is the carry, but the payoff profile is concave — losses during vol spikes far exceed the accumulated carry income. Recognizing convexity of carry allows portfolio managers to compare two seemingly similar carry trades and identify which one has a more favorable curvature profile. In rates markets, a yield curve steepener funded at the short end may have positive carry convexity if the curve steepens nonlinearly during risk-off episodes. In FX, high-yielding EM currency carry often embeds sharply negative convexity due to sudden stop dynamics.

How to Read and Interpret It

To measure convexity of carry, analysts compute the carry return across a range of scenarios for the underlying variable — typically spanning ±1, ±2, and ±3 standard deviation moves — and examine whether the carry income is symmetric, accelerating, or decelerating. A simple heuristic: if carry at +2σ and −2σ scenarios is higher than at the base case, the trade has positive carry convexity; if carry collapses in tail scenarios, the trade has negative carry convexity. Quantitatively, traders often estimate a "carry gamma" metric — the second derivative of P&L with respect to the underlying variable, holding position constant. Positions with carry gamma below −0.05% per 10bp move in the underlying should be flagged for stress testing against historical tail episodes.

Historical Context

The 2018 EM currency crisis provides a textbook example of negative carry convexity in FX carry trades. During 2017 and early 2018, EM carry baskets — long Turkish lira, South African rand, and Argentine peso against USD — generated annualized carry returns of 6–10%. However, as the DXY rallied roughly 8% between April and August 2018 and the Fed continued hiking, carry income evaporated and capital losses accelerated nonlinearly. The Turkish lira lost approximately 40% against the dollar from January to September 2018, meaning the cumulative carry of ~8% was wiped out many times over — a direct manifestation of deeply negative carry convexity that linear carry metrics had failed to flag.

Limitations and Caveats

Convexity of carry is scenario-dependent and sensitive to the assumed distribution of the underlying variable. Fat-tailed or bimodal distributions — common in credit and EM FX — can make convexity estimates highly unreliable if calibrated on short historical windows. Additionally, convexity of carry is not static: it shifts as the position moves in- or out-of-the-money and as the broader vol regime changes. Strategies that appear convex in a low-volatility regime may revert to concave payoffs when implied volatility surfaces reprice. Finally, carry convexity analysis typically ignores funding costs, which can themselves be convex in stress periods.

What to Watch

• Spread between realized and implied volatility across carry-heavy asset classes (FX, credit, commodities) as a leading indicator of carry convexity breakdown
• Positioning crowding in EM carry trades via COT report and prime brokerage flow data
• Shape of the volatility term structure for clues about whether optionality embedded in carry trades is being repriced
• Fed rate path repricing events, which historically trigger the most severe negative carry convexity episodes in duration and FX carry