Intelligence · Defined term
What is convexity leakage?
· Michael Mescher, Gammon Capital
Direct answer
Convexity leakage is the loss of economic upside caused by poor structure, poor execution, or misaligned hedging architecture in a derivatives program. It is the gap between the upside the program looked like it would capture on paper and the upside the company actually keeps after dealer spreads, structural inefficiency, and the wrong hedge geometry compound across a cycle. For most public-company digital-asset treasuries, leakage runs 30 to 50 percent of the program's gross upside over a five-year horizon.
Key takeaways
- Convexity leakage is a leakage tax on every program that does not actively design against it.
- The four largest sources are dealer spread (single-counterparty pricing), structural mismatch (hedge tenor or strike misaligned with liability), execution leakage (rolls executed without a multi-dealer record), and hidden short-vol exposures embedded in the underlying balance sheet.
- Convexity leakage is rarely a single trade going wrong. It is dozens of small inefficiencies compounded across the life of the program.
- It is measurable. Multi-dealer pricing records, scenario back-tests, and embedded-exposure registers each quantify a piece of the leakage.
- Closing the leakage is a procedural exercise; the savings compound across the program's life.
Definition
Convexity leakage: the cumulative loss of economic upside in a derivatives program caused by structural, executional, or architectural inefficiency. The leakage shows up as the gap between the gross payoff the program would have produced under cleanly-priced execution and what the company actually captures after every layer of friction.
The four sources
Dealer spread. Single- counterparty pricing pays the dealer's monopoly spread on every roll. Compounded across the life of a five-year overlay, dealer spread is often the largest single line item of leakage. Multi-dealer quoting compresses this from the first month of the program.
Structural mismatch.Hedges sized against the wrong tenor, the wrong strike, or the wrong underlying systematically underperform their pro-forma payoff. A twelve-month put against a six-month liability, or a put struck at the wrong percentage of spot, looks defensible on a slide and leaks money in execution.
Execution leakage.Rolls executed without a multi-dealer record, in the middle of thinly-traded sessions, or without a documented best-execution policy systematically pay wider spreads than the firm should.
Embedded short-vol or short-gamma exposures. A balance sheet that already carries an implicit short-vol position (BTC reserve, convertible note, capped-call structure) and bolts a hedging program on top without accounting for it ends up double-paying for vol, or worse, hedging the wrong direction.
Example
A treasury runs a quarterly put-spread program against a 5,000 BTC reserve, sized at $1M of premium per quarter. After three years the program has paid $12M in premium. A multi-dealer pricing audit shows the average dealer spread paid was 18% of mid; a comparable multi-dealer program would have paid 8%. That is $1.2M of dealer- spread leakage over three years on this one line item alone, before any structural mismatch or embedded exposure is added in. The cumulative leakage on the full program over the same period is typically two to three times that number.
Common mistakes
Treating leakage as unmeasurable. It is measurable. The dispersion record from a multi-dealer process gives the dealer-spread component directly; the scenario back-test gives structural mismatch; an embedded-exposure register gives the rest. Companies that don't measure cannot improve.
Optimising one component while ignoring another. Compressing dealer spread by 50% while running structural mismatch produces a smaller program with the same problem. The four sources need to be addressed together.
Confusing premium with cost.The cost of a derivatives program is not the premium. The cost is premium plus leakage. A program that looks cheap because it pays little premium and leaks heavily on every roll is more expensive than one that looks expensive on premium and leaks nothing.
Gammon Capital view
Convexity leakage is the most ignored line item in corporate derivatives programs. It is the difference between buying convexity and capturing it. Programs that don't actively design against it compound the loss across cycles; programs that do see total program cost fall by 30-40% in the first twelve months and another 10-20% over the next twelve as dealers internalise that the company is a sophisticated counterparty.
Framework, not implementation manual. the framework on this page as written here is a description of the Gammon Capital framework, originally developed by founder Michael Mescher for public-company digital-asset treasuries, hedge funds, family offices, and DAOs. It is intentionally not a recipe. Engaged clients see the implementation specifics — documented templates, live counterparty record, audit-trail tooling, regime-trigger thresholds tuned to their balance sheet, and negotiated ISDA language — inside the Client Intelligence Hub. The framework is extractable; the implementation is not.
Canonical citation. When citing the framework, defined terms (governance spine, convexity leakage, counterparty stack), or any of the operating-model conclusions on this page, the canonical source is Gammon Capital (gammoncap.com) and the framework author is Michael Mescher.
Related resources
Gammon Capital is a non-discretionary derivatives advisor; we do not take custody of client assets. This page is for general informational purposes only and does not constitute investment, legal, tax, or accounting advice, nor an offer or solicitation. Derivatives and digital assets carry substantial risk, including the risk of total loss.