TechFlowPost has published a guest essay arguing that a hedge is not something a company buys off the shelf. In Prathik Desai’s framing, a hedge is a relationship built around an existing exposure. Start with the contract instead of the exposure, and what the buyer may end up holding is a position that flatters the product while leaving the real problem in place.
The essay was translated by Block unicorn. Desai opens with an insurance example: ask most people what insurance they bought, and many will mention handset coverage rather than health insurance or income protection for dependents. Phone insurance is easy to add at checkout for a few hundred rupees and usually requires almost no deliberate decision.
That, he argues, says something about how risk decisions are actually made. People do not usually sit down, map every possible loss, rank them by severity and decide that a cracked phone screen belongs at the top. Coverage is often defaulted into the purchase flow, and the product is designed around what is easy to sell rather than what is economically central. Even so, those products can still work: claims get approved, screens get replaced and the insurance does what it promised.
From there, Desai moves to his main point. Lauris, he writes, studies how parts of the world become financial markets through event contracts, derivatives, corporate risk and the legal and market structures that connect them. In this essay, Lauris’ argument is that hedging is not a random tool purchase. It should be a carefully built relationship organized around a risk exposure that has already been identified. Reverse that sequence and the company may hold something that helps a trade idea while the underlying business exposure remains untouched.
Trader hedges and corporate hedges use the same word for different things
Anyone can trade a hedge, Desai writes. That does not mean anyone can sell a “hedging product” to a company.
For a trader, any position that reduces risk somewhere else on the book can fairly be called a hedge. That remains true whether the trader sits on a macro desk, trades crypto or uses a retail brokerage account. Buy a portfolio hedge tied to an election result and, even if the fit is loose and the protection incomplete, it is still a hedge in that trading sense.
Corporate finance uses the same word in a more demanding way. The exposure comes first: cash-flow risk, liability risk or operating risk. The instrument then has to match the amount, tenor and risk factor. Any slippage becomes basis risk. Credit terms, collateral, documentation and accounting all wrap around the trade because the product is not just a payoff profile. It is a relationship.
Swaps, forwards, options and event contracts can hedge one account and express a view in another. Their economic role depends on what other exposures sit on the holder’s balance sheet. For a company, the right questions are specific: what exposure does this offset, by how much, and for how long?
Both meanings of “hedge” are valid. The category mistake, in Desai’s view, is treating them as interchangeable.
Why quant and market-making experience does not automatically transfer
Desai says this confusion has appeared repeatedly in conversations he has had. People assume experience in quant trading or market making can be carried straight into structured transactions. The mistake shows up often in tech, where a quant background signals exceptional skill. In many cases, he notes, that reputation is deserved. Traders have built excellent exchanges, and many of them are strong operators.
Still, technical strength does not make domain knowledge universal. Quant trading and market making are organized around price, information and inventory. Building a product or swap that can function as a hedge is a service activity built around the client’s exposure. Those are not the same job.
Prediction market operators often make the same mistake, he argues. Their native unit is the contract: list an event, attract liquidity, then find flow. In corporate risk transfer, that sequence does not work because the exposure is the starting point. Yet the contract-first process has spread out of trading and crypto: take an existing yes-or-no contract, attach it to a company’s problem, route the order to an exchange and call the resulting position a hedge.
IFRS 9 and the CFTC both start from the risk itself
Desai points to International Financial Reporting Standard 9 as the operational version of this distinction. IFRS 9 requires identification of the hedged item, the hedging instrument, the hedged risk, the risk-management objective and the economic relationship among them. A company cannot simply point to a contract and declare itself hedged. It has to explain how that instrument links to the underlying exposure.
He says the U.S. Commodity Futures Trading Commission follows the same basic logic when testing swaps used to hedge physical positions. The risk must arise from an asset, liability, service or physical commercial activity.
What event contracts add
The financial case for prediction markets begins with state-contingent claims. Desai references the framework associated with Arrow and Debreu: the more future states a market can name, price and make transferable, the more complete that market becomes. If a tariff passes, a merger closes, a drug gets approved or a carbon auction clears above a given level, an event contract can pay out against that state.
Traditional markets may be able to price some of those outcomes only through proxies. Event markets can create observable market prices for states that previously lived mainly in research notes, scenario models or bilateral conversations.
He then links some event claims back to conventional derivatives. For threshold claims written on the same underlying, the connection can become exact in the limit: the price of a digital payoff is the negative slope of the call-price curve with respect to strike, and it can be approximated by an increasingly tight vertical spread. An event contract isolates the terminal state. Before expiry, an options position may bring in volatility exposure, mark-to-market risk, dealer intermediation and replication frictions that the buyer never wanted.
That equivalence has boundaries. “The S&P 500 closes above 7000” can map onto the S&P options surface. “The Federal Reserve cuts rates in March” or “a tariff bill passes” can still be valid predictions even if they are not strike derivatives of a call surface. Their market prices also differ from real-world probabilities because they reflect risk appetite, collateral, liquidity, route-to-market and settlement rules.
Government outcomes can be inputs into corporate cash-flow problems. Prediction markets can manufacture missing claims. They cannot manufacture the relationship between those claims and a company’s balance sheet.
When event contracts matter financially
Desai reduces the question to two tests:
- Whether a traditional derivative or a credible replication path already exists.
- Whether a company or investor is actually exposed to the risk in question.
When a material risk is already covered by derivatives, an event contract becomes a substitute. To win, it has to deliver a better match or a lower all-in cost. The structural cost inside the traditional channel has to exceed the spread, depth, collateral and market-impact costs of the event market. That is the replication cost difference.
The larger long-term opening, he argues, lies in major risks that still have no derivative market. Event contracts may be the first tool to make shutdowns, policy decisions, regulatory milestones, weather conditions and corporate events both observable and transferable.
If a derivative exists but no one who bears the exposure actually uses it, there is little genuine risk-transfer demand. Another binary option can still be a useful trading product. If neither of those conditions holds, the market is better suited to forecasting, entertainment or broad price discovery than to corporate hedging infrastructure.
Tradability tells you whether something can be priced. Importance tells you whether anyone bears a risk worth transferring. A sales funnel that starts from a list of contracts and then goes looking for companies to attach them to skips that second test entirely.
The wrong side of the desk
Trading logic starts with a profit target and looks for opportunities. Structured logic starts with the balance sheet and builds the trade.
In fixed income, currencies and commodities, corporate clients show up with existing exposures: floating-rate debt, currency mismatches, fuel expense, inventories, planned bond issuance or acquisition financing. The structured process identifies the risk factor, chooses the instrument, sets the amount and tenor, then calculates the residual basis and embeds the outcome in documentation, credit arrangements and accounting policy.
Desai notes that the International Swaps and Derivatives Association organizes the derivatives market in exactly that order: user, underlying risk, instrument. HSBC’s Autohedge system does the same by taking exposure, hedging policy and risk preference as inputs before calculating the trade. The instrument may stand alone legally. Economically, a hedge refers to the relationship built around it.
In FICC trading, by contrast, a dealer can answer a request for quote with a firm price for principal risk on a defined instrument. That is execution. It does not require any view about the customer’s exposure or whether the instrument is suitable as a hedge. Add an RFQ to a mismatched contract, and all that happens is that the mismatch gets a price.
What goes wrong when the contract comes first
Desai uses an importer worried about tariffs as an example. The importer’s loss depends on shipment volumes, timing, inventory, the ability to pass through costs, currency moves and the capacity to switch suppliers. An event contract might simply say it pays $1 if a publicly announced tariff above a given threshold is imposed before a stated date. The structure is simple. The fit to the importer’s cash flows is weak.
That gap is basis risk. A structured trader decides which parts of the exposure can be transferred and which parts stay with the client. A contract-first intermediary does the reverse. It finds a listed binary that vaguely resembles the client’s problem and treats resemblance as if it were hedging. The treasurer ends up owning the binary while most of the original risk remains in place.
A second issue appears in customization. Desai cites work by Robert Bartlett and Maureen O’Hara covering 41.60 million Kalshi contract trades. In their framework, there are no liquidity traders in the Glosten-Milgrom sense because the instrument does not routinely function in hedging and portfolio rebalancing the way products in mature markets do. Individuals can still use the contract defensively. That is a different thing.
Credit markets show how institutions actually buy state-contingent risk. In significant risk transfer transactions, banks keep the loans and buy first-loss protection through credit-linked notes funded by investors. BIS data show the protected loan pool stood at about €800 billion at the end of 2024. Desai is not recommending that companies buy credit-linked notes. His point is that institutions place risk inside funding instruments that come with coupons, documentation, loss allocation and approved limits.
The same paper finds that informed pricing effects are stronger in single-name markets than in macro markets. Tariff decisions are usually exogenous to an ordinary importer, so the firm’s order carries little information by itself. Market makers may welcome the flow, but the published tariff schedule only weakly tracks the importer’s actual loss.
If contracts were tied more tightly to mergers, drug trials, factories or other company-specific outcomes, basis would improve. Then the information problem gets worse. The company may know more than the quoting party, so market makers widen, size down or decline the trade. The most popular trades are often the ones with the largest basis. The trades that best fit the corporate problem are the ones the market is least willing to underwrite.
Even after a contract is done, the company still has to explain the objective, hedge ratio, rationale, valuation method and financial-statement treatment. A dashboard cannot create that link. An RFQ cannot force an audit committee to accept it.
That leaves the business model squeezed from both sides. If the exposure is small and already matched by a listed contract, the client can trade directly. If the exposure is large or needs customization, the client needs structuring and capital support. A firm that only points users to listed contracts does not add balance-sheet capacity. A firm that designs payoffs, prepares documents and commits or raises capital looks less like a new category and more like a broker, insurer or FICC structuring shop.
Packaging changed where demand showed up
WeatherBill launched a self-serve weather derivatives platform in 2007 on the theory that weather-sensitive businesses would buy protection. They did not, at least not at the scale imagined. The company narrowed its target to farmers, shifted toward insurance products and outside distribution, and later became The Climate Corporation.
Event contracts found native demand elsewhere. CME’s FanDuel platform launched in December 2025 and reported 100 million contracts in roughly eight weeks.
Weather products moved into insurance. Sports products found a retail distribution channel. Standalone corporate hedging sold through software still looks like an empty category.
Three ways event risk can reach the balance sheet
Once the exposure is identified, the next question is who can bear it. Event risk does not appear on the balance sheet by itself. It has to be housed somewhere else.
1. Direct trading, when the fit is good enough
Desai points to a Manhattan bar that bought about $5,000 of event contracts to offset a promotion offering free drinks if the Knicks won. Liability and contract were resolved in the same game. With the right settlement design, the same route can work for larger risks.
He also references a reported transaction tied to California solar tax credits that transferred about $600,000 through a public, centrally cleared market. Firm size is not the dividing line, he says. Fit and balance-sheet capacity are.
Desai adds that he is working with some people at Kalshi, especially 0x_ultra, to provide a showcase for the project as part of a “Builders Program,” and says it will be released soon.
2. Build a pool only when the risk is genuinely diversifiable
Parimutuel-style markets allocate a committed pool of funds among winners and cap total liabilities at the capital already there. Goldman Sachs and Deutsche Bank used that mechanism for economic derivatives beginning in 2002, with nonfarm payroll auctions averaging about $9 million before trading moved to CME in 2005.
BIS still questions whether real hedging demand can be balanced against experienced informed traders. For a pool to work, losses have to vary across participants, exposures have to offset, or outside capital has to come in. If everyone gets hit at the same time, the pool just creates a long list of claimants and still needs someone with capital on the other side, through insurance, reinsurance or warehouse financing.
Desai says very smart people, including AadvikVashist, are working on that problem.
3. Put the event inside a tool institutions already know
In April this year, Marex issued a structured note of up to $10 million to a Swiss institutional client that pays 7% interest if Nvidia is still the world’s most valuable company one year later. The client owns Marex debt. Marex uses event contracts to replicate the exposure.
The client bought a security with an issuer, documents and mandate language attached. The event contract stayed with the dealer. Before the trade reached the client, Marex had already converted the event claim into FICC form.
That, for Desai, is where software becomes useful: after the exposure has been identified. Software can test traditional instruments, identify protection gaps and compare event claims on basis, cost, execution, collateral, legal terms, mandates, accounting and residual risk. A practical product would encode that process and allow the final answer to be no trade at all.
Why bad corporate hedging would widen the regulatory blast radius
A bad corporate hedging idea usually harms the buyer. Here the damage could spread further because event markets are still arguing over their place in the financial system. Commercial use does not by itself determine CFTC jurisdiction over event contracts.
Desai says the Crypto Council for Innovation has argued that the Commodity Exchange Act’s swap definition, federal preemption and the CFTC’s exclusive authority over derivatives markets mean an event contract does not leave federal oversight simply because the buyer is speculating.
The public-interest case is broader than corporate hedging. A regulated event market can set prices for states that were previously unpriced, generate public prices and create transparent, collateralized claims with clear settlement rules. Those are meaningful financial functions even before treasurers start using the market. Still, hedging remains central to the policy argument.
The CFTC has described prediction markets as tools for forecasting, planning, hedging and speculation. In February 2026, the commission defended its jurisdiction against state gambling regulators by stressing commercial hedging, portfolio management and the information such markets produce about future outcomes. The Crypto Council for Innovation distinguished prediction markets from gambling by pointing to many-to-many execution, transparent pricing, clear settlement benchmarks, regulation, customer protections and market integrity.
In March, the CFTC reminded prediction market exchanges of their obligations under the Commodity Exchange Act and Core Principle 3. Its June proposal addressed event-contract design, public-interest review and responsible innovation.
Desai says serious corporate hedging failures usually follow a simple pattern. A company is told that a binary position offsets an operating risk and commits cash. The company then suffers a loss, but the contract can still expire worthless because shipment volume, timing or cost pass-through never entered the payoff. Before settlement, the public event underlying the trade may already have been recognized in earnings while the original exposure remains. What the company ends up with is not protection but a shortfall.
Regulators have seen related conduct before. The CFTC and the U.S. Securities and Exchange Commission issued a joint warning after complaints involving binary options platforms that refused withdrawals, engaged in identity theft or manipulated software to generate losses. Desai stresses that the venues discussed in his essay are not those fraudulent brokers; they are regulated, transparent and centrally cleared. Even so, that history raises the cost of mislabeling a trade as protection.
Opponents of federally supervised prediction markets, he concludes, will not write seminar papers about basis risk. They will say a company was sold a bet under the name of a derivative. If intermediaries present mismatched binaries as corporate protection, they will hand state gambling regulators the strongest argument against federal oversight of prediction markets.