Digital options are contracts whose payment depends on whether a stated condition is met. The simplest version, a cash-or-nothing option, pays a fixed sum if the condition is satisfied and nothing otherwise. The size of the market move beyond the threshold does not increase that payment.
Their origins require a distinction between three developments: the mathematics used to value contingent payments, the classification of digital contracts, and their later distribution through exchanges and online platforms. These are related histories, not interchangeable invention dates. This article examines the first two; the broader history of binary options follows the product’s subsequent commercial development.
The Pricing Foundations: Why 1973 Matters
The theoretical foundation came from modern option pricing. In 1973, Fischer Black and Myron Scholes published their option valuation model, while Robert Merton developed related reasoning and extensions. The central advance was a method for connecting an option’s price to the cost of managing its exposure through trading, rather than relying solely on a forecast of the underlying asset.
The framework reached beyond ordinary calls and puts. Merton’s Nobel lecture on option pricing describes how his 1973 work extended the model to dividends, changing exercise prices, early exercise and exotic provisions, including options that disappear when a price barrier is crossed.
That did not establish a single birthday for cash-or-nothing contracts. It established a method for analyzing them. Once a contract’s future payment could be stated mathematically, the next question was whether that payment could be valued and hedged under the model’s assumptions.
Consider two promises tied to a share price. One pays the amount by which the share exceeds $100. The other pays exactly $10 if the share exceeds $100. Their payment patterns differ, but both depend on the same uncertain future price. A sufficiently general pricing framework can address both. The second promise does not require an entirely separate theory simply because its payoff has a sharp edge.
A Documented Milestone: Rubinstein and Reiner in 1991
A clear historical marker is Mark Rubinstein and Eric Reiner’s paper dated July 31, 1991, titled Binary Options. A version appeared in the October 1991 issue of Risk as Unscrambling the Binary Code. The paper classified and valued contracts with different conditional payments, separating those dependent on the final asset price from those dependent on the price path.
The Berkeley archive of Rubinstein and Reiner’s binary options paper also references a March 1991 article that called these contracts digital options. That supports a narrower, more defensible historical claim than naming one inventor: the terminology and product family were already being discussed when Rubinstein and Reiner organized their analysis.
Their contribution was a systematic treatment within a Black-Scholes setting. It was not a claim that every possible conditional payment had just been invented. A publication date documents an explanation; it does not necessarily identify the first privately negotiated trade.
This distinction matters because histories of financial products often compress gradual development into a convenient launch date. For digital options, the evidence supports a sequence of pricing foundations, documented classification and later market access.
What Cash-or-Nothing Actually Means
A cash-or-nothing call pays an agreed amount when the final reference price is above the strike. A corresponding put pays when it is below the strike. The contract must also state what happens if the reference price exactly equals the strike.
The related asset-or-nothing option pays the asset, or its value, when the condition is met. Its successful payment therefore varies with the asset price. This distinction appears in the payoff definitions used in research on digital and binary barrier options.
| Contract | Illustrative condition | Payment if satisfied |
|---|---|---|
| Cash-or-nothing call | Final price above $100 | Fixed $10 |
| Cash-or-nothing put | Final price below $100 | Fixed $10 |
| Asset-or-nothing call | Final price above $100 | One unit of the asset, or its value |
Suppose the cash-or-nothing call in the table costs $4. If the share finishes at $100.01, its gross payment is $10 and the buyer’s profit is $6 before fees and financing. If the share finishes at $150, the figures are unchanged. A much better directional forecast earns no extra payment.
If the share finishes below $100, the payment is zero and the buyer loses the $4 premium. “Nothing” describes what the contract pays, not what it cost. Confusing payment with profit makes a simple contract look more generous than it is.
Why Digital Payments Belong to Ordinary Option Mathematics
The connection between digital and conventional options becomes clearer by constructing their payments directly.
Take an asset-or-nothing call that delivers one share if its final price exceeds $100. Pair it with a short cash-or-nothing call requiring a $100 payment under exactly the same condition. Above the strike, the combined position receives a share and pays $100. At or below the strike, both components pay nothing.
The resulting payoff is the same as an ordinary call with a $100 strike: the amount by which the share price exceeds $100, or zero. This is a payoff identity, not a trading recommendation. It demonstrates why digital payments can be treated as components of familiar options rather than an unrelated financial invention.
A Narrow Call Spread Gives Another Illustration
Consider buying a call struck at $100 and selling a call struck at $101, both expiring together and covering one share. Ignore premiums and fees for the moment.
At a final share price of $99, neither call pays anything. At $100.40, the purchased call pays $0.40 and the sold call pays nothing. At $102, the purchased call pays $2 while the sold call costs $1, leaving a net payment of $1.
The combined payoff rises from zero to $1 over a narrow price interval, then stays at $1. Narrow that interval further and scale the position to preserve the payment, and the payoff increasingly resembles a cash-or-nothing contract away from the threshold.
The resemblance is not exact for a spread with a finite width. Inside its strike interval, it pays a sliding amount rather than switching immediately between zero and the full payment. That small interval is where much of the practical difficulty sits.
From Conditional Payment to Price
The pricing idea can be expressed without a page of equations: value the promised payment using the appropriate pricing probability, then discount it to the present.
The word “appropriate” does considerable work. Derivatives pricing uses risk-neutral probabilities, which account for how markets price risk. These need not equal an investor’s forecast of how frequently an event will actually occur. The Bank of England’s handbook on option-implied probabilities distinguishes these pricing probabilities from subjective probabilities and explains their role in discounted payoff valuation.
For a simplified example, assume a contract pays $100 if a condition is met. Suppose its risk-neutral probability is 40%, interest is zero, and there are no trading costs or counterparty losses. Its theoretical value is $40.
Now suppose a buyer pays $45 and independently estimates a genuine 40% chance of success. On those assumptions, the expected gross payment is $40 and the expected net result is a $5 loss. Correctly describing the contract’s two outcomes says nothing about whether the purchase price is attractive.
A Useful Consistency Check
Combine two contracts that divide every possible settlement outcome between them. One pays $100 when the reference price is at or above the strike; the other pays $100 when it is below. Exactly one pays, so the combined settlement is always $100.
Under ideal pricing assumptions, their combined present value must equal the discounted value of that $100 payment. Desmond Higham’s cash-or-nothing option exercises demonstrate this relationship using a different equality convention that produces the same combined payment.
The reasoning also shows why settlement wording matters. If two contracts both exclude equality, their combined payment is not guaranteed in that outcome. A tidy mathematical identity depends on matching contractual terms, not just matching product names.
Simple Payments, Difficult Exposure
A fixed payment can look easier to manage than an ordinary option’s variable payment. Near the threshold, the opposite problem appears: a tiny change in the reference price can decide the entire amount.
Return to a contract paying $10 if a share finishes above $100. A settlement price of $99.99 produces nothing; $100.01 produces $10. The share price differs by two cents, but the contract’s payment differs by its full amount.
This discontinuity also creates market conduct concerns. The 2014 Fair and Effective Markets Review discussion of barrier and digital options examined the incentives created when small underlying price movements determine large derivative payments. It distinguished ordinary hedging from attempts to influence the market around a trigger.
For our example, the practical questions are straightforward. Which price determines settlement? Is it a last trade, an auction value or another reference? At what exact time is it observed? How is rounding handled? A screen showing $100 is insufficient if the contract uses a differently calculated settlement value.
These are not decorative clauses. They define the event being purchased. Two contracts can share a strike and expiry date yet produce different results because their reference prices or observation rules differ.
Why Exchange Listings Were Not the Beginning
A later milestone came on May 22, 2008, when the SEC approved a CBOE proposal to list and trade binary options on broad-based security indexes. The SEC’s approval order for CBOE binary index options described fixed settlement payments, European-style exercise and clearing through the Options Clearing Corporation.
That was a development in how these contracts could be issued, traded and settled. It was not the invention of their payoff structure, which had already received published analytical treatment years earlier.
The distinction is similar to the difference between designing a contract and building a market around it. Standardized terms and clearing arrangements address trading organization. They do not create the underlying mathematical idea.
The separate history of exchange-traded binary options covers those market arrangements. For the origins of digital options, the relevant point is that exchange approval belongs later in the chronology than the pricing theory and documented classification.
The Contract Came Before the Online Interface
Nothing in a cash-or-nothing payoff requires an internet platform, a countdown clock or an expiry measured in seconds. Change the observation date from tomorrow to six months away and the same conditional payment can still be written. Change the trading interface and the payoff need not change at all.
That separation helps explain why institutional digital options and retail binary products should not be treated as identical commercial arrangements. A shared payment shape does not establish identical pricing, counterparty exposure, settlement procedures or customer protections. Those questions require examination of the actual contract and venue.
The rise of online OTC binary options platforms is therefore a distribution story, distinct from the earlier development of digital option mathematics.
The most defensible account of the origins is not a single invention date. Modern option pricing supplied a framework; published work such as Rubinstein and Reiner’s 1991 paper organized the contract family; later trading arrangements brought particular versions to different audiences. The enduring idea is straightforward: isolate a condition and attach a payment to it. Valuing that promise, and settling it fairly, requires considerably more care.