What it means
Probabilities need to obey consistency rules, because they cannot be negative, a certain event has probability one, and mutually exclusive events have additive probabilities. A set of prices derived from incompatible probabilities can create an exploitable gap.
The Stanford Encyclopedia of Philosophy explains the theorem in terms of betting quotients, noting that if those quotients fail to satisfy the relevant probability axioms, a set of bets can guarantee a net loss to one side, assuming the specified bets are actually placed and settled. A simple example uses complementary events, since either an event happens or it does not.
If a buyer treats both contingent payments as worth too much, buying both can cost more than their combined guaranteed payoff. Each bet can look acceptable in isolation under the person's stated beliefs, so reviewing exposures independently can miss a loss built into the total payoff structure.
Coherence differs from accuracy. A perfectly consistent set of probabilities can still be a poor forecast of the world, and the theorem does not establish that coherent beliefs predict well or that every coherent investment earns a profit.
The direction of trading matters too, because formal arguments often assume willingness to buy or sell at the stated betting price while a real participant may refuse one direction, impose limits or have different prices for buying and selling. Risk preferences can complicate the interpretation, as a stated probability is not automatically a cash price a person will accept for any stake.
Wealth, risk aversion and the size of the position can affect decisions outside the simplified theorem. Trading frictions also matter, since fees, bid-ask spreads and position constraints can close an apparent price gap, so a mathematical combination is not necessarily an executable real-world arbitrage at the prices used in a spreadsheet.
Payoff definitions must match. Two contracts that sound complementary may use different observation times, settlement conditions or definitions of success, so they cannot be treated as a certain combined payoff until those details are verified.
Counterparty performance is another assumption, because a promised payment is not the same as a certain payment if default or dispute can prevent settlement, and credit and legal risks need their own assessment alongside payoff arithmetic. The philosophical argument uses vulnerability to loss to motivate probability coherence, and Stanford discusses qualifications and debates over that interpretation.
Those debates should not be hidden by claiming the theorem proves all aspects of rational decision-making. For a non-finance manager, compare related scenarios for internal consistency, ensure they are mutually exclusive, collectively complete where appropriate and based on matching definitions, then assess forecast quality, trading costs and practical risk separately.
In practice
Real-world examples.
Example
A risk team assigns probabilities of 70% and 50% to two exhaustive outcomes. It finds their combined probability is 120%, above one, and revises the assumptions before using them in a valuation. The revised inputs are checked again to confirm they add to 100%.
Example
An analyst prices two contingent claims separately. A combined payoff table reveals that the proposed purchase cost exceeds the maximum amount the pair can pay. The purchase is not approved until the prices are corrected.
Example
A trader sees apparent complementary contracts on different venues. The trader checks settlement definitions, fees and counterparty risk before calling the difference a guaranteed opportunity. One contract turns out to settle on a different date, which removes the apparent gain.
Formula
Calculation
Illustrative book: one claim pays $100 if event A occurs, and another pays $100 if A does not occur. If both are bought for $60 each, total cost is $120 and combined payoff is $100 in either outcome. The $20 loss is built in before costs, assuming the events are complementary and settlement is certain.
The $60 prices imply probabilities of 0.6 for each event, which sum to 1.2 rather than 1. If the same claims were priced at $40 each, the pair would cost $80 for a certain $100, a sure $20 gain to the buyer and a sure loss to the seller. At $50 each, the pair costs $100 for a certain $100, so no sure loss or gain arises.Case study
Seen in the real world.
Fictional case: An investment team evaluates contingent products in separate spreadsheets. Its analyst notices the probability assumptions conflict and builds a combined payoff table showing a sure loss under the proposed prices. The team revises the inputs and checks contract definitions. It also documents that removing this inconsistency does not prove the remaining probabilities are accurate or the investment is profitable.
The fictional team then adds a consistency step to its review template, so that any set of related scenarios is summed and compared against the payoff table before approval. Forecast quality is handled in a separate review that compares past predictions with outcomes. Keeping the two checks apart stops the team from treating tidy numbers as good forecasts.
Watch out
Common mistakes.
- Confusing coherent probabilities with accurate predictions.
- Ignoring buy-sell constraints, costs and settlement definitions.
- Claiming inconsistent beliefs cause an actual loss even when no bets are placed.
Questions
People also ask.
Does the theorem predict the outcome?
No. It concerns the payoff consequences of inconsistent prices or probabilities.
Does coherence guarantee profit?
No. Coherence avoids a particular inconsistency, not ordinary investment risk.
Are real-market opportunities automatically executable?
No. Costs, limits, contract definitions and counterparty risk matter.
From the founder's library
Take it further with the book.
Build your financial confidence beyond this definition. Shihan's full-length guide, Accounting Fundamentals, takes the same plain-English approach and turns it into a complete, practical playbook for non-finance managers, business owners and students - with chapter-end quiz answers and presentation slides included.
25% off with code MMHQ25, applied at checkout. Priced in USD - checkout may show the equivalent in your local currency.
View the book and save 25%