Representations and warranties vs indemnities

These three words are usually said together, almost as one phrase, as though they were a single protection. They are not. Representations and warranties promise that something is true; an indemnity promises who pays, and how much, if it is not — and the distinction changes what a buyer actually recovers when something goes wrong after closing.

What representations and warranties actually promise

A representation is a statement of fact as of a given date — that the financial statements are accurate, that there is no undisclosed litigation, that taxes have been filed — typically qualified by knowledge and materiality language and narrowed by disclosure schedules that carve out known exceptions. On their own, without an indemnity behind them, a breach is pursued through ordinary breach-of-contract remedies: the buyer sues, has to prove the statement was false, and has to prove the resulting loss, with no defined cap or basket unless one has been separately negotiated.

What an indemnity actually does

An indemnity is a separate promise to pay defined losses without forcing the buyer through a full breach-of-contract case. It is worth telling apart into two forms: a general indemnity that backs the representations, paying out if one of them proves false, subject to the survival period, basket and cap negotiated into the agreement; and a specific, standalone indemnity that names one known risk directly — a pending claim found in diligence, a defined environmental condition, a known tax exposure — and pays if that named event occurs, with no need to prove any representation was false at all.

Where the real difference sits

• A representation is a promise about a fact; an indemnity is a promise about who pays, and how much, if things go wrong
• A general indemnity is triggered by a breach of a representation; a specific indemnity is triggered by a named event occurring, whether or not anything was misrepresented
• Representations without an indemnity still give a remedy, but it runs through ordinary contract law rather than a defined basket-and-cap process
• A specific indemnity for a known issue usually sits outside the general cap limiting overall seller exposure, because it addresses something already identified and priced into the negotiation, not an unknown risk

Why buyer and seller pull in different directions

A buyer who found something concerning during diligence — a specific pending claim, an uncertain lease assignment, an environmental question — wants that item covered by a standalone specific indemnity, precisely because it removes the need to prove anything later, while also pushing for broad, lightly qualified representations to catch whatever diligence did not find. A seller resists both: an unqualified representation about a business they may not fully understand themselves is a large exposure to accept, and a specific indemnity naming a known risk directly and agreeing to pay if it materializes is a far harder number to negotiate down than the same risk buried inside a general representation.

What commonly goes wrong

A buyer sometimes relies entirely on broad representations and skips negotiating a specific indemnity for an issue diligence already flagged, then discovers the representation never actually covered it because a disclosure schedule carved the known issue out — which is exactly what disclosure schedules exist to do. The reverse failure is a seller agreeing to a specific indemnity for a named risk with no cap, assuming it will never actually materialize, and ending up with open-ended exposure on the one item everyone already knew was a real risk going in.

How to decide

Which tool fits which risk usually turns on whether the risk is known or unknown at signing. Broad representations and a general indemnity exist to cover the whole universe of things diligence did not catch; a known, already-identified risk is better handled with a specific indemnity naming it directly, since a general representation qualified by disclosure often ends up not covering it at all. Reviewing the disclosure schedules against the representations line by line, rather than assuming the representations alone provide blanket protection, is where this actually gets tested.