Guaranteed maximum price and lump sum answer different questions

An owner asking for cost certainty is usually offered two answers. A lump sum, where the contractor prices the work and carries the difference. Or a guaranteed maximum price, where the work is reimbursed at cost up to a ceiling, with savings below it shared or returned.

Both produce a number the owner will not exceed. They are not, however, alternative routes to the same place, and the choice between them is decided by something that has nothing to do with risk appetite: how complete the design is when the contract is signed.

Lump sum and stipulated sum are the same thing

Worth settling before anything else, because the terms are routinely presented as three options when there are two.

Stipulated sum is the formal name for a lump sum. The standard American owner–contractor agreement for a fixed price uses "Stipulated Sum" in its title; the industry says "lump sum" in conversation and "fixed price" in board papers. All three describe one arrangement: the contractor commits to a single price for a defined scope and keeps the difference between that price and its cost.

A guaranteed maximum price is the genuinely different form. Comparing a GMP to a lump sum and to a stipulated sum as though they were three choices produces a decision matrix with a duplicate row in it.

The third form worth naming is cost-plus with no ceiling at all — reimbursement of cost plus a fee, with the owner carrying the outcome. A GMP is that arrangement with a cap bolted on, which is why it inherits cost-plus administration rather than lump-sum administration.

What a lump sum requires

A lump sum is a price for a defined scope. Its integrity depends entirely on the definition being complete enough that the contractor can price it without guessing.

Where that condition holds, the form is excellent. The owner's exposure is fixed, administration is light, and the contractor is rewarded for its own efficiency. Change is the only route to a higher price, and change requires an instruction.

Where the condition does not hold, the form behaves differently. A contractor pricing an incomplete design has three options: price the risk, qualify the bid, or price low and recover through change. All three are used. The first produces a bid the owner considers uncompetitive. The second produces a price that is not actually fixed. The third produces the pattern owners describe as a claims-heavy contractor, which is frequently a rational response to being asked to fix a price on something that was not fixed.

The ceiling in a lump sum is only as firm as the drawings underneath it.

What a guaranteed maximum price does differently

A GMP reimburses actual cost and caps the total. Two consequences follow, and they are the real reason to choose it.

The owner sees the cost as it is incurred. Subcontract awards, rates, quantities, and the composition of the general conditions are all open. That visibility is the product; it lets an owner intervene while decisions are still open rather than after a price has been assembled behind a single figure.

The scope can complete during the work. Because the price is built up rather than fixed at award, design development that would be change under a lump sum can be absorbed within the buy-out, provided it stays inside the intent the ceiling was set against.

The cost of both is administrative. A GMP requires an owner able to audit open books, agree what is and is not reimbursable, and manage a contingency that sits inside the ceiling. An owner without that capability holds a cost-plus contract with a number written on it.

A lump sum tells you what you will pay. A guaranteed maximum price tells you what you are paying for. Those are different products, and an owner with an unfinished design can only really buy the second.

When the ceiling is set decides what it is worth

A lump sum price is fixed at award. A guaranteed maximum price usually is not — it is proposed part-way through design development, after the contractor has been engaged for preconstruction and has had time to price a set of documents that is advanced but unfinished.

That timing is the whole mechanism, and it cuts both ways. Set the ceiling too early and it is priced against so little information that the contractor loads it, or qualifies it so heavily that it is not a ceiling. Set it too late and the owner has lost the leverage that came from being able to walk, because the design is now complete and the preconstruction contractor is the only party who knows it.

The document that matters more than the number is the basis of the guaranteed maximum price — the schedule of drawings, specifications, clarifications, assumptions, and allowances the ceiling was priced against. Every later argument about whether something is inside the GMP or is a change is settled by reading that list. An owner who negotiates the figure hard and accepts the basis document without reading it has negotiated the wrong half of the agreement.

Allowances deserve particular attention, because an allowance is an admission that a scope was not priced. A ceiling carrying many of them is a forecast wearing the costume of a commitment.

Contingency sits in different places, and the difference matters

Both forms involve contingency. They are not the same money and they are not controlled by the same party, and conflating them is one of the more expensive misunderstandings available to an owner.

Under a GMP, a contractor's contingency normally sits inside the ceiling. It exists to absorb the contractor's own risks — buy-out that lands above estimate, coordination gaps, productivity. It is typically drawn at the contractor's discretion, within rules the contract sets, and whatever remains at the end is caught by the savings mechanism below.

The owner's contingency sits outside the ceiling, and it exists for a different purpose: owner-driven change, scope the owner adds, and decisions the owner has not yet made.

The failure mode is an owner who counts the contractor's contingency as protection against its own changes. It is not. When the owner's own decisions consume it, the contractor's risk cover is gone, the ceiling is reached early, and the conversation turns to whether the remaining work was ever inside the basis document. What each pot is actually for, and who may authorize a draw from it, is worth settling in the contract rather than discovering during it. What a contingency actually needs to cover takes that question further.

A lump sum has contingency too — the contractor's is simply invisible, priced into the number and never discussed again. The owner still needs its own.

What happens to savings is an incentive, not an accounting detail

If the work finishes below the ceiling under a GMP, the difference goes somewhere. How that is written determines how the contractor behaves for the whole job.

Returning all savings to the owner is the purest position and the weakest incentive: the contractor carries the risk of overrun but captures nothing from diligence. Splitting savings between the parties buys real effort on buy-out, value engineering, and subcontractor management — the contractor now has a reason to hunt for the cheaper compliant solution rather than the easier one.

The split is also where the mechanism can be gamed. A generous share against a ceiling that was set loosely rewards the contractor for the looseness rather than for the performance. The two terms are negotiated together or not at all, and a share that looks attractive in isolation is worth reading against the basis document that sets the ceiling it measures from.

Under a lump sum, none of this exists. Savings belong entirely to the contractor, invisibly, and the owner neither sees them nor argues about them. That simplicity is the form's genuine advantage and the reason it remains the right answer on a finished design.

Change behaves differently under each

Under a lump sum, change is the only route to a higher price, so change is where the contract's energy goes. Every instruction is a priced event, and the administration is a sequence of discrete negotiations against a fixed baseline.

Under a GMP, design development that would be a change under a lump sum can often be absorbed within the buy-out, provided it stays inside the intent the ceiling was priced against. That absorbs friction — until the ceiling comes into view, at which point the same question reappears in a harder form, because there is no longer room to absorb anything.

The practical consequence is that a GMP moves the argument later and concentrates it. Owners who read a quiet first year as evidence the contract is working are sometimes reading the absorption, not the performance. The change pipeline is a leading indicator is the reporting that tells the difference.

Where each one fails

A lump sum fails when the design was incomplete and the contract becomes an administrative contest over what the price included. The failure is not the form; it is the mismatch between the form and the definition.

A GMP fails in two ways. Where the ceiling was set against a scope that was too loosely described, the ceiling is reached and the argument is whether the remaining work is inside the original intent or is change — which is the basis document being read in anger. And where the owner does not exercise the visibility, the open book is filed rather than read, and the owner has paid for oversight it did not perform.

Both forms share a third failure that belongs to neither: a delivery model chosen before the design maturity was honestly assessed. Where an owner wants a fixed price and the design will not support one, a two-stage arrangement — engage early, price the second stage when the documents can carry it — is usually the honest answer rather than forcing either form. Two-stage procurement and the second stage sets out what the second stage has to contain to work.

Choosing

Three questions settle most cases.

Is the design complete enough to price without assumption? If a competent contractor could not produce a firm price without qualification, a lump sum will not deliver one either.

Does the owner need to see inside the cost? Where an owner needs to understand the buy-out — because the market is volatile, because the scope is still developing, or because there is a fleet of similar assets to learn from — a GMP delivers information a lump sum will not.

Can the owner administer an open book? Cost verification, reimbursable definitions, and contingency governance are real work requiring real capability. Without it, the ceiling is the only control the owner actually has.

A fourth question applies only if the answer to the first was no: when will the ceiling be set, and against what? A GMP chosen for the right reason still fails if the basis document is thin, so the date and the document list belong in the decision rather than in the mobilization that follows it.

The instinct that fixed price equals lower risk is understandable and often wrong. A fixed price on an unfinished design is a number that will move; the only open question is by how much, and through which mechanism.