Unused construction contingency, and where it goes at the end

Unused construction contingency is the most predictable disappointment in the final accounting of a project. The balance shown in month eighteen is not the balance that returns, and the difference is consumed by three mechanisms that are visible well before anybody looks.

Why the balance falls fastest at the end

Contingency drawdown is not linear. It is slow in the first third, steady through the middle, and sharply faster in the last quarter.

Three things cause that shape.

Closeout finds everything that was deferred. Punch items, commissioning failures, incomplete work by trades that have demobilized, and coordination issues that were worked around rather than resolved all arrive in the same three months.

Trade contractors present their own accumulated claims at the end, and the general contractor settles them out of the contingency rather than passing them up as changes.

And the incentive changes. Where unused contingency returns to the owner in full, there is no commercial reason for the contractor to leave money in it at closeout. Where it is shared, there is a partial reason. Where it falls into the cost of the work, there is none at all.

That last point is not cynicism. It is what the contract instructs both parties to do, and it is settled by the drawdown and reversion language described in contingency drawdown rules.

The three mechanisms that consume unused construction contingency

Absorption of trade claims. Every trade contractor has an accumulated position at the end: extras never formalized, delays never claimed, extended supervision. The general contractor settles these to close out the subcontracts, and the settlements come from the contingency because they are not owner changes.

This is legitimate and it is also where the majority of late contingency goes on most projects.

Reclassification. Items previously carried as potential change orders, on which the owner had taken no decision, are resolved by drawing on contingency rather than by pursuing the change. The owner never sees the change and never sees the drawdown as related to it.

Closeout scope. Commissioning shortfalls, warranty items pulled forward, final cleaning, temporary facilities extended beyond the planned period, and attic stock. Individually small, collectively substantial, and almost always funded from contingency because none of them is a change.

None of the three is improper. Together they routinely consume between half and all of the balance visible six months earlier.

What the reversion language actually says

Four arrangements are common and they produce very different outcomes.

Full reversion. Unused contingency returns to the owner at completion, outside the shared savings calculation. The best position for an owner and the least common.

Into the savings pool. The balance forms part of the underrun and is divided at the shared savings percentage. So a $900,000 balance under a sixty forty split returns $540,000.

Into the cost of the work. The balance is simply available to be spent and whatever remains at the end is part of the final cost. In practice nothing remains.

Silence. The contract does not say. The outcome is then decided by the final accounting and by whoever argues more persuasively, which is rarely the owner.

Owners negotiating a guaranteed maximum price spend considerable effort on the contingency percentage and very little on this paragraph, which decides what the percentage was worth.

A worked example

A 59 million dollar project with a contractor contingency of $2.4 million. The contract provides that any unused balance forms part of the cost of the work savings, shared 60 percent to the owner.

At month eighteen, with the building substantially complete, the contingency shows a balance of $1.3 million. The owner reports an expected return of $780,000 to the investment committee.

Over the following five months the balance falls to $190,000. Trade contractor settlements take $640,000, closeout scope takes $310,000, and $160,000 is spent resolving three items the owner had assumed would be presented as change orders.

The final return to the owner is $114,000.

Every drawdown was permitted. None required consent. The owner learned the balance had moved when the final accounting arrived.

With notification above a threshold and a monthly drawdown log, the owner would have seen the trajectory in month nineteen and could have negotiated the trade settlements as part of a closeout discussion rather than receiving them as an outcome.

What to do from month eighteen onward

The window for influencing this is the last six months, and it requires three things.

Ask for the contingency balance and the drawdown log monthly, not the balance alone. The pattern in the final quarter is where the money goes.

Ask what open trade contractor positions exist. A general contractor closing out subcontracts knows what each trade is claiming, and that list is the best available forecast of where the contingency will go. Most will share it if asked before the settlements are made.

Separate closeout scope from contingency. Commissioning, attic stock, temporary facilities and final cleaning are all foreseeable and should have been in the price. Where they are being funded from contingency, that is worth a question, because it means they were not carried in the first place.

None of the three is confrontational. All three are questions about documents that exist, asked in the months when the answers can still change something.

Settling it at signature instead

Everything above is mitigation. The actual answer is four lines in the contract, agreed before it is executed.

Unused contingency reverts to the owner in full at completion, outside the shared savings calculation.

Drawdowns above a threshold are notified with a cause within a stated period.

Settlement of trade contractor claims above a threshold requires owner notification, and the owner may participate.

Closeout scope, named specifically, is part of the cost of the work rather than a contingency item.

Contractors resist the first and generally accept the other three. Even without the first, the other three change the outcome materially, because visibility during the final quarter is worth more than a reversion clause nobody monitors.

And the negotiation on all four is a fifteen minute conversation at signature against a six figure outcome at closeout, which is the usual ratio on contract administration provisions.

The same asymmetry applies to the other funds on the project. An owner who has settled reversion on the construction contingency and left the allowance schedule silent has closed one door and left another open, which is why the two are read together in allowance reconciliation.

Both funds behave the same way at the end of a job: they are consumed by items nobody classified, in a period when the owner attention has moved on to occupancy, and they are settled in a closeout reconciliation prepared by the other party.

The defense in both cases is the same, and it is not a clause. It is a monthly document with a cause column.

One last point about timing. The contingency balance is the number most often quoted to lenders and investment committees during a project, because it reads as a measure of safety. Quoting the month eighteen balance to a committee without qualifying how much of it is already spoken for is how an owner ends up reporting a return that does not arrive.

The honest version of that report carries three figures rather than one: balance, committed but not yet drawn, and expected at completion. All three are available from the same log.

What we do

We read the reversion language alongside the drawdown rules and say what the contingency is actually worth to the owner rather than what it says on the front page. Where a project is approaching closeout, we read the drawdown log and the open trade positions and set out what is likely to remain. The pre signature work is part of the readiness review. After signature, monthly owner cost assurance keeps the same reading running against each payment application.

Questions people ask