The morning of May 23, 1963 was cool and clear in Washington, D.C., sixty-four degrees with not a trace of rain. Nothing about the morning suggested what was coming. In a hearing room on Capitol Hill, Vice Admiral Hyman Rickover took his seat before the House Appropriations subcommittee and began, in the exacting way that had made him admired and feared in equal measure, to talk about the cost of building warships.
Congress listened to Rickover on this subject the way it listened to few others, because he had earned the standing the hard way. He was the father of the nuclear navy. Two decades earlier he had forced a skeptical Navy to build the world’s first nuclear-powered submarine, then ran the propulsion program with a rigor that became legend, personally interviewing every officer who would ever command one of his reactors. He did not deal in theory. He built the machines, sent men to sea in them, and held himself responsible for what came after.
Six weeks earlier, the ocean had been nothing like that quiet Washington morning. On April 10, 1963, the USS Thresher, the most advanced submarine the country had yet produced and one whose reactor Rickover had helped bring to life, went down during deep-diving trials off Cape Cod and did not come back up. All 129 men aboard were lost in cold, black water eight thousand feet down. It remains the deadliest peacetime loss in the history of the American submarine force. The cause was not the reactor, and not an enemy. It was a single brazed pipe joint, one connection among thousands, that gave way under pressure, flooded the engine room, and cost the boat her power at the depth where she could least afford to lose it.
What troubled Rickover most was how small the failure was, and how invisible. No one had set out to build a flawed boat. The joint had passed the inspections that existed. The problem was that the system had no way to see the weakness until the sea found it first. That conviction, that in complex work the fatal flaw hides where no one is looking, was what he carried into the hearing room that clear May morning. When he turned to the cost of defense, he was making the same point on a far larger scale. The government was buying on an enormous scale and had no dependable way to know whether the prices it paid were sound, because no two contractors kept their books the same way, and no one could compare one to the next. Whatever else might be wrong, the deepest problem was structural. It was accounting.
Congress came to agree. In 1968 it directed an inquiry into whether uniform cost-accounting standards were even feasible; in 1970 it created the Cost Accounting Standards Board (“CAS”), which over the following decade issued a body of rules governing how defense contractors must measure their costs and, above all, hold to whatever methods they have disclosed. That consistency requirement is the spine of the system. The individual measurement rules come and go, and the government is retiring several of them now, folding them into ordinary commercial accounting where the two have converged; the duty to stay consistent with what you disclosed does not move. One of those measurement subjects sounds clerical and in fact decides a great deal of money: depreciation, long governed by the standard numbered CAS 409.
One ledger, two readers
Depreciation is the quiet mechanism by which the cost of a long-lived asset is spread across the years it earns its keep. Inside a defense contractor, that single figure answers to two masters. The tax function uses it to reduce the company’s taxable income. The contracts function uses it to set how much of the asset’s cost the government will reimburse, by a method the contractor is bound to keep consistent with what it disclosed. The register is one document. The two readings rarely meet.
This is the seam. It runs through the fixed-asset register, and reading it means holding the tax return and the government contract in view at the same time, against two sets of rules written by different authorities for different purposes and almost never read together. Adjust how an asset is depreciated to capture a tax advantage and the contract-cost figure moves with it, automatically and usually unremarked. The connection is structural, and it is the kind of thing that stays invisible precisely because no single function owns both sides of it.
The new law
In December 2025, the annual defense authorization, the National Defense Authorization Act (“NDAA”), delivered the most consequential cost-compliance relief the contracting world has seen in a generation. It moved on two fronts, and they do not move at the same speed.
The first is settled. The threshold for certified cost-or-pricing data rises from $2.5 million to $10 million for prime contracts entered after June 30, 2026. Congress wrote that line directly into the statute; it takes effect on its own terms, with no rule required.
The second moved in two pieces, and only one of them is waiting on anything. The line at which the cost standards attach to a single contract rose from $2.5 million to $35 million in the statute itself, on the day the NDAA was signed. Congress did not ask for a rule there; it rewrote the section. Full coverage is the piece that waits: Congress directed that it begin at $100 million of government work rather than $50 million, but left that to be put in place by regulation within 180 days, and that rule is still in proposal as this is written. So the relief did not arrive all at once, and it did not arrive on one schedule.
There is a further wrinkle that no contractor should have to discover on their own. Cost standards reach a company through the clause its contracting officer wrote into the award. The statutory floor has moved while the regulation has not caught up, so what a specific contract requires turns on the clause that contract carries.
On its face, much of the middle market is being released from machinery it carried for years. The timing is the trap.
What’s at stake. The part of the relief worth the most is the part you have to reach for, and reaching is itself a regulated act. Letting a pricing threshold rise on its own costs nothing. Simplifying your cost accounting to shed coverage, or changing a method to capture the new freedom, is a formal event in the government’s eyes. Routed through the seam above and performed without care, it can hand the government a basis to recover money across the contracts a company still holds. “Less paperwork” can disguise the most expensive decision of the fiscal year.
How it works
The trap has three moving parts. The relief is real; the hazard is in how a company reaches for it.
The relief is prospective; the existing contracts are not. The higher thresholds govern new awards. They do not strip cost coverage from contracts a company already holds. A contractor that simplifies its books and abandons the methods it disclosed on existing covered work has not been deregulated. It has fallen out of compliance on the contracts still in hand.
Altering a cost method is a defined event. Government accounting has a precise name for changing how costs are measured, a change to a cost accounting practice, and a procedure that travels with it. For a voluntary change the rule runs one way: the government will not pay more because a contractor changed, and where the change raises the government’s costs, the difference flows away from the contractor. Omit the required notice and an auditor may withhold a portion of payments until it is supplied.
The auditors anticipate exactly this. The Defense Contract Audit Agency (“DCAA”), the Pentagon’s audit arm, has long instructed its examiners to establish whether a contractor’s accounting changes were disclosed, and disclosed in time. A wave of threshold-driven simplifications across 2026 and 2027 is among the most predictable audit themes in years; the companies that simplified quietly will be the ones the questions find.
Here the seam closes. The same year that loosened the contract rules also rewrote the tax code beneath them. A separate 2025 law, the summer tax act, restored full first-year expensing for equipment, added a deduction for qualifying production property, and again let domestic research be deducted immediately. A contractor reaching for those advantages is reaching into the very register the cost rules read. Claim the deduction without tracing its consequence on the contract side, and the tax victory and the audit exposure prove to be a single decision viewed from two directions.
One disciplined caveat, because precision is the point: most of the depreciation benefit is timing, cash recovered now rather than later, real but not permanent. The durable gains are the research deduction and the compliance relief itself. In asset-intensive companies, though, the first-year cash from correcting the register routinely funds the entire exercise in the year it is done.
Where things stand
The pricing-data line is live by statute: $10 million for prime contracts entered after June 30, 2026, on its own terms. The single-contract cost-standards line is also live by statute, at $35 million, and has been since the Act was signed. Full coverage is the piece still in motion: Congress mandated $100 million, gave the rule-writers 180 days, and that window has passed with the implementing rule still in proposal.
A separate rulemaking landed while this was being written. On July 8, 2026 the board that writes these standards issued a final rule, effective August 7, retiring the measurement standards for compensated absence and for material costs outright, and most of the standards for capitalization and depreciation, in favor of ordinary commercial accounting. That rule does not touch a single coverage threshold. What it does is remove the detailed measurement rulebook while leaving the duty underneath it exactly where it was: a contractor must still account the way it said it would, and a change to a disclosed practice is still a regulated event with a cost impact attached.
Which is the whole point. Two relief tracks and a rescission all landed inside eight months, and every one of them makes the same thing more valuable, not less: a clean, current record of what your practices actually are, kept through the transition rather than reconstructed after it.
What to do now
Treat the question, whether to take the relief and by what method, as a problem of engineering rather than announcement. If a company has already simplified its accounting to claim the new freedom, the exposure may already exist on the contracts it still holds. If it is about to, the discipline is the same. The starting material is the fixed-asset register and the cost methods built upon it. Document the present state. Design the intended one. Price the distance between them, in contract-cost terms on one side and tax-cash terms on the other, before the next method is altered, not after a contracting officer’s letter asks why the last one was.
Rickover’s objection, sixty years ago, was that the government could not trust numbers it had no way to compare. The principle now runs in the other direction. A contractor that cannot read its own asset register through both lenses at once, tax and contract, cannot know whether this year’s relief is a windfall or an invoice. Those who can read both will spend the coming eighteen months turning an unwieldy law into cash. Those who cannot will learn which it was when someone else performs the comparison for them.
FactorTax reads a single, engineered asset register through both lenses at once, federal tax and federal contract cost, and tells a contractor, before anything is altered, what the NDAA transition is actually worth. To see your own numbers through both readers’ eyes, schedule a discovery call.