How does retention release work?

The two moieties, and what actually releases them

Retention is deducted from each payment — commonly a small percentage, capped at a limit — and held as security that the work will be completed and the defects made good.

The first half comes back at taking-over. The second comes back when the defects liability period ends and the outstanding defects have been closed. Those are the triggers, and they are events, not dates: a defects period that never formally closes is a retention that never formally releases.

Orphaned retentions are the failure mode

A retention held past its release trigger, with nobody accountable for noticing, is a classic audit finding. It is also an acid on subcontractor relationships — the subcontractor knows exactly what you owe them and exactly how long you have held it, even when you have forgotten.

And on the UK evidence it is a genuine insolvency exposure. Government research found billions held in retention at any one time, with hundreds of millions a year lost to upstream insolvency — and 44% of surveyed contractors had lost retention money that way. That is money already earned, sitting in somebody else’s account, disappearing when they fail.

The alternative, and the reform

A retention bond replaces cash retention with a bond — the security stays, and the subcontractor keeps its cash. It costs a premium, and on a cash-tight subcontractor that premium is often cheaper than the working capital it frees.

Retention reform is also a live regulatory question in the UK, with proposals to ban or ring-fence retentions under active consideration. Worth watching if you have UK exposure, and worth not building a business model on retention float either way.

Terms

• Retention
• Retention Ledger