In short
Bank guarantees in construction contracts are unconditional undertakings by a bank to pay the principal on demand, without the principal proving default. They are provided instead of cash retention, typically reducing at practical completion and released at the end of the defects period. They consume your bank facility for that whole time and are very difficult to stop once called.
A contractor signs a $2.4 million council contract. Clause 5 requires security of five per cent, reducing by half at practical completion. The contractor’s bank issues two guarantees of $60,000 each. Nobody thinks about them again.
Three years later the business is bidding a $6 million job and the bank will not extend the facility. The reason turns out to be that eleven guarantees totalling $340,000 are still outstanding across seven completed contracts, four of which finished more than two years ago. Nobody asked for them back. The facility they consume is the facility the new job needs.
Contract security is one of the least understood parts of a civil contract and one of the most consequential, because it sits at the intersection of your contract terms, your banking, and your capacity to take on the next job. This guide explains the instruments, the mechanics, and the commercial decisions worth making deliberately. For how security interacts with a prequalification financial assessment and your facility headroom, see our guide to demonstrating financial capacity in tenders.
Three things called security
Before anything else, a vocabulary problem. Australian civil contracts use “security” for at least three unrelated concepts, and contractors routinely conflate them.
| Term | What it means | Where to read about it |
|---|---|---|
| Contract security | Money or an undertaking you provide to the principal, which it can convert to cash if you default | This guide |
| Security of payment | The statutory regime that gives you a fast route to being paid | Security of payment in Australia |
| Security over assets | A charge your bank or financier takes over your property, plant or receivables | Plant and equipment finance |
They interact — the third is usually how the first is obtained — but they are governed by different rules and confusing them leads a contractor to look for an answer in the wrong place. Everything below concerns the first.
What contract security is actually for
Security exists because a principal is exposed to two risks that money in your hands does not cover.
- Performance risk during the works. If you stop, fail or become insolvent mid-contract, the principal has to engage someone else to finish, usually at a premium and always late. Security funds that difference without the principal having to sue you first — and if you are insolvent, without joining a queue of creditors.
- Defect risk after completion. If defects appear during the defects liability period and you do not rectify them, the principal can have them fixed and recover the cost. That is why security typically halves at practical completion rather than being released.
Understanding that second purpose explains the shape of every security regime you will meet: full amount during construction, reduced amount through the defects period, released at the final certificate. It also explains why arguing to have security released at practical completion rarely works — you would be asking the principal to give up the only protection it has for the period when defects actually emerge.
Two related instruments do different jobs and should not be confused with performance security. Bid or tender security, occasionally required on large or complex procurements, protects the principal against a bidder withdrawing or refusing to execute after award. A rehabilitation bond, common in quarrying and extractive operations, secures an environmental obligation to a regulator rather than a contractual obligation to a principal — see our guide to quarry ownership and pit operation.
The instruments, compared
Bank guarantees in construction sit alongside several alternatives, and most civil contracts allow security to be provided in any of them — which means the choice of instrument is the contractor’s more often than contractors realise.
| Instrument | How it works | Cost to you | Main drawback |
|---|---|---|---|
| Cash retention | The principal withholds a percentage from each progress payment | The full amount of your own cash, unavailable until release | Worst for cash flow — you fund it out of working capital as you go |
| Bank guarantee | Your bank undertakes to pay the principal on demand up to a stated amount | A fee, plus consumption of your facility limit, plus whatever secures the facility | Consumes borrowing capacity for years; usually secured against real property |
| Insurance or surety bond | An insurer or surety issues an equivalent undertaking | A premium, underwritten on your financial standing | Not accepted by every principal; underwriting takes longer than a bank guarantee |
| Parent company guarantee | A related entity guarantees your performance | No cash cost; exposes the parent | Only available in a group, and only worth what the parent is worth |
| Director’s personal guarantee | Directors personally guarantee performance or payment | No cash cost; unlimited personal exposure | Puts personal assets at risk. Take advice before offering one |
| Retention held in trust | Cash retention held in a statutory trust account | Same as cash retention | Better protected than ordinary retention, but still your money held by someone else |
The commercial comparison most contractors should be making is between the first two rows. A bank guarantee is nearly always better for cash flow than cash retention, because the money stays in your account and only a facility limit is consumed. The trade-off is that the facility is a scarce resource with its own constraints, and it is secured against something. Which of those two costs hurts more depends entirely on your balance sheet — the analysis in our guide to cash flow in civil construction contracts is the right frame for it.
Unconditional: the word that does the work
This is the single most important paragraph in the guide, and it is the part most often misunderstood.
A bank guarantee in Australian construction is normally an unconditional undertaking. The bank promises to pay the principal on written demand, up to the stated amount, without reference to you and without the principal having to establish that you defaulted. The bank is not asked to form a view about the dispute. It is not entitled to. It pays and debits your account or your facility.
Three consequences flow from that, and they surprise contractors every year.
- Your dispute with the principal is irrelevant to the bank. You may be entirely right about the variation, the extension of time, or the defect. The bank still pays, because its obligation is to the principal under a separate instrument, not to you under the construction contract.
- The undertaking is autonomous. Courts treat these instruments as akin to cash precisely so that they are reliable. That reliability is the reason principals accept them instead of cash, and it is bought at your expense.
- The fight happens afterwards. Once the money is gone you are the claimant, not the defendant. You are suing to get it back rather than resisting a claim, which is a materially worse position and a much slower one.
A conditional bond — where the issuer only pays on proof of default or of loss — exists and is far better for a contractor, but it is uncommon in Australian civil contracting and principals generally will not accept one. If a contract offers the option, take it.
What a bank guarantee really costs you
The fee is the least of it. There are four costs and most contractors only price the first.
- The issuance and ongoing fee. Usually a percentage per annum of the face value, charged for as long as the guarantee is outstanding. Modest per guarantee; not modest across eleven of them for several years.
- The facility consumed. A guarantee is a contingent liability that occupies a limit. That limit is finite and it is the same limit your overdraft and equipment facilities compete for. This is the cost that constrains growth.
- What secures the facility. Most banks will not provide an unsecured guarantee facility to an SME contractor. It is secured against real property — frequently a director’s home — or against a cash deposit, which defeats much of the point.
- The duration. This is the cost nobody models. A guarantee issued for a nine-month job typically stays outstanding for the defects period as well, and then for however long it takes somebody to ask for it back. Two to three years per contract is normal; longer is common.
The duration point compounds badly. A contractor running six jobs a year with two guarantees each, held for two and a half years, has around thirty outstanding at steady state. At five per cent of contract value, that is a substantial multiple of annual turnover in contingent commitments — which is exactly the number a financial assessor looks at when deciding whether you can take on a larger contract.
A practical caution on the form of the instrument: contracts commonly require the guarantee to be issued by an authorised deposit-taking institution, sometimes with a stated credit rating, and to be in the exact form of an annexure to the contract. Do not let your bank issue its own standard wording without checking it against the annexure. A guarantee in the wrong form can be rejected, and discovering that after award — when the guarantee is a condition precedent to commencement or to the first payment — costs time you do not have.
Insurance bonds and the surety market
An insurance or surety bond does the same job as a bank guarantee but is issued by an insurer rather than a bank. For a growing civil SME it is worth understanding, because it addresses the constraint that actually binds.
| Bank guarantee | Insurance / surety bond | |
|---|---|---|
| Consumes bank facility | Yes | No — this is the main advantage |
| Security required | Usually real property or cash | Usually indemnities rather than a property charge |
| Underwriting | Your existing bank relationship | Separate underwriting of your financial standing and track record |
| Speed to issue | Fast once a facility exists | Slower initially; comparable once a facility is established |
| Acceptance | Universally accepted | Accepted by many but not all principals; check the contract |
| Cost | Facility fee | Premium, which may be higher or lower depending on your standing |
The strategic point: bonding capacity and bank facility are separate pools. A contractor that provides all of its security through the bank is competing against itself for the same limit every time it wins work. Establishing a surety facility alongside the bank facility increases total capacity without increasing the charge over the director’s house — which is often the real constraint on growth. This belongs in the same conversation as the growth-stage financial planning covered in our guide to scaling a civil contracting business.
Two cautions. Check the contract before assuming a bond is acceptable — many government and council contracts specify an approved financial institution and some do not contemplate an insurer. And read the counter-indemnity you sign with the surety, because it defines what the surety can recover from you if it pays out, and it is frequently broader than contractors expect.
Recourse: when a principal can call
The bank’s obligation is unconditional, but the principal’s contractual right to make a demand is not. The contract sets out when the principal may have recourse to security, and this clause is worth reading before signing every single time.
Recourse clauses fall broadly into three drafting patterns, in increasing order of danger to you:
- Recourse on established entitlement. The principal may call where it has an entitlement to payment from you that has been determined or agreed. The narrowest and best for a contractor.
- Recourse on a bona fide claim. The principal may call where it claims in good faith to be owed money. Much broader, and the common position — it means a disputed claim can trigger a call.
- Recourse at large. The principal may have recourse to security at any time it considers appropriate, with no qualification at all. This effectively converts your guarantee into the principal’s cash on demand.
Two further provisions matter as much as the trigger. Notice before calling — some contracts require the principal to give notice a few business days before making a demand, which is the only realistic window in which a contractor can act. And an obligation to reinstate — where the principal calls on part of the security, most contracts require you to top it back up, which means a call costs you the money twice over until the dispute resolves.
Note also how this interacts with statutory payment rights. Set-off and recourse to security are a common principal response in a payment dispute, and the interaction between security recourse and security of payment rights is not always intuitive. Where a call is threatened in the middle of a payment dispute, that is a moment for advice rather than correspondence.
Can you stop a call?
Rarely, quickly, and expensively — and the honest answer is that most civil SMEs will not.
Because the undertaking is autonomous, courts are reluctant to restrain a call. The realistic route is an urgent injunction restraining the principal from making the demand, on the basis that doing so would breach the construction contract — not restraining the bank from paying. That requires:
- An arguable case that the call breaches the contract’s recourse provision.
- Moving before the demand is made, which usually means acting within the notice period if the contract gives one.
- An undertaking as to damages — you must be able to compensate the principal if the injunction turns out to be wrong.
- Legal cost and management attention, immediately, on a matter of urgency.
The practical implication is not that you should plan to injunct. It is that the leverage is in the drafting, not the litigation. A notice period and a narrow recourse trigger, negotiated before signing, are worth more than any remedy available afterwards. The dispute pathways once things have gone wrong are covered in our guide to dispute resolution after adjudication.
Reduction, release and the money that never comes back
Security comes back in stages, and every stage requires somebody to do something. Nothing happens automatically.
| Milestone | What normally happens | What you must do |
|---|---|---|
| Practical completion | Security typically reduces by half | Request the reduction in writing, referencing the certificate. It is not automatic |
| During the defects period | The balance is held | Rectify promptly. An unrectified defect is the main reason a release is refused |
| Final certificate | The balance is released | Request return of the original instrument, and confirm the bank has cancelled it |
| After release | Nothing | Confirm with the bank that the facility limit has actually been freed |
The final step is the one that goes wrong. A guarantee returned to you but never cancelled at the bank still consumes your facility. The principal has no interest in that step and will not do it. Getting the physical instrument back and confirming the cancellation with the bank are two separate actions, and only the second frees the limit.
An expiry date helps considerably. A guarantee expressed to expire on a date — rather than running until returned — cancels itself if nobody acts, though principals frequently resist a fixed expiry and some contracts require an open-ended instrument. Where a date is available, take it, and diarise it.
The completion mechanics that trigger each stage — certificates, defects, the final claim — are covered in full in our guide to practical completion, defects liability and the final claim. Security release is one of the items most often left undone when a job winds up and the team moves on.
Cash retention and trust account regimes
Cash retention is the older mechanism and remains common on smaller contracts and in subcontracts. The principal simply withholds a percentage of each progress payment, typically until it reaches a cap.
Three things distinguish it from a guarantee, and all three favour the guarantee.
- It is funded from working capital as you go. Every claim is short by the retention percentage, at exactly the time you are paying for the work.
- It sits in the principal’s account. Unless a statutory trust applies, retention is a debt owed to you and it ranks with other unsecured debts if the principal fails — the exposure covered in our guide to principal and head contractor insolvency.
- It is harder to get back. A guarantee is a discrete instrument someone must return. Retention is a number in a ledger that quietly stops being discussed.
Several Australian jurisdictions have responded to the insolvency exposure with retention trust regimes, requiring cash retention withheld from a subcontractor to be held in a trust account rather than mixed with the head contractor’s own funds. The regimes differ by state in their thresholds, the contracts they apply to and the obligations they impose, and they have been introduced and expanded at different times. Where one applies to you as a party holding retention, it brings account, record-keeping and notification obligations that are enforced. The statutory framework is covered in full in our guide to security of payment in Australia.
For a contractor that is both holding retention from subcontractors and having retention held by a principal, the important practical point is that these are separate obligations. Being owed retention does not excuse mishandling retention you hold.
Negotiating security before you sign
Security terms are more negotiable than most contractors assume, particularly with councils and private developers. Six asks, in rough order of how likely they are to succeed:
- Choice of instrument. Permission to provide a bank guarantee instead of cash retention, or an insurance bond instead of a bank guarantee. Frequently granted, and the single most valuable ask.
- A notice period before a call. Three to five business days’ written notice before the principal makes a demand. Reasonable-sounding, rarely refused outright, and the only thing that makes a remedy possible.
- An expiry date. A stated expiry tied to the end of the defects period, rather than an open-ended instrument.
- A narrower recourse trigger. Moving from “recourse at large” to recourse on a bona fide claim, or to an established entitlement.
- A reduction schedule. Where the contract holds the full amount through the defects period, ask for the standard halving at practical completion.
- A lower percentage or a cap. Least likely to succeed, and the one contractors ask for first.
Timing matters more than the asks. These are questions for the clarification window or the pre-award negotiation, not for after you have been named preferred tenderer and have lost most of your leverage. Note also that a departure from stated contract conditions may need to be lodged as a formal qualification, with the compliance risk that carries — see non-conforming and alternative tenders. The pre-execution stage generally is covered in our guide to contract award and mobilisation.
One thing worth pricing rather than negotiating: if the security regime is genuinely onerous — full amount held for a long defects period on an open-ended instrument with recourse at large — that is a cost of the job and belongs in the risk register and, if it is bad enough, in the go/no-go decision.
Taking security from your subcontractors
The same mechanism runs down the chain, and a contractor letting significant packages should think about it deliberately rather than copying whatever the head contract says.
- Match the periods. If your defects liability to the principal runs twelve months from practical completion, subcontract security that expires at completion of the subcontract works leaves you exposed for the whole of that period.
- Be realistic about who can provide what. A small subcontractor may not have a guarantee facility at all. Cash retention is often the only workable instrument at that end of the market — and if you hold it, the trust account obligations may apply to you.
- Do not over-secure. Demanding security disproportionate to the package narrows your subcontractor field and raises your prices, which costs you more than the risk it covers.
- Release it. A head contractor that is scrupulous about chasing its own security and casual about releasing its subcontractors’ has a reputation problem in a small market.
The wider discipline of letting and administering packages — scope gaps, back-to-back terms, and paying down the chain — is covered in our guide to engaging and managing subcontractors.
The security register nobody keeps
The single highest-value administrative change available to most civil SMEs on this subject takes an afternoon: build a register of every instrument outstanding and every dollar of retention held against you.
One row per instrument, with:
- Contract, principal, and contract number
- Instrument type, issuer, reference number and face value
- Date issued and expiry date, if any
- Date of practical completion and the reduction that should have followed
- End of the defects liability period, and the date the release was requested
- Date the original was returned, and the date the bank confirmed cancellation
- For retention: amount withheld to date, the release milestones, and amounts actually received
Then work the overdue rows. On a first pass most contractors find guarantees outstanding on jobs finished years ago and retention nobody has asked for. Both are recoverable, both are free money in the sense that they are already yours, and releasing the facility they consume is frequently worth more than the cash — because it is the capacity that lets you bid the next job.
Two habits keep it current: review the register at the same time each month alongside your job costing, and make “security released and cancelled” a line item on the job close-out checklist, so it is done while somebody still remembers the contract.
Checklist
- Do you know the total face value of every guarantee and bond you have outstanding, right now?
- How much of that relates to contracts that finished more than a year ago?
- Do you know how much of your bank facility is consumed by contingent instruments rather than borrowings?
- Have you asked your bank to confirm that released guarantees have actually been cancelled?
- On your current contracts, what is the recourse trigger — established entitlement, bona fide claim, or at large?
- Does any of them require notice before a call?
- Do your instruments have expiry dates, and are those dates diarised?
- Does the contract let you choose the form of security, and have you chosen the cheapest workable one?
- Has the guarantee wording been checked against the annexure form before issue?
- Have you investigated a surety facility as a separate pool from your bank facility?
- Does your subcontract security run for as long as your own defects liability to the principal?
- If you hold cash retention from subcontractors, do trust account obligations apply to you?
- Is security release a line item on your job close-out checklist?
- Do you know what a director’s guarantee you have already signed actually covers?
The short version
- Contract security, security of payment and security over assets are three different things. Only the first is what a principal holds against your performance.
- Security covers two risks: that you do not finish, and that you do not fix defects. That is why it halves at practical completion instead of being released.
- A bank guarantee is an unconditional undertaking. The bank pays on demand without the principal proving anything, and your dispute is irrelevant to it.
- Once it is called, you are the claimant chasing money rather than the defendant resisting a claim.
- The real cost of a guarantee is not the fee. It is the facility it consumes for two to three years and the charge over property that secures it.
- An insurance or surety bond does the same job without consuming the bank facility, and is worth establishing as a separate pool of capacity.
- Read the recourse clause. “At large” recourse converts your guarantee into the principal’s cash on demand.
- Stopping a call requires an urgent injunction against the principal, moving before the demand. The leverage is in the drafting, not the litigation.
- Nothing is released automatically. Request the reduction at practical completion and the release at the final certificate, in writing.
- A returned guarantee that was never cancelled at the bank still consumes your limit.
- Cash retention is worse than a guarantee on every dimension except simplicity, and unless a statutory trust applies it ranks as an unsecured debt if the principal fails.
- Negotiate the instrument, the notice period and the expiry date before signing. Ask in the clarification window, not after award.
- Build a security register. Most contractors doing it for the first time find years-old instruments still consuming capacity they need for the next job.
Sources and further reading
This guide is general information for Australian civil construction businesses and is not legal, banking, insurance or financial advice. The operation of a security or retention provision depends entirely on the words of the particular contract and the instrument issued under it, and the descriptions here are of common patterns rather than of any specific contract. Retention trust account regimes, and the circumstances in which a court will restrain a call on an unconditional undertaking, differ between Australian jurisdictions and continue to develop. Personal and parent company guarantees create exposure beyond the contract and should not be given without advice. Always work from the executed contract, the instrument as issued, and current advice from a construction lawyer and your financier.
- Australian standard-form construction contracts and their security provisions, which establish the pattern described in §02, §07 and §09 — security provided at a percentage of the contract sum, reduced at practical completion and released at the final certificate, with recourse available in defined circumstances. The forms in general use in Australian civil work, and how their risk allocation differs, are sourced in full in our guides to AS 4000 and AS 2124 and contract forms beyond construct-only. Amended and bespoke conditions frequently depart from the standard position, and the executed contract governs.
- The legal characterisation of unconditional undertakings described in §04 and §08, under which the issuer’s obligation is autonomous from the underlying construction contract and a court will generally restrain the beneficiary rather than the issuer. This is a developing area of Australian construction law and the availability of relief turns on the wording of the recourse provision and the facts. It is a matter for advice, urgently, if a call is threatened.
- State and territory security of payment legislation and the retention trust account regimes referenced in §10, which differ between jurisdictions in their thresholds, the contracts they cover, and the account, record-keeping and notification obligations they impose on a party holding retention. Sourced in full in our guide to security of payment in Australia.
- Related TenderBuilt guides carrying the primary-source detail referenced above: demonstrating financial capacity in tenders (how facility headroom and contingent liabilities are assessed at prequalification), cash flow in civil construction contracts, practical completion, defects liability and the final claim, principal and head contractor insolvency, and plant and equipment finance.