The email arrived on a Tuesday afternoon and ran to four lines. The facility could no longer be supported on the terms previously indicated. We were welcome to discuss alternatives. Completion was booked for two weeks on Friday.
What made it worse was that we had done, by our own reckoning, everything right. We had a broker. We had an offer on the lender's letterhead. We had a solicitor who had been working on it for six weeks. What we had not done was read the offer as the conditional document it plainly was.
An offer in principle is a marketing document
The vocabulary is designed to sound more solid than it is. A decision in principle, an agreement in principle, an indicative offer: all of these are produced from information nobody has verified, often by a system rather than a person, and none of them commit the lender to anything.
The stage that starts to mean something is a formal offer that has actually been through credit — a real underwriter, or a credit committee, looking at real numbers. Even then, what you are holding is a list of conditions with a number printed at the top. Most people read the number.
A lender's offer is a list of conditions with a number at the top. Most people read the number.
What "subject to" meant in ours
We went back through the offer after the email, which is exactly the wrong week to read it properly. The conditions precedent ran to a page and a half.
A satisfactory valuation from the lender's own panel. A debenture over the company's assets. Personal guarantees from both directors. A solicitor's certificate of title. Management accounts to a date no more than three months before drawdown. Written confirmation that there had been no material adverse change in the business since the application. Bank statements to within thirty days. Evidence of the source of the deposit. Insurance with the lender noted on the policy. And a landlord's consent that, in our case, nobody had started chasing.
Every one of those is a place where the deal can stop. Several of them are outside your control entirely, and two of them are re-tests of things you thought had already been assessed.
The three things that actually killed it
The valuation came in low. We had agreed £480,000 for the premises; the lender's valuer put it at £430,000. The offer was 70% loan to value, so the facility fell from £336,000 to £301,000 and a £35,000 hole opened up two weeks before completion. Down valuations are the single most common way a property-backed deal dies, and the borrower almost never gets to choose or challenge the valuer.
The management accounts were re-tested. The offer required accounts to a recent date, and ours by then covered a quarter that had been softer than the one we applied on. Nothing dramatic — a dip. But the affordability calculation was rerun against the newer figures and came out tighter, which turned a comfortable file into a marginal one.
And the lender's appetite for our sector had changed between application and completion. Nobody says this in writing. You find out because a broker who has placed forty deals tells you they have watched the same lender go quiet on three files that month.
That third reason is the one worth internalising. A deal can fail for reasons that have nothing to do with you, your accounts or your business, and no amount of preparation prevents it. What preparation does is make sure that when it happens, it is survivable.
The money you lose anyway
The facility never drew down, and we were still out of pocket. The valuation fee of £1,100 was paid up front and is not refundable, because the valuer did the work. A commitment fee, paid to reserve the funds, was refundable in our case only in part. Our solicitor had six weeks on the file. The lender's legal costs, which you agree to pay regardless of whether the deal completes, arrived anyway. The broker fee was contingent on drawdown, which was the one piece of good news.
All in, a deal that produced nothing cost somewhere between four and six thousand pounds, plus the deposit exposure on a purchase we then had to renegotiate. That is the real reason to take conditions seriously: the failure is not free.
What we do differently now
We read the conditions list before we celebrate. On any offer, the first questions are which conditions are outstanding, who owns each one, and what the realistic lead time is for the slowest. A landlord's consent nobody has requested is a four-week condition sitting quietly in a two-week timetable.
We ask three direct questions of every offer, in writing. Has this been to credit, or is it an underwriter's indication? Which fees are non-refundable, and from what point? And what would cause you to withdraw this offer? A good broker or relationship manager answers all three plainly, and the quality of the answer tells you what the offer is worth. What a lender actually asks for before they approve a loan covers the documents side of the same question.
We keep a second lender warm to the point of a written offer, and we accept that this costs a second set of fees. On a deal with a hard completion date, that duplication is cheap insurance. It also does something subtler: it removes the desperation that makes people accept worse terms in the last fortnight.
We form our own view on value before we agree a price, using comparable sales and, where the sum justifies it, our own independent valuation. You cannot stop a down valuation, but you can avoid being the last person to know the property was optimistically priced.
We keep the management accounts current and clean, because they will be re-tested at drawdown. A file that was approved on figures to March and completes in September is judged on September, and what a lender sees in your bank statements applies to that second look just as much as the first.
And we negotiate the timetable rather than accepting it. A longer completion window, or a finance condition in the purchase contract where the seller will wear one, converts a catastrophe into a delay. Sellers refuse this often enough that it is worth pricing the risk into the offer when they do.
The part nobody tells you
The deal survived, in the end, with a different lender, a lower loan, a bigger deposit taken out of working capital, and terms that were worse by roughly the amount of the shortfall. That is the ordinary outcome. Deals rarely die outright; they get more expensive and less pleasant, and the cost lands on the borrower who had run out of alternatives.
Two related points are worth reading before you sign anything on the next one: personal guarantees on business loans, because the replacement facility asked for more of them than the original, and what loan covenants actually commit you to, because the tighter file came with tighter tests attached. If you are weighing the property decision itself rather than the funding, commercial mortgage versus leasing is the piece to start from.
The lesson is not that lenders are unreliable. It is that an offer is a conditional document, and the person who reads the conditions the day it arrives is in a completely different position from the person who reads them the day it collapses.
Common questions
Is a decision in principle binding on the lender?
No. A decision in principle, agreement in principle or indicative offer is based on information the lender has not verified, is often generated automatically, and commits nobody to anything. It is useful for confirming that a deal is roughly the right shape and for showing a seller you are serious, and that is all. The stage that carries weight is a formal offer that has been through an underwriter or credit committee against verified information, and even that is issued subject to conditions precedent. If you need to know where you stand, ask directly whether the file has been to credit or is still an indication.
Can a lender withdraw a formal loan offer before completion?
Yes, and it happens more often than borrowers expect. Formal offers are issued subject to conditions precedent, which typically include a satisfactory valuation, security being granted, management accounts to a recent date, confirmation of no material adverse change, and evidence of the source of funds. If any condition is not met the offer falls away. Lenders can also withdraw where their appetite for a sector or a loan type changes between offer and drawdown, which has nothing to do with the borrower. The practical protection is to know which conditions are outstanding, chase the slowest ones early, and keep an alternative lender live.
What fees do I lose if a business loan falls through before drawdown?
Usually the valuation fee, because the valuer has done the work regardless of the outcome, and your own solicitor's time to the point the file stopped. Lenders' legal costs are normally payable by the borrower whether or not the facility completes, so expect that invoice too. Commitment or arrangement fees vary: some are refundable if the lender withdraws, some only in part, and some not at all, which is why you should establish from the outset which fees are non-refundable and from what point. Broker fees are frequently contingent on drawdown, but confirm that in the terms of business rather than assuming it.
How do I reduce the risk of a lender pulling out at the last minute?
Treat the conditions list as the real document. Establish which conditions are outstanding and who owns each, and start the slow ones — landlord consents, insurance endorsements, certificates of title — immediately rather than at the end. Form your own view on the property's value using comparable sales before agreeing a price, since down valuations are the most common cause of collapse. Keep management accounts current, because they get re-tested at drawdown. Run a second lender to written offer stage on any deal with a hard completion date, and negotiate a longer completion window or a finance condition where the seller will accept one.
