Why a Bank Declines a Good Transaction

A bank is not a single mind that weighs a proposal and reaches a view. It is a sequence of desks, systems and committees, each with its own mandate, its own template and its own reasons to stop. A transaction that would satisfy every one of them, if it reached them all in the right form, can still fail at the second step and never be seen by the rest.

That is how good transactions get declined. Not because someone examined the asset and judged it poorly, but because the file met an obstacle that had nothing to do with the asset, at a point in the process where nobody was looking at assets yet. Understanding where those obstacles sit is the difference between blaming the market and preparing a file that reaches the people who can approve it.

The transaction never reaches the person who would approve it

Most credit processes begin with a filter that no human designed for your case. An automated screen checks the borrower against a scoring model. A policy list checks the jurisdiction of the asset and the country of residence of the beneficial owner. A product code decides which desk owns the file. None of these steps asks whether the deal is good. They ask whether the deal is theirs.

An international investor fails these filters routinely and invisibly. There is no local credit record for the scoring model to read, so the model returns nothing rather than something bad, and nothing is treated as a negative. The borrower is a company in another country, so the corporate verification path is longer than the standard one and the file is routed to a queue rather than a desk. The asset sits abroad from the perspective of whichever branch received the enquiry.

The result is a decline produced by a system that never formed an opinion. This is worth stating plainly, because borrowers assume the opposite. They assume that somewhere in the institution a professional looked at their property, their income and their plan and decided against them. In most early declines of cross-border files, nobody did. The transaction stopped before the assessment began.

There is a structural reason for this design, and it is not incompetence. A retail and commercial credit process exists to handle volume: to say yes quickly to the profiles it recognises and no cheaply to everything it does not. The filters are what make that speed possible, and they are calibrated on the population the institution actually serves. An international investor with a holding structure is, by construction, outside that population. The machine is not failing when it stops such a file. It is doing exactly what it was built to do, applied to a case nobody built it for.

The practical consequence is that improving the economics of the deal changes nothing at this stage. Offering more equity, accepting a lower loan or reducing the price cannot solve a routing problem, because the filter is not measuring any of those things. What solves it is arriving at the right desk with the file already in the shape that desk expects.

Appetite is a moving target and it is rarely published

Every lending institution runs an internal appetite that shifts with its balance sheet, its regulatory position and its recent experience. It has more room for some asset classes than others this quarter. It is at or near its allocation in certain jurisdictions. It has just absorbed a loss in one sector and has quietly stopped writing new business there for a while.

None of this appears in the marketing material, and the borrower never hears it as the reason. A file declined because the institution is full will be declined in the language of criteria and fit. The banker often does not know either, or knows and cannot say. So the borrower receives a refusal that reads like a judgement on the transaction when it was a statement about the lender’s own position on a particular Tuesday.

Appetite also has a shape, not just a level. An institution may finance hotels but not hotels with an operator it does not know. It may lend on land but only with planning in place. It may accept foreign borrowers but only through certain corporate forms. A transaction can sit one degree outside a specific appetite and be declined as decisively as one that sits far outside it, and from the borrower’s side the two look identical.

Appetite is also owned by people, and people move. A file that fits an institution’s policy perfectly can still fail because the officer who understood this type of transaction has changed roles, and the colleague who inherited the enquiry has no reference point for it. Conversely, a single credit officer with experience of cross-border structures can carry a file through an institution that looks, from the outside, unlikely to finance it. Borrowers tend to think in terms of banks. The decision is usually made by a person inside one, and knowing which person is often worth more than knowing which bank.

This is why the previous relationship is a weak guide. A bank that financed a similar transaction two years ago may have no room for it now, and a bank that declined one last year may be actively seeking them today. Testing current appetite before submitting is not diligence for its own sake. It is the difference between a submission and a lottery ticket.

The file makes the transaction impossible to assess

Set aside routing and appetite, and a third category remains, larger than either. The transaction reaches the right desk, at an institution with genuine appetite, and is still declined because the file does not let anybody say yes.

Credit officers work under time pressure and against internal deadlines. A file that answers its own questions can be assessed in a sitting. A file that requires three rounds of requests to establish who owns the borrower, where the income comes from and what repays the loan consumes far more of that officer’s week than the transaction is worth to them. Somewhere in the second or third round, the cheapest available decision becomes no, and it is taken without malice and without anyone concluding that the deal was bad.

The gaps that trigger this are consistent. The ownership chain is not mapped to the ultimate beneficial owners. The sources and uses do not add up, or exist only in the borrower’s head. The income is real but documented in a format the analyst cannot reconcile. The repayment plan is implied rather than stated. Adverse scenarios are absent, which reads either as optimism or as concealment, and both cost the same.

Timing inside the institution compounds all of this. Credit committees meet on a calendar, and a file that arrives incomplete two days before a sitting does not get assessed late, it gets assessed at the next one, which may be weeks away. The borrower experiences the gap as inexplicable silence and often fills it with pressure, which rarely helps. A complete file submitted with the committee calendar in mind can save more time than any amount of chasing, and the calendar is information any banker will share when asked directly.

There is also the harm done by circulation. A file sent simultaneously to six institutions comes back with six partial reactions, and each institution that declines leaves a trace in a market where credit people speak to each other. A transaction that has visibly toured the market arrives at the right lender already tired, and the lender begins by wondering what the others saw. The remedy is not secrecy but sequence: build the file, choose the readers, approach deliberately.

What the decline is really telling you

Read this way, most declines of good transactions are messages about process rather than verdicts on value. That does not make them harmless, and it does not make them all the same, but it does mean they can be diagnosed.

A decline in days, with no questions asked, is a filter or an appetite decision. Nothing about the deal was examined, and the response is to find an institution whose ordinary business this is. A decline after several rounds of documentation requests is a file problem: somebody wanted to assess it and could not, which is the most fixable category of all, and the most expensive to ignore, because the same file will produce the same outcome everywhere. A decline after genuine underwriting, with questions about coverage, valuation and exit, is a statement about the transaction, and it deserves to be taken at face value.

There is a fourth category worth naming, because it is invisible from outside: the transaction that is declined for a reason the institution cannot state. A connected exposure elsewhere in the group. An internal concentration limit on a postcode or a sector. A compliance question about a jurisdiction in the ownership chain that the bank will not discuss with the client. These declines are real, they are final at that institution, and no file quality overcomes them. Their signature is a refusal that stays vague under direct questioning, from a banker who was previously specific about everything else. When that happens, the useful conclusion is simply that this door is closed for reasons that belong to the bank, and the only productive move is another door.

The one response that never helps is speed for its own sake. Sending the same file to the next institution the following morning repeats the experiment without changing the variable. The pause after a refusal is not lost time. It is the only moment in the process when the borrower can see, at no cost, exactly which part of the machine stopped them.

The institutions that finance cross-border transactions as ordinary business exist: banks with international desks, family offices, specialist credit institutions. They apply the same standards to the numbers as anyone else, and they are no softer on solvency. What they have is a process built to read a foreign borrower, a corporate chain and an asset in another country without treating any of it as an exception. Reaching one of them with a complete file is the whole of the work, and it is work the borrower controls.

None of this is an argument that banks are wrong to work as they do, or that a borrower is owed an approval. It is an argument for aiming. A good transaction shown to the wrong institution, in the wrong format, at the wrong point in its cycle, will be declined as reliably as a bad one, and the borrower learns nothing from the experience except discouragement.

The bank that declined your transaction probably never decided it was bad. It decided it was not theirs, or not readable, or not this quarter. Those are three different problems with three different answers, and none of them is a reason to change a transaction that works.

Montclare does not act as a lender. It structures transactions, prepares the file and coordinates financing with authorised institutions.