They Want a Personal Guarantee I Do Not Want to Give
The request usually arrives late, and it arrives phrased as routine. The terms are broadly agreed, the valuation is in, and then the lender asks for a personal guarantee, presented as the last formality before the file goes to committee. For an investor who built a corporate structure precisely to keep personal wealth and transaction risk apart, it is not a formality. It is a request to undo the structure with a signature.
Refusing outright, with nothing offered in its place, usually ends the financing. Signing reluctantly creates a liability that outlives the discomfort of the moment. The workable path runs between the two, and it starts with understanding what the guarantee is actually doing inside the lender’s analysis, because whatever it is doing can usually be done by something else.
What the request really signals
A personal guarantee does three jobs for a lender. It provides recourse beyond the borrowing entity, so that the debt does not stop at a company whose only asset is the one already mortgaged. It aligns the principal with the outcome, on the theory that people steer harder when their own balance sheet is aboard. And it compensates for whatever the lender could not see, measure or verify in the file.
The third job is the one that matters, because it is the one the borrower can influence. When the structure is unfamiliar, when the credit history sits in another country, when the accounts arrive in a foreign format, when the chain of ownership has not been documented to the last link, the lender is left with gaps. The standard institutional response to a gap is not more analysis. It is a guarantee. The signature is cheaper for the bank than the underwriting it replaces.
There is also plain habit. Many institutions lend against standard security packages, and the personal guarantee is printed into the template before anyone has read the file. A request that is standard is not therefore necessary, and credit officers themselves distinguish between what the template asks and what the credit needs, when a borrower gives them a reason to. Nobody makes that distinction unprompted.
Read this way, the request is information. It tells you which parts of the transaction the lender could not underwrite from the file it was given: the structure, the income, the history, or the risk itself. Identifying which gap the guarantee is covering is the first step to closing that gap by other means.
What signing actually costs
A personal guarantee is not a gesture of goodwill, and it is rarely priced into the deal as if it were real security, which it is. Depending on its wording it can give the lender recourse to personal assets, present and future, for the full amount of the debt, enforceable long after the transaction that justified it has been repaid down to a fraction of its original size. It sits outside the corporate structure and survives it.
It also quietly reverses decisions that were made deliberately. Limited liability, ring-fenced entities and the separation of assets are not decoration. They are the architecture of professional investment. A broad personal guarantee reconnects everything that architecture separated, so that the financing remains corporate on paper and becomes personal in substance. The structure still exists. It just no longer does its job.
Guarantees also accumulate. An investor active across several transactions, each with its own lender and its own signed guarantee, has built a concentration of personal risk that no single counterparty sees in full and nobody is managing. Each individual signature looked reasonable. The stack is not, and it surfaces at the worst possible moment, when more than one transaction is under stress at once.
The signature also reaches further than the person who gives it. Depending on the jurisdiction and the borrower’s marital property regime, a personal guarantee can expose assets the investor thinks of as the family’s rather than the business’s, and enforcement may arrive in a country other than the one where the guarantee was signed. An investor with a cross-border life has to read the guarantee across every jurisdiction it touches, which is exactly the kind of reading nobody does in the week of completion. The corporate structure was built, among other things, so that no single signature could ever do this. That is worth remembering at the moment one is requested.
The honest measure of any security is proportionality. Security should cover the risk the lender actually carries. An unlimited personal guarantee standing behind a conservatively structured loan, secured on an income-producing asset with documented repayment, covers risk the lender does not have. That mismatch is not a grievance to swallow. It is the argument to negotiate with.
What replaces it: security, information, structure
The first substitute is better real security, properly presented. The asset itself, valued and encumbered with a sensible margin between debt and value. An assignment of the rental income that services the loan. Where the borrower is a company, a pledge over its shares, which gives the lender a route to the whole entity rather than a claim on its owner. A lender holding good security, sized with room to spare, is holding protection a personal signature does not improve.
The second substitute is a file that leaves no gaps to fill. Cash flows and debt service capacity set out plainly. Adverse scenarios examined rather than avoided. A repayment strategy with dates and a fallback. The corporate chain reconstructed to the ultimate beneficial owners, with powers, approvals and signatures documented before anyone asks. Most guarantee requests are underwriting shortcuts, and a complete file removes the gap the shortcut was covering. It is slower than signing. It is also the version of the transaction that survives scrutiny.
The third substitute is scope. Where some personal undertaking genuinely cannot be avoided, it does not have to be unlimited, and it does not have to be permanent. It can be capped at a fixed amount rather than the whole debt. It can attach to specific risks, completion of works, a shortfall in interest, rather than to everything. It can carry release conditions in writing: a period of clean debt service, a level of occupancy, delivery of the asset. An undertaking with a cap, a trigger and an exit is a different instrument from an open guarantee, even if the lender files it under the same name.
Price belongs in the same conversation, because security and margin are two ends of one trade. A lender that keeps demanding the guarantee after the file is complete and the security is sized can be invited to state what the signature is worth: if it is real protection, its removal has a price in margin or in loan size, and if the lender cannot name one, the request was habit rather than analysis. Borrowers rarely ask the question, and it is clarifying for both sides. Sometimes the answer is a modestly smaller loan without the guarantee, which is very often the better transaction once the personal exposure is counted at its true cost.
Sequence decides most of this. Security packages are negotiable while the file is being shaped and nearly fixed once terms have been through committee. The moment to put alternatives on the table is the first serious conversation, not the week of signing, when every change reopens an approval someone has already obtained and defended internally.
Helping the lender say yes without it
A refusal to sign only works if someone inside the bank can defend it, and that someone is the credit officer who has to present the file to committee. The productive move is to treat the officer as the audience, not the adversary. A committee that receives “the client declines the guarantee” hears risk. A committee that receives “the guarantee has been replaced by an assignment of rents, a share pledge and a documented repayment fallback” hears a package. The difference between those two sentences is prepared by the borrower, not by the bank.
That means putting the alternative in writing before the request hardens. A short note that names the risk the guarantee was meant to cover, shows the security and information that now cover it, and states plainly what the borrower will and will not sign. Credit officers reuse well-drafted arguments internally, often word for word, and a borrower who supplies the argument has effectively drafted part of their own approval. A borrower who only says no has left the drafting to someone with no reason to try.
It also means conceding the legitimate part early. If the works phase genuinely depends on the sponsor, offer the capped completion undertaking before it is demanded, with its release condition attached. Volunteering the narrow instrument is what makes the refusal of the broad one credible. The position “nothing personal, ever” invites a template response. The position “personal exposure where it corresponds, sized to the risk, with an exit” reads as a professional counterparty, and it is the position that tends to win.
When the answer is a different lender
Some institutions lend on templates and will not move, whatever the file says. If the borrower’s profile does not fit the template, a foreign holding structure, income earned abroad, no local credit history, then the guarantee is the price of the mismatch, and no quality of preparation changes the printed form. Negotiating harder with a lender like that does not produce a better structure. It produces the same structure later.
The alternative is not to pay the price but to change the counterparty. There are lenders whose business is precisely the borrower the template cannot read: international structures, cross-border income, assets held through companies. In one transaction, a foreign investor who had been declined by a bank obtained financing from a lender specialised in real estate assets and international borrowers, without disproportionate personal guarantees and without dismantling the structure. The change of lender resolved a problem of credit criteria, not a problem of solvency.
That distinction carries the whole subject. If the guarantee request reflects a real weakness in the transaction, thin income against the debt service, no credible repayment story, leverage stretched to the ceiling, then the problem is the transaction, and it should be repaired rather than papered over with personal risk. If the request reflects what the lender could not see or was never built to assess, the remedies are the file and the choice of counterparty, in that order.
The instinct not to sign is usually sound. What makes it workable is arriving with something to hold in the guarantee’s place: security that covers the risk the lender actually runs, information that answers the questions the signature was silencing, and a lender whose criteria fit the borrower standing in front of them. A guarantee refused and replaced leaves a stronger transaction than a guarantee signed and regretted.
Montclare does not act as a lender. It structures transactions, prepares the file and coordinates financing with authorised institutions.