Why Large Gold Deals Can Fail Even When the Paperwork Looks Substantial

More than a decade ago, I became involved in evaluating a small number of proposed gold and mining transactions through people I had met in the industry.

Some looked substantial on paper.

There were documents. There were intermediaries. There were discussions about mines, bullion, assay, banking and the movement of significant quantities of gold.

None of the transactions I helped bring forward ultimately completed.

I don't use that history as proof that the proposals were fraudulent or that large private gold transactions don't happen. I also don't know that every failure had the same cause.

What stayed with me was something more useful.

The difficult question was often not the amount of gold being discussed.

It was establishing what actually existed, who owned or controlled it, who had authority to transact, what could be independently verified and how money, title and metal could ultimately move.

That distinction still matters.

Why can a large gold deal fail even when the paperwork looks substantial?

A large gold transaction can fail because documentation is only one part of an executable transaction.

The parties still need to establish what product actually exists, who owns or controls it, who has authority to transact, what evidence supports those claims, where the gold came from, what compliance requirements apply and how payment, title and metal will settle.

A document can answer one of those questions.

It rarely answers all of them.

That leads to a simple rule:

Every document answers a particular question. The mistake is using one document as proof of a different question.

A gold deal can look convincing long before it is executable

Large transactions tend to accumulate paperwork.

That isn't inherently suspicious.

A complex transaction can legitimately involve corporate records, authority documents, product specifications, assay information, contractual documents, custody records and banking material.

The problem begins when the existence of the package itself is treated as proof.

A document showing that gold was assayed doesn't prove who currently owns it.

A document identifying a mine doesn't prove that finished bullion exists.

A document authorising an intermediary doesn't prove that the principal owns the metal.

A banking message doesn't prove that every commercial claim elsewhere in the transaction is true.

The more useful approach is to identify the proposition being made, then ask what evidence would actually establish that proposition.

Start with the product: what gold actually exists?

"Gold" can describe very different things.

A proposed transaction might concern:

• refined bullion
• dore
• gold stored with a custodian
• current mine production
• future production
• a mineral resource or reserve
• a financial interest connected to gold
• a tokenised interest linked to physical bullion

Those aren't interchangeable assets.

A mining project can contain a reported mineral resource without there being finished bullion available for immediate delivery.

Dore can contain substantial gold while still requiring sampling, assay and refining.

Stored bullion is different again because questions of custody and title become central.

Tokenised gold introduces another layer because the existence of backing bullion doesn't, by itself, establish what legal interest the token holder owns.

The first question therefore sounds simple:

What exactly is the product being offered?

Until that is clear, almost every later document is harder to interpret.

Existence, custody and ownership are different questions

Suppose evidence establishes that a quantity of gold exists in a vault.

That proves something important.

It still doesn't necessarily establish who owns it.

Custody tells you who holds or controls an asset physically or operationally.

Ownership asks who has the relevant legal title or property interest.

Control asks who can issue valid instructions concerning the asset.

Those relationships can overlap, but they shouldn't be assumed to be identical.

This distinction is already familiar in other parts of the gold market.

A vault operator can hold bullion without owning it.

A fund can own bullion for the benefit of its investors without each investor owning a specific bar.

A token structure can involve physical gold without giving its holder direct property title to the backing metal.

The same discipline belongs in a large physical transaction.

Follow the gold. Follow the ownership. Follow the authority.

Authority to transact is its own question

Large private transactions can involve brokers, representatives and other intermediaries.

The number of intermediaries doesn't prove that a transaction is illegitimate.

It can, however, make the authority chain harder to establish.

A letter of authority, mandate or similar document can be relevant evidence that someone has been authorised to perform a particular role.

Its existence alone does not establish that the underlying principal owns the asset, that the document is authentic or that the authority granted is broad enough for the transaction being proposed.

That is why an authority document needs to be read for the question it actually answers.

• Who issued it?
• To whom?
• For what purpose?
• For what period?
• What authority does it grant?
• Can the issuing principal itself be verified?

A credible authority chain connects back to an identifiable party with a defensible relationship to the asset.

Proof that gold exists is not proof that someone can sell it

This is where several distinct questions are often compressed into one vague idea of "proof".

One piece of evidence can sometimes support more than one field.

It shouldn't automatically be treated as proving all of them.

A useful transaction review keeps the questions separate until they can be connected.

Dore and assay answer narrower questions

Dore shows why product identification matters.

Gold dore is an intermediate material rather than the same thing as a finished retail or wholesale bullion bar.

At The Perth Mint, for example, published refinery material describes dore being melted and stirred so that representative samples can be taken before refining. Assay is then used to determine the precious-metal content of the material.

That can tell the parties something important about composition.

It does not, by itself, prove present ownership, lawful provenance, authority to sell or an executable delivery path.

An assay answers an assay question.

That sounds obvious.

In complicated transactions, it is surprisingly useful to keep saying it.

Provenance and compliance can't be treated as paperwork at the end

Gold has a physical history.

Where it came from can matter to refiners, financial institutions, regulated dealers and other counterparties.

LBMA currently lists Responsible Gold Guidance Version 9 as the applicable responsible-sourcing guidance for gold production for financial years beginning on or after 1 January 2022.

The framework concerns responsible sourcing by Good Delivery refiners and should not be confused with a general government licence for every gold transaction.

Singapore's current precious stones and precious metals regime covers regulated dealers, including intermediaries for regulated dealing.

Ministry of Law guidelines published in May 2026 address AML/CFT/CPF obligations under the applicable Act and regulations.

That does not mean every private gold transaction falls under the same regulatory provision.

It does mean compliance cannot safely be reduced to "the parties have documents".

The relevant parties, activity, jurisdiction and source of metal all matter.

Mining reports answer a different question

Mining disclosure is another area where a technically important document can be asked to prove too much.

The JORC Code deals with public reporting of Exploration Results, Mineral Resources and Ore Reserves.

As of 7 September 2026, JORC continues to identify the 2012 Edition as the current Code.

Its August 2026 update confirms that the Code Review remains underway. JORC's March 2026 update said the provisional replacement was moving through final review before formal approval and a later ASX consultation and transition process.

Canada's current NI 43-101 framework likewise governs standards of disclosure for mineral projects.

Current Canadian regulator records continue to show the consolidated instrument as the operative framework, while proposed changes should not be treated as an effective replacement unless and until they take effect.

In the United States, Subpart 1300 of Regulation S-K contains the SEC's mining-property disclosure requirements for registrants engaged in mining operations.

These frameworks can provide important information about mineral projects.

They do not, by themselves, establish title to a parcel of refined bullion being offered for immediate sale.

That distinction is fundamental:

Mineral-resource evidence is not bullion-title evidence.

Banking language doesn't prove the underlying gold transaction

Banking terminology can make a transaction procedure look highly formal.

That formality needs context.

MT103 is a legacy SWIFT Single Customer Credit Transfer message type.

Since 22 November 2025, ISO 20022 has become the standard for cross-border payment instructions under SWIFT's CBPR+ framework, although MT103 can still appear in limited contexts and through contingency conversion.

Whether the transaction refers to an MT103, an ISO 20022 payment message or another banking mechanism, the same distinction applies: the payment message does not independently prove that the seller owns the gold, that the gold exists or that every other representation in the proposed transaction is accurate.

The same restraint applies to standby letters of credit and guarantees.

ICC publishes rule frameworks used for independent undertakings, including ISP98 for standby letters of credit and URDG 758 for demand guarantees.

Which rules apply depends on the wording of the undertaking and whether those rules have been incorporated, while applicable law can also matter.

The presence of one of these instruments should not be turned into a universal badge of transaction legitimacy.

The name of a banking instrument tells you what mechanism is being discussed. It doesn't prove the underlying gold transaction.

This article is not a guide to structuring banking instruments.

A large discount is a question, not a verdict

A proposed gold price far below an observable market reference deserves scrutiny.

It does not, by itself, prove fraud.

The reason is that not every form of gold sits at the same stage of processing or has the same costs attached to it.

The World Gold Council's work on central-bank artisanal and small-scale gold purchasing shows legitimate pricing frameworks that can incorporate processing, refining, transport, operational costs, fineness and price risk when gold is purchased before it reaches monetary or London Good Delivery form.

That does not validate any particular private discount.

It demonstrates why the correct question is:

What explains the difference?

The further a proposed price moves from a relevant market reference, the more important it becomes to understand the product, location, purity, processing stage, costs, risks and commercial structure behind it.

Price can trigger investigation.

It cannot decide the conclusion by itself.

Settlement is where the earlier claims have to meet

Eventually the transaction has to work as one system.

• The metal must be identifiable.
• Title or control must be established.
• Authority must connect to the right party.
• Compliance requirements must be satisfied.
• Payment has to move through an acceptable path.
• Custody or delivery instructions need to work.
• Title needs to pass when the transaction says it will pass.

A proposed transaction can look impressive during individual stages and still fail when those stages cannot be connected into one coherent settlement path.

This is why settlement is not merely the final administrative step.

It is where earlier claims are forced to meet reality.

Why this matters to Gold.com.sg now

Years later, I acquired Gold.com.sg without a detailed plan for what it would eventually become.

Those earlier transactions had long since been put aside.

Building the site has brought some of the same questions back in different forms.

They appear when we look at bullion custody.

They appear when we examine the difference between physical gold and a financial interest.

They appear in tokenised-gold structures when backing, ownership and redemption are separate questions.

They appear in refining, provenance and market infrastructure.

The pieces were there long before the project was clear.

Only now have they become useful together.

Gold.com.sg isn't here to approve private transactions or perform due diligence for readers.

Its role is narrower.

It can help make the questions clearer.

And in complicated gold markets, that is often where useful analysis starts.

Frequently asked questions

Related reading