
Rules
Part of Bullion authentication guide: evidence and limits
Bullion authentication problems: evidence and uncertainty
Bullion authentication problems: copied serials, tampered holders, conflicting tests, altered surfaces, and inconclusive results.
What to take away
- A real certification number can be copied onto a false holder.
- A sealed package can be genuine, replaced, opened, or paired with another item.
- Surface tests can miss internal substitutions.
- Cleaning, damage, and corrosion can erase diagnostics without proving a fake.
- Inconclusive is a valid result when the method cannot answer the question.
Bullion authentication problems often arise from a mismatch between the claim and the evidence being checked. Separate coin, bar, package, holder, label, serial, and grading claims.
Problem 1: the certification number exists
Finding a number in a database is a starting point, not the end. Use the certification lookup offered by PCGS, NGC, or ANACS, then compare the returned coin description, grade, label type, holder generation, and certification images with the physical object. Look for differences in marks and toning, not only design type.
Counterfeits reach ordinary buyers through ordinary channels. In a case announced by the U.S. Attorney's Office for the District of New Jersey, a man admitted selling fake gold bars to a coin collector and 49 counterfeit Morgan dollar coins to two other buyers.
He also imported counterfeit coins and bars into the United States in falsely manifested packages. A database entry describes a record, not the object in the buyer's hand. A copied number can point at genuine data while the item differs.
Problem 2: the slab looks opened
Photograph seams, welds, chips, clouding, glue, heat distortion, insert movement, and label alignment. Do not pry it open. Contact the grading service named on the label through its official submission route; for coins, services include PCGS, NGC, ANACS, and CAC Grading. Opening can destroy holder evidence and may affect service guarantees.
Problem 3: a bar passes XRF but fails another test
XRF commonly characterizes material near the tested surface. A plated or compositionally layered item may produce a plausible surface spectrum and a conflicting ultrasonic, conductivity, density, or dimensional result. Verify each instrument and repeat multiple controlled points before escalating to an internal or bulk method.
Problem 4: a serial matches the assay card
Matching print supports internal consistency between bar and card, but it does not independently prove either is genuine. Compare font, placement, seal, security feature, dimensions, mass, and material response with the precise issuer reference. For eligible PAMP bars, PAMP's VERISCAN provides an issuer-specific check; the Royal Canadian Mint's Bullion DNA feature applies to supported products. Check whether the manufacturer offers a serial verification service and what that service confirms.
Problem 5: surfaces are altered
Harsh cleaning, polishing, filing, corrosion, plating, putty, tooling, and artificial toning can change visual and instrument evidence. Do not remove the material. Record it and let the qualified examiner decide whether safe analysis is possible.
Federal rules draw a line the tests do not. The regulations under the Hobby Protection Act define an imitation numismatic item as one that claims to be, but is not, an original numismatic item.
Such items also include a reproduction, copy, or counterfeit of an original. If made in or imported into the United States, they must be plainly and permanently marked "COPY" on the obverse or reverse, never on the edge.
An unmarked piece is not thereby proven counterfeit, and the absence of a positive authentication opinion is not automatically a counterfeit finding either.
Problem 6: reference values disagree
Check product generation, gross versus fine weight, troy versus avoirdupois units, millimeters versus inches, proof versus bullion issue, alloy, and package inclusion. Dealer listings sometimes round or merge variants. Prefer the issuer's dated product specification.
Problem 7: a known-genuine control fails too
Stop. The instrument, mode, battery, calibration, temperature, sensor, package compensation, or operator method may be wrong. Do not label the unknown item from a method that cannot classify the control correctly.
Problem 8: experts disagree
Ask each examiner to state scope, reference, instrument, raw findings, and uncertainty. One may be judging grade, another holder authenticity, and another surface chemistry. For coins, submission routes include PCGS, NGC, ANACS, and CAC Grading; for bars, SGS provides precious-metals testing, but confirm that a U.S. facility accepts individual submissions and can answer the specific authenticity question. Ask about fees and shipping before sending anything; a second opinion is most useful when it tests a different property or reviews the same evidence independently.
Resolution table
First check
- Number matches, images differ
- Verify holder and coin side by side
- XRF passes, ultrasound fails
- Calibration, points, geometry
- Weight fails, dimensions pass
- Units, scale, package, alloy
- Visual concern only
- Exact genuine reference
- Damage blocks diagnostics
- Preserve surfaces
Better escalation
- Number matches, images differ
- Grading service
- XRF passes, ultrasound fails
- Qualified lab
- Weight fails, dimensions pass
- Repeat measurement and material test
- Visual concern only
- Experienced authenticator
- Damage blocks diagnostics
- Service or laboratory opinion
Common questions
Can a fake coin be placed in a real holder?
A holder can be tampered with, so compare physical coin, holder, label, and archived service data.
Does inconclusive mean likely genuine?
No. It means the available evidence or method did not support a reliable conclusion.
Should plating be removed to help XRF?
Not without owner authorization, transaction review, and a qualified test plan because removal is destructive.
Can altered surfaces affect grade without affecting authenticity?
Yes. Authenticity and condition are separate conclusions, though severe alteration may also block authentication.







