PNJ's diamond scandal: is the moat destroyed, or is this temporary?

A subsidiary accused of falsifying diamond certificates for two years. The parent's stock loses half its value in three weeks. Thirty businesses in the same industry close because they cannot take the pressure. Amid all those numbers I see one question genuinely worth asking, and almost nobody is asking it properly.

What happened

In early July 2026, Thanh Hoa police broke up a transnational diamond smuggling ring. [FACT] Among the accused was the former director of a certification company wholly owned by PNJ, charged with buying smuggled diamonds, grinding off the original GIA numbers, and issuing new certificates under his own company's name.

Worth noting: this is not only PNJ's problem. The ring also moved goods through SJC's own certification unit. More than thirty other gold and jewellery businesses in Ho Chi Minh City closed during the month, not because they were under investigation, but because customers all wanted to sell back and they did not have the cash to pay.

PNJ held. But the cost was not small: [COMPANY STATEMENT] the company disclosed spending more than 7,000 billion dong buying jewellery back from customers in just 27 days, at one point 3.7 times the same period's revenue. The stock fell 51% from its high in less than half a year. The company then voluntarily provisioned 865 billion dong, making Q2 2026 its first quarterly loss in five years.

At this point most investors split into two camps. Camp one: "the stock is down 50%, it must be cheap." Camp two: "the company is in a scandal, stay away." I think both camps are asking the wrong question.

The question I ask first, and it is not "has the price fallen enough"

When analysing a business in crisis, I borrow Buffett's distinction between two things that look identical on a price chart but are fundamentally different: is this a temporary difficulty, or is the moat permanently destroyed?

A jewellery brand has a moat not because the name is pretty. It has a moat because customers believe what they are buying is genuine: the right gold purity, diamonds matching their certificates. People pay 15–20% more than at an ordinary gold shop precisely for that belief. Take the belief away and PNJ is just a chain selling precious metals at margins no different from the neighbours.

And here is what I want to say plainly: the P-Lab affair does not hit a peripheral corner of the business, it hits the very thing that creates the moat. Not an operational error, not an out-of-fashion product. It hits the one function that defines the difference between PNJ and a market-stall gold shop.

So the question worth asking is not "is the stock cheap yet". It is: how much of that trust has been lost, and can it be regained?

Answering that needs data, not sentiment. I see two notable signals:

Leaning toward "containable": this is the behaviour of one individual at one subsidiary, not direction from group management. And more importantly, the ring spread across the industry, including the certification unit of the largest competitor. This is a hole in Vietnam's entire diamond certification system, not a defect unique to PNJ.

Leaning toward "possibly deeper than assumed": 7,000 billion of buy-backs in 27 days shows the scale of doubt is not small, far from the initial statement that "smuggled goods did not enter the system". And the volume sold during 2022–2024, when this subsidiary is accused of operating improperly, is already in customers' hands. I do not know how many such time bombs are out there, and I do not think anybody knows for certain.

I do not have enough data to answer decisively right now. And I think admitting that matters more than pretending otherwise.

What almost nobody notices: half of PNJ is no longer PNJ

This is the part I find most interesting, and it has nothing to do with the scandal.

I use a simple test to separate a business with a real franchise (customers think nothing can replace it) from a pure commodity seller (customers only care about price). Gold bullion is the classic example of the latter: a tael of SJC gold at PNJ and a tael of SJC gold at the shop next door are exactly the same thing to the buyer.

In the first half of 2026, 24K gold bullion's share of PNJ's revenue rose from 21% to 38%. Total revenue growth of 49% sounds impressive, but when I unpack it, most of that growth comes from selling something with no moat, at a far thinner margin than jewellery.

In other words: even if the P-Lab affair is resolved cleanly tomorrow, PNJ is still gradually turning a substantial part of itself into a high-volume gold shop. I do not think this is absolutely bad; gold is rising, liquidity is good, and it is less risky than holding jewellery inventory. But I treat it as a story of franchise dilution, running quietly alongside the scandal, and almost nobody mentions it because all attention is on diamonds.

The integrity test I use is not "are they lying"

There is one question I always ask about any management team, borrowed from how Buffett assesses people: is this person willing to tell me what they do not know, or do they only say what makes them look good?

PNJ earns a rare credit here: they voluntarily provisioned 865 billion dong for an event occurring after the accounting cut-off, a choice not technically required. They publicly disclosed the buy-back-to-revenue ratio, a figure listed companies rarely volunteer because it is unflattering. To me that is the behaviour of a management team choosing to face bad news early rather than bury it in a later quarter's footnotes.

But I also note a small crack: the initial message was "smuggled goods did not enter PNJ's system, everything is stable", while the actual buy-back scale afterwards was far larger than that message implied. I do not call that an outright lie. But I treat it as the gap between "the full truth from the start" and "the truth revealed gradually as circumstances force it", and that gap, however small, goes in the notebook.

Why I do not think "cheap" is the opportunity here

This is where I see most Vietnamese individual investors make their heaviest error, including those who have read Graham.

"Cheap" is not a low price against an old high. To me "cheap" is the gap between price and the business's real value. A stock down 50% whose real value is also down 50% is not any cheaper; it is simply reflecting what has been lost.

To avoid speaking in generalities, I tried to quantify it. I calculated owner earnings, the profit genuinely belonging to shareholders after deducting reinvestment needs in working capital, since PNJ is inventory-intensive, under three scenarios reflecting how well the company gets through this:

The multiples here are well below what I would normally pay for a business with a stable moat, because I currently classify PNJ as "carrying uncertainty", not "clearly excellent". The reason is simple: the independent audit result does not exist yet. For the uncertain category, the margin-of-safety principle I follow demands a 40–50% discount to intrinsic value, not the 20–30% applied to a stable business.

Applying that discount to the base-case intrinsic value (~32,800) gives the range I would buy in: roughly 16,000–20,000.

The current price is around 33,000, essentially equal to base-case intrinsic value, with no discount at all for what I do not yet know. In other words: the margin of safety here is close to zero. The market is pricing as though the crisis is over and the audit will certainly come back clean, while in reality neither has been confirmed.

And here is the practical question I ask: in accepting this level of uncertainty, what reward am I demanding over simply holding a government bond at a certain 5% a year? If the answer is "almost no extra reward for the risk being carried", I treat the case as not yet attractive, however persuasive the story.

What I will watch next

I am not closing this file. Here is what I am watching over the coming months, in descending order of importance:

• The independent audit of the entire import-production-sales chain that PNJ committed to. This is the real deciding milestone; everything else is waiting.
• Whether the criminal investigation widens to PNJ management beyond P-Lab.
• The reviewed Q3 2026 financial report: are there further special provisions, and what is real operating cash flow?
• Retail revenue growth excluding 24K gold: can it stay above 10%, or is it stalling on customer sentiment?
• The buy-back-to-revenue ratio: does it keep falling back to a normal baseline?
• Insider transactions: are executives and related parties buying more, or still selling?

So what do I think now

This is a business I can understand in a paragraph, with a real moat, not an imaginary one, but that moat has just been punctured at its most vulnerable point. I do not need to know exactly how deep the puncture is to know that now is not the time to pay close to intrinsic value for it.

Management appears honest in how they are handling the crisis; they disclosed unflattering numbers nobody required them to disclose, and that counts. But honesty during a crisis does not erase the question of why the crisis happened in the first place.

I will wait for the independent audit. If it confirms this was one bad individual at one subsidiary, and the price is still at a wide enough discount, this could become a good opportunity. For now, paying close to fair value for a business in the middle of an unfinished scandal is not prudent investing; it is betting on an outcome I do not yet know.

The QuietVC lesson

Most of the money I see investors lose does not come from choosing the wrong business. It comes from answering the wrong question: asking "has the price fallen enough" when the right question is "is the thing that creates this business's value still intact". The two questions sound alike. They lead to completely different decisions.

Silent discipline. Compound value.