PNJ after five limit-down sessions: three questions before I look at the price

A stock down 73% from its high earlier this year.

Five limit-down sessions in a row. On the fifth, 76 million shares changed hands.

The sellers believe they just escaped. The buyers believe they just caught the bottom.

I do not know who is right. I only know I should not look at the price yet.

I spent a day rereading PNJ's disclosures over the past month. This piece records what I found, and the three questions I ask myself before looking at the price: one about people, one about cash, one about the crowd.

The figures come from company disclosures as reported in the press, through the close of 2 October 2026. This is how I see it, not advice to buy or sell.

What happened

On 2 July the police charged a diamond smuggling ring. Among the accused was the former director of a certification company wholly owned by PNJ.

Customers flooded in to sell diamonds back. PNJ honoured its buy-back promise and paid immediately. In the first 20 days the company spent 5,900 billion dong repurchasing goods, while sales brought in only about 1,590 billion.

August brought good news. Police confirmed the smuggled diamonds had not entered PNJ's system. An independent testing body concluded 100% of samples were natural diamonds. The stock recovered.

Then came September:

• 3 September: the reviewed half-year report showed six-month profit of 728 billion, 456 billion below the company's own version. The auditor added an emphasis of matter on the diamond incident.
• 4 September: the board approved the company borrowing money from its own chairwoman. Her two daughters registered to sell 25 million shares to fund that loan.
• 11–18 September: the 25 million shares were sold at around 37,000, raising roughly 900 billion. About 700 billion was lent to the company.
• 25 September: the company presented a new 2026 plan: a 7,071 billion provision, a 6,271 billion loss. The original plan was a 3,409 billion profit. The 2026 dividend goes to zero. The company needs to raise up to 8,000 billion. The same day, the chairwoman's brother registered to sell 9 million shares.
• 29 September: the company sought approval for a private placement of up to 550 million shares. There are 511.7 million outstanding.

From 28 September to 2 October the price went from around 33,000 to 23,050. An extraordinary general meeting is set for 21 October.

The number that is not in the P/E

For years PNJ sold diamonds with a promise attached: if a customer wants to sell back, the company buys at 80–95% of the price.

That promise helped sales. It is also a liability, just not yet due. Customers can bring goods back at any time, and they will do it exactly when the company is weakest.

According to notes in the reviewed report as reported in the press, the total value of goods covered by buy-back policies is around 26,053 billion dong. Equity at the end of June was 13,456 billion. The promise is nearly twice equity.

On the goods repurchased through early September, for every 100 dong paid out the company lost about 29.

Before July, PNJ's P/E, P/B and ROE showed none of this. Nor did I see it.

Question one: how are the insiders behaving?

Buying jewellery is buying trust in the seller. Few of us can tell a good diamond from an average one.

Buying a stock is the same. I am not in their boardroom. I can only watch what they do when things get hard.

So I split the question in two.

Half one: do customers still trust them? I run a business too, and I understand why they chose to keep their word. Abandoning the promise then could have cost them the gold and jewellery business entirely. The stones were also proven genuine. But the result is unclear: in July and August revenue still rose 4.5%, while retail fell 1.5% and gross margin dropped from 18.5% to 10%. Two months is not enough to conclude anything.

Half two: do the managers behave like owners? I want the operators to treat the company as their family's largest asset, especially when it is hard.

Here there are two readings of the same event.

The company's reading: the transactions were registered and disclosed lawfully. The sale proceeds went back into the company exactly when it needed them. The loan required no collateral. The chairwoman herself did not sell her own stake.

The critical reading: shareholders bear the loss first; creditors are paid first and with interest. The chairwoman's family converted 25 million shares into cash, then into a loan, one week before the company announced a record loss. One of the two sellers is vice-chair and sits on the audit committee.

I see no evidence of wrongdoing. But this is where I stop the longest.

Question two: who is holding the cash?

In a panic, whoever has cash sets the terms. Whoever needs money listens.

So I ask two more things.

Who needs money, and how urgently? At the end of June the company had about 4,000 billion in cash and short-term investments, against nearly 4,000 billion of bank debt. In the first 20 days of July alone it spent 5,900 billion on repurchases. Everything else follows: a 120-day payment schedule, the loan from the chairwoman's family, the 8,000 billion plan.

Which row am I sitting in? When a company struggles, creditors go first and shareholders last. The interest rate on the chairwoman's loan has not been disclosed; the press estimates around 9–10% a year. A ten-year government bond pays about 5%. The lender receives 9–10% and is paid first. The shareholder receives whatever is left, if anything.

I do not blame anyone for leaving the last row for a better one. I simply note they changed seats, and ask myself where I am sitting.

Then I look at the placement. Whoever brings 8,000 billion in cash right now writes the terms. Existing shareholders without new money get diluted.

Under the proposal, the issue price may not be below the 30-session average or the most recent book value. Book value is currently around 26,300 a share, above the market price. But after recording the full-year loss, I estimate it falls to about 12,600. For 550 million shares to raise 8,000 billion, the price only needs to be 14,545.

So comparing today's price against 84,600 earlier this year is the wrong comparison. The company then had 511.7 million shares and a 3,409 billion profit plan. The company ahead may have more than a billion shares.

I see no reason to hurry before knowing who buys, and at what price.

Question three: has the crowd left?

I like buying when others are afraid. A stock limit-down five sessions in a row gets my attention.

But fear has levels, and I want to know which one I am looking at.

Has pessimism bottomed? As I understand it, when pessimism bottoms, nobody wants to buy. On 2 October buying pressure at one point pushed PNJ above the reference price, and 76 million shares traded. Plenty of people are still eager. And the bad news calendar is not finished: 9 million shares registered for sale through 30 October, the meeting on 21 October, the Q3 report, the investigation's conclusion.

Am I buying one stock or a basket? Buying while everyone panics usually works with a basket. A few go to zero and the rest make up for it. Putting money into a single stock in crisis is a different bet entirely.

Cheap against what? Cheap against the old price says nothing. I need cheap against value. To get value per share you need to know how many shares there are. That number is somewhere between 512 million and 1.06 billion.

Pessimism is a necessary condition. It has never been a sufficient one.

I am not crossing PNJ off. I am writing it in the notebook and waiting until fewer people are talking about it.

The QuietVC lesson

A promise to customers is also a liability. Buy-back commitments, guaranteed returns, price guarantees: they do not appear in the P/E. They live in the notes to the accounts, and they fall due on the worst possible day.

A falling price does not create a margin of safety. A margin of safety comes from being able to count what you are buying. Until I can count the obligations and know the share count, down 73% still does not tell me it is cheap.

In a crisis, watch which row the insiders sit in. Anyone can say anything. The seat they choose is the real statement.

"Too hard" is a valid answer. For me, PNJ sits in that drawer right now. A 38-year-old brand may well come through this; the stones were proven real. But I cannot count it, so I watch.

Four things I will track: the price and the buyers of the placement; whether the chairwoman's family contributes capital alongside shareholders; how much stock is still coming back in the Q3 report; and the investigation's conclusion.

Three questions: about people, about cash, about the crowd. None of them asks "how far has the price fallen".

Silent discipline. Compound value.