NOVAGOLD insiders buy into a weak gold tape

NOVAGOLD RESOURCES INC. had four insiders buying on July 2, 2026, and that matters more because the tape around gold has been ugly than because the dollar amounts were large. Richard Williams, Peter Adamek, Gregory Anthony Lang, and Mark Benjamin Machlis all filed purchases that day. The stock closed at $6.44, up $0.37 or 6.10% from the prior close of $6.07, after trading between $6.25 and $6.64.

That is the right frame. NOVAGOLD is not a producer with a quarter of operating cash flow to lean on. It is a development-stage precious metals company whose primary asset is Donlin Gold in Alaska, a large-scale project the company says it holds 60% of. When gold is sliding, as it has been, the market stops paying for optionality with the same generosity. It starts asking whether the project can still clear the bar on capital intensity, permitting, and timing.

The sector tape is doing the opposite of helping

Gold has been under pressure in recent weeks. Spot prices fell from January 2026 peaks above $5,500 per ounce to levels near $4,000 in late June, which put the metal on track for its worst quarterly performance in 13 years, with roughly a 15% decline over three months and a 7.5% drop year to date. That is not a gentle reset. That is the kind of move that changes how investors underwrite developers, because the commodity price is the first line in every project model.

The macro backdrop has not offered much relief. Expectations that major central banks, including the Federal Reserve, may keep policy rates higher for longer have weighed on non-yielding assets like gold. Some banks have trimmed year-end 2026 gold forecasts into the $4,900 to $5,000 range, but that still leaves the market with a commodity that has already taken a hard hit from its peak. If you own a developer, you are not just buying geology. You are buying the right to wait for a better commodity tape without burning too much capital in the meantime.

That is why the peer set matters. Seabridge Gold, another development-stage company with a large North American gold asset, traded around $26 to $27 in early July. Agnico Eagle Mines, by contrast, has held up far better because it is a producer with ongoing output and shareholder returns, even while dealing with a temporary pit-wall issue at one operation. Those are different businesses, and the market treats them differently for good reason. NOVAGOLD sits much closer to Seabridge than to Agnico Eagle. It has a project story, not a production story.

Donlin Gold is the whole argument

NOVAGOLD’s most recent public update, filed June 24, emphasized progress on Donlin Gold engineering and permitting workstreams. That is the sort of update developers live on. It does not create revenue. It does not remove the financing question. It does, however, keep the asset moving up the value chain, which is the only way a project like this earns a higher multiple before first production.

The company also reported a second-quarter 2026 net loss of $25.5 million, or $0.06 per share, narrower than the year-earlier figure but still a reminder that project spending is the operating reality here. For a development-stage miner, losses are not the surprise. The question is whether the spending is buying progress that the market can eventually price. Donlin Gold is the answer NOVAGOLD keeps putting forward.

That is why the insider cluster is worth reading against the backdrop of the asset, not in isolation. When multiple senior officers buy into a name like this, they are not signaling near-term earnings momentum. They are signaling that, at least at this price and this point in the project cycle, they are willing to add exposure to the same long-duration bet the market is discounting. That can be meaningful. It can also be routine. The distinction is in the size, the clustering, and the context.

What the July 2 cluster actually says

InsiderTrades data shows a cluster here, with 2 distinct insiders and 12 recent declarations in the cluster picture. The July 2 filings include Richard Williams, Peter Adamek, Gregory Anthony Lang, and Mark Benjamin Machlis. Williams appears multiple times in the recent declarations list, which tells you this was not a one-off tick-the-box filing. It was repeated buying by the same name alongside other senior officers.

The euro-normalised filing values were small. Williams bought EUR 2,306.52 in one filing and EUR 888.33 in another, with another Williams filing at EUR 892.33. Adamek bought EUR 2,059.83. Lang bought EUR 4,415.69. Machlis bought EUR 1,147.09 and EUR 441.47. Those are not large sums in absolute terms, and they are microscopic relative to NOVAGOLD’s EUR 3.72 billion market cap. InsiderTrades data pegs one of the conviction proxies at about 0.00% of market value, which is exactly the point. This is not a balance-sheet-changing gesture. It is a confidence signal.

The signal score sits at 48 in the legacy framework, and the rationale is straightforward enough: an operating director filed, the trades came as part of an insider cluster, and the amounts were small in market-cap terms. That is a fair read. It is also a reminder not to over-romanticize the print. Small buys from multiple officers can matter, but they do not magically turn a development story into a de-risked one.

The cohort math is decent, not dramatic

NOVAGOLD RESOURCES INC. insider-trading story">

InsiderTrades data for the relevant bucket, Directeur · Large, gives you a useful historical anchor. The sample size is 58,770. The 90-day win rate is 49.7%. The average 90-day return is 1.4%. The average 365-day return is 20.53%. That is the kind of profile that tells you insider buying in this bucket has had some positive drift over time, but not enough to pretend the signal is a shortcut.

Read that carefully. A 49.7% win rate is basically a coin flip. A 1.4% average 90-day return is modest. The 20.53% 365-day average is more interesting, but it is still historical cohort data, not a forecast for NOVAGOLD and not a promise that this cluster will work. If you are weighing this name, the right use of that cohort is to keep you from dismissing the filing as noise while also keeping you from treating it like a catalyst.

The strategy layer in the dossier is similarly restrained. InsiderTrades data shows a 90-day holding period, a maximum position size of 0.08%, an out-of-sample Sharpe of 0.53, and an out-of-sample CAGR of 17.1% on a restricted EU venue universe. Those figures survive only in a narrow universe and a short, single-regime window, and they do not survive search-aware deflation. So yes, there is some historical edge in the broader framework. No, that does not mean this trade is suddenly a high-conviction alpha machine.

Why the buying matters more here than it would at a producer

A producer can buy back stock, raise dividends, and point to ounces in the ground that are already being sold. A developer cannot. It has to persuade the market that the project is real, financeable, and worth waiting for. That makes insider buying more interesting in a name like NOVAGOLD than it would be in a mature miner with a long production record. The insiders are buying the same future the market is debating, and they are doing it while the commodity tape is weak.

That does not mean they are calling the bottom in gold. It means they are willing to own the project through the drawdown. There is a difference. Gold’s recent slide from above $5,500 to near $4,000 has made that willingness more notable, because developers are exactly where sentiment gets hit first when the metal rolls over. The market starts discounting dilution, delay, and capex inflation before it worries about the geology.

NOVAGOLD’s setup is therefore a clean test of whether insider buying can still matter in a hard tape. The answer is yes, but only at the margin. The cluster does not erase the macro headwind. It does not solve the financing question. It does not turn Donlin Gold into a cash generator. What it does is tell you that the people closest to the asset were willing to add stock on July 2 while the sector was still digesting a rough quarter.

The read breaks down where the project risk starts

The obvious risk is that a development-stage gold company can look cheap for a long time and still go nowhere if the project timeline slips or the commodity backdrop stays hostile. NOVAGOLD has no operating revenue to cushion the story. It reported a second-quarter loss, and the June 24 update was about engineering and permitting workstreams, not production. That is the reality of the name.

There is also a valuation trap in names like this. A large market cap can make a small insider purchase look almost ceremonial, and in this case the euro-normalised amounts are indeed tiny relative to the company. If you are looking for a clean, high-conviction insider tell, this is not it. The buys are better read as a cluster of internal alignment than as a bold capital allocation statement.

Still, the cluster is not meaningless. Multiple senior officers buying on the same date, in a weak gold tape, after a quarter that still showed a loss and ongoing project spending, is the kind of filing that deserves a second look. It is especially relevant because the company’s value proposition is so concentrated in one asset. When the whole story sits on Donlin Gold, insider behavior around that story matters more than usual.

What to watch next

The next thing to watch is whether the company keeps advancing Donlin Gold without forcing the market to absorb a worse financing overhang. That is the pressure point for every developer. Progress on engineering and permitting helps. So does a better gold tape. Neither is guaranteed. If gold stabilizes, the market can get more forgiving on long-dated projects. If it keeps sliding, the burden on NOVAGOLD gets heavier.

You should also watch whether the July 2 cluster proves to be a one-day event or the start of a broader pattern. Insider buying is most useful when it repeats, when it comes from multiple names, and when it shows up against a difficult backdrop rather than after the stock has already run. This one checks those boxes. It does not clear the bar for certainty. That is fine. Certainty is not the point.

For now, the read is simple. NOVAGOLD insiders bought into a weak gold market, and they did it in a name where the project story still dominates everything else. That makes the filing worth your time. It does not make it a thesis by itself.

This is not investment advice.