Silver's per-share problem: why production growth isn't shareholder value

Silver is the most reliable example of a sector where production growth is not shareholder value. The cohort has grown absolute output meaningfully over the past decade. The per-share metrics tell a different story. The arithmetic is mechanical; it is not the operator's bad luck; the incentives that produced it are by design.

We have written before about why per-share production growth is the fourth of the seven numbers a careful investor reads on any operator. In silver, the gap between the absolute and the per-share number is wide enough that the absolute number is positively misleading.

The arithmetic

Take a hypothetical silver producer that grows production from 10 million ounces to 15 million ounces over a decade. Absolute growth: 50%, or 4.1% CAGR. That is a respectable number on a press release.

Now apply the share-count growth that most silver producers have actually delivered over the same period. The cohort has grown share counts by between 60% and 200% over ten years, through a combination of equity issuance to fund acquisitions, equity issuance to fund growth capex, and significant option and warrant overhang from prior financings. Pick the middle of the range — 120% share-count growth — and per-share production has actually shrunk from 1.0 to 0.68 ounces per share. The operator's deck reports the headline 50% production growth. The shareholder, having held throughout, owns 32% less production per share than they started with.

That is not an extreme case. It is the cohort median.

Why it happens

The arithmetic is not the operator's fault on any single transaction. Each individual equity issuance is defensible: the acquisition added reserves; the growth project required capital; the operator chose equity over debt for balance-sheet reasons. The problem is the aggregation. Each defensible decision adds shares; the cumulative result is the cohort's per-share decline.

The reason it persists is structural. Senior silver compensation plans, almost without exception, are weighted toward absolute production targets and absolute reserve growth. The proxy filings are public. The result is incentive-aligned with absolute growth, not per-share growth. A CEO whose bonus is tied to growing absolute production has no reason to resist the equity-funded acquisition that grows it; the shareholder, holding a smaller per-share claim, is on the other side of the trade.

First Majestic (AG) is one of the cohort. The Mexican producer has grown absolute output meaningfully over the past decade through Gatos and other transactions; the per-share record over the same period is meaningfully weaker. The operator's deck features the absolute number. The proxy discloses the compensation structure. Both facts are public.

The compare to gold

Gold has the same potential incentive problem, but the senior gold cohort has produced a meaningfully cleaner per-share record than the senior silver cohort over the same window. The structural reason is asset quality: a gold operator that owns a Tier-1 asset can grow output through the asset itself rather than through equity-funded acquisitions, which keeps the share count discipline intact. A silver operator without a Tier-1 asset has to grow by acquisition, which is the path that produces dilution.

Agnico Eagle (AEM) is the cleanest senior gold example. The 2022 merger with Kirkland Lake was per-share accretive on day one — the rare gold deal that grew per-share reserves and per-share production simultaneously. The Detour Lake mill optimization since the merger has been an organic per-share story. Per-share reserves are a tracked metric in the compensation plan. The result is in the ten-year share-count line: relatively flat, despite an actively acquisitive period.

That posture is not available to most silver producers because the asset universe is different. The silver-primary cohort is small; the polymetallic mines that produce silver as a by-product (lead-zinc, copper-gold) sit inside the books of other operators (Freeport being one); the Tier-1 silver-primary deposits that could anchor a clean per-share growth story are few and contested.

Growth at the expense of the share count is a tax the operator quietly charges you.

How to read the per-share record

The Halvren check on any mining operator's per-share record is mechanical, and we run it for every name on the coverage list every year. Pull the operator's average share-count for the most recent year and for the year ten years prior. Pull production and reserves for both years. Compute the CAGR of per-share production and per-share reserves over the decade. Compare to the absolute CAGR for both metrics.

If the per-share CAGR is meaningfully positive and tracks the absolute, the operator has grown by improving the asset rather than buying assets with new shares. If the per-share CAGR is meaningfully negative while the absolute CAGR is positive, the operator has grown headline output by transferring value from the existing shareholder to the seller of the acquired asset.

The senior silver cohort, in aggregate, falls into the second bucket. There are exceptions. Pan American Silver's record post-Tahoe is more nuanced than the cohort average. Wheaton Precious Metals' streaming model is structurally per-share-accretive because the financing currency is a function of the deal price rather than the cohort dilution rate. Royal Gold and Franco-Nevada are similarly clean for the same reason: the business model is not a mining business in the per-share-dilution sense.

The "honest growth" path

A silver operator that wants to grow without diluting the existing shareholder has three options. The first is to fund growth through retained free cash flow, which requires the asset base to produce enough free cash to fund it — a high bar in silver, where the cohort cash-cost average sits in the high-teens-of-dollars-per-ounce range. The second is to fund growth through debt rather than equity, which requires the balance sheet to support it. The third is to buy back stock when the market price implies a lower per-ounce value than the operator's own asset base, which requires the operator to have a view of intrinsic value that exceeds the market's.

The third option is the rarest, and it is the cleanest signal. A silver operator that uses free cash flow to retire stock at trough prices, and accepts the optical drag of slow absolute growth, is making a per-share-positive bet that few in the cohort have been willing to make. We have not seen it executed at meaningful scale in the senior silver cohort over the past decade.

The streamer alternative

A reader who wants silver exposure without the per-share dilution problem has a structural workaround: the precious-metals streaming and royalty operators. Wheaton Precious Metals, Franco-Nevada, Royal Gold, and a handful of smaller streamers buy contractual rights to a fixed percentage of future production from a primary operator's mine, paid for upfront, and receive metal deliveries (or cash equivalent at the spot price) for the life of the asset. The streamer model is structurally per-share-accretive on every deal that closes at acceptable economics because the financing currency is contractual rather than the streamer's own equity.

The compounded effect on a streamer's per-share metrics over a decade is the opposite of the senior silver producer cohort. Wheaton Precious Metals has grown per-share attributable silver and gold production over ten years while keeping the share count growth in the single-digit range. The streamer multiple is higher than the producer multiple — for good reason — but the per-share total return record over the same window is, in many cases, meaningfully better.

The streamer model is not a perfect hedge against the silver-producer problem. The streamer takes counterparty risk on the underlying mine operator; if the operator fails, the streamer's deliveries are at risk. The streamer also has less optionality on a silver-price up-cycle than a debt-financed producer, because the upside is shared with the mine operator under the original contract terms. The streamer's beta to silver is lower; the streamer's per-share quality is higher; the choice depends on what the reader is underwriting.

For a Halvren-style portfolio that prioritizes per-share compounding over commodity beta, the streamer exposure is meaningfully more attractive than the senior silver producer exposure. We hold no senior silver producers on the desk currently. We track the streamer cohort more carefully than the producer cohort. Wheaton in particular has the cleanest per-share track record in the wider precious-metals universe; the streamer-royalty trade is the desk's preferred way to be paid for any precious-metals exposure that the macro thesis ever calls for.

The structural read on silver is therefore: the cohort produces a poor per-share record because the asset universe forces equity-funded acquisitions; the streamer model bypasses that problem by financing through contract rather than equity; and the silver investor who has been told for fifteen years that "production growth is the thesis" has been buying the wrong basket. The names are the same. The structural mechanics are different.

What we underwrite

Silver is a small position on the desk, and the reason is the per-share record. We hold First Majestic on the desk in part as a forcing function: writing about the per-share problem in a name that exemplifies it keeps the framework honest. The operator's writeup (see the desk page) records the open Pillar II questions clearly. The principal's view on the cohort has not changed in five years.

If the silver tape gives us a meaningful drawdown and a senior operator with a clean per-share record emerges from it, the position changes. Until then, the read is the same: production growth is not shareholder value.

Why the framework matters beyond silver

The per-share-versus-absolute framework applies far outside silver. Any operator that grows headline output through equity-funded acquisition is potentially destroying per-share value, regardless of the sector label. The framework is most visible in silver because the cohort dilution rate is unusually high; it is most useful elsewhere precisely because the dilution there is less obvious. A copper operator that has grown reserves through an equity-funded acquisition every two years is on the same arithmetic, even if the absolute production growth looks more defensible. The compounded per-share record over a decade reveals the same pattern.

The Halvren framework runs the per-share check on every operator on the coverage list, every year. Most pass; the operators on the desk pass cleanly. The cohort we do not own — including most senior silver producers, several gold-producer growth stories, and a small number of copper names whose acquisition pace has been unusually high — fails the check on the same arithmetic. The check is the same; the operators it eliminates differ by sector. The lesson is the same in every case: the operator's deck is the operator's story, and the share count is the part of the story that costs the existing shareholder real money to fund.

The math is in the proxy filings, every year. We read them.