Abstract: Caledonia’s Q2 profit rose, but Blanket production fell 17.6% year-on-year and AISC increased to $2,678/oz; high gold prices are buying time for operational volatility and Bilboes construction, rather than eliminating cost issues.

—— ◆ ——

◌ ◎ I. Production Declined, Costs Rose, Yet Profits Increased: Why?

On August 10, Caledonia Mining disclosed its Q2 2026 results. The company’s consolidated gold production was 17,737 ounces, a 17.3% decrease year-on-year; core asset Blanket Mine produced 17,360 ounces, a 17.6% decrease year-on-year. Meanwhile, mine costs rose from $1,123/oz to $1,675/oz, an increase of 49.2%; All-In Sustaining Costs (AISC) rose from $1,805/oz to $2,678/oz, an increase of 48.3%. By traditional mining analysis frameworks, this should have been a clearly pressured earnings report.

However, the company’s Q2 gross profit increased from $33.81 million to $39.18 million, net profit attributable to shareholders increased from $20.49 million to $23.81 million, and operating cash flow also saw a slight increase from $28.08 million to $28.44 million. There is only one core variable explaining this divergence: the company realized gold price increases from $3,186/oz to $4,259/oz, a year-on-year rise of 33.7%.

If we use a simplified observation of realized price minus AISC, the difference per ounce in Q2 2025 was approximately $1,381, while in Q2 2026 it was approximately $1,581. Although costs increased by $873/oz and gold prices increased by $1,073/oz, there remained an additional buffer of about $200. This calculation cannot replace the income statement, as sales timing, royalties, corporate expenses, taxes, and hedging all affect final profits; however, it clearly illustrates how higher revenues temporarily enveloped operational setbacks via gold prices.

Therefore, what this performance truly needs to answer is not “whether gold companies benefit from high gold prices,” but rather: Can the cash buffer created by high gold prices repair Blanket’s grade, recovery rates, and unit costs before price fluctuations, while simultaneously supporting the Bilboes project into peak construction? If yes, the 2026 cost increase is transitional volatility; if no, current profits are more like time bought for the mine by gold prices.

There is also a denominator misconception here. Gold mining companies often compare using per-ounce metrics, but declining production simultaneously raises costs, reduces sellable ounces, and weakens the dilution effect on sustaining capital expenditures. Even if per-ounce profits remain high, total cash may not grow synchronously due to reduced sales volume. Caledonia’s Q2 operating cash flow being almost flat compared to the same period last year serves as a warning: significant increases in unit price advantages did not translate one-to-one into more operating cash.

Taxes and royalties also increase with rising gold prices, meaning some costs have revenue-linked attributes. High gold prices do not therefore fall entirely into miners’ pockets; they are redistributed among governments, employees, contractors, financiers, and shareholders. For resource-rich countries, this is an opportunity to expand fiscal and foreign exchange revenues; for mines, the key is retaining sufficient cash to complete underground development and new projects. Comparing only gold prices and cash costs easily overlooks growth capital and national distribution mechanisms.

The company continues to pay dividends and must maintain its listed company financing capability. If Blanket costs remain persistently high, management may need to make stricter prioritizations between dividends, sustaining capital, and Bilboes investments. Delaying any of these items could have long-term consequences: reducing development affects future stopes, reducing Bilboes investment delays scale transformation, and over-reliance on debt increases fixed interest and refinancing risks.

◌ ◎ II. What Happened at Blanket: Increased Tonnage Does Not Automatically Mean More Gold

The company attributed the decline in Blanket’s Q2 production to lower feed grades and recovery rates, noting that increased processing tonnage could only partially offset this. Gold mine production is determined jointly by throughput, grade, and recovery rate. Increasing ore tonnage helps dilute certain fixed costs, but if the gold content per ton of ore declines and the mill recovery percentage also drops, final gold ounces may still decrease, while electricity, hoisting, grinding, reagents, and tailings treatment costs continue to accrue.

This is precisely the mechanism behind the rapid rise in unit costs. Costs do not simply rise because “more money is spent”; they can also rise because the same fixed and semi-fixed expenditures are spread over fewer gold ounces. Blanket’s Q2 production decreased by 3,710 ounces year-on-year, a significant drop; even if total operating expenditures changed little, the shrinking denominator pushes up per-ounce costs. When superimposed with deepening mining, equipment maintenance, wages, electricity, and spare parts, AISC rises faster than cash costs.

Quarter-over-quarter data provides another perspective. Blanket’s Q1 production was 14,767 ounces, rising to 17,360 ounces in Q2, a quarter-on-quarter increase of 17.6%. This indicates the mine is not deteriorating continuously; production rhythms have recovered from Q1 lows. Consequently, the company maintained its full-year production guidance of 72,000–76,500 ounces, but raised its full-year mine cost guidance to $1,600–$1,800/oz and its AISC guidance to $2,500–$2,700/oz.

Maintaining production guidance while raising cost guidance conveys a restrained signal: management believes ore quantity and production still have room for recovery, yet acknowledges that recovery requires higher unit costs. For industry analysts, the most critical factor in Q3 is not just whether production continues to grow quarter-on-quarter, but whether at least two of the following improve synchronously: grade, recovery rate, and cost per ton of processed material. Otherwise, swapping tonnage for production may continue to elevate grinding and energy burdens.

◌ ◎ III. The External Environment of High Gold Prices: Strong Price Protection, But Demand Structure Is Not Unilaterally Upward

World Gold Council (WGC) data shows that the LBMA noon gold price average in Q2 2026 was $4,506.29/oz, down 8% from the historical high in Q1, but still 37% higher than Q2 2025. Global total gold demand (including OTC) was approximately 1,269 tonnes in Q2, flat year-on-year; H1 demand reached 2,522 tonnes, up 2% year-on-year, with demand value hitting a historical high of approximately $380 billion.

Demand internally is inconsistent. In Q2, gold ETFs saw net outflows of approximately 45 tonnes; bullion and coin demand was 307 tonnes, roughly flat year-on-year; central bank gold purchases reached 289 tonnes, a significant surge compared to the previous year; high gold prices suppressed jewelry consumption, which dropped to 278 tonnes in Q2, a 17% year-on-year decline. This indicates that high gold prices do not mean simultaneous expansion across all end-user demands, but rather that central banks, OTC investments, and macro-hedging forces compensated for part of the traditional consumption and ETF volatility.

For mines, high prices first improve revenues, but do not automatically lower mining costs. In fact, high gold prices may bring wage negotiations, contractor quotes, tax rates, and resource-country distribution requirements upward. Caledonia’s realized price of $4,259/oz was lower than the WGC’s Q2 LBMA average, potentially influenced by sales timing, settlement basis, and hedging portfolios; the spot average price cannot be directly substituted for company revenue.

The company also arranged price protection for the Bilboes financing: purchasing put options with a minimum strike price of $3,500/oz on a monthly volume of 3,000 oz from January 2026 to December 2028. The purpose is not to cap upside potential, but to establish a cash flow floor for Blanket during the peak capital expenditure phase of Bilboes, thereby enhancing lenders’ visibility into debt service capacity. In other words, the company has acknowledged that Bilboes’ construction cannot rely solely on the assumption that gold prices will remain at current elevated levels over the next three years.

◌ ◎ IV. Data Analysis: Gold Prices Provided an Extra $1,073, While Costs Absorbed $873

Operating MetricsQ2 2026Q2 2025Change and Implication
Consolidated Gold Production17,737 oz21,442 ozDown 17.3%, pressuring sales volume
Blanket Production17,360 oz21,070 ozDown 17.6%, primarily driven by grade and recovery rates
Gold Sales Volume17,811 oz20,487 ozDown 13.1%, consistent with production trends
Realized Gold Price$4,259/oz$3,186/ozUp $1,073, a 33.7% increase
Mine Cost$1,675/oz$1,123/ozUp $552, a 49.2% increase
AISC$2,678/oz$1,805/ozUp $873, a 48.3% increase
Realized Price Less AISC$1,581/oz$1,381/ozSimplified margin increased by approx. $200; higher prices still cover deteriorating costs
Gross Profit$39.18 million$33.81 millionIncreased by approx. 15.9%, primarily supported by price
Operating Cash Flow$28.44 million$28.08 millionEssentially flat, indicating that more profit has not yet fully translated into stronger cash increments
Long-Term and Financing MetricsDisclosed DataComparison BasisQuestions to Answer
2026 Production Guidance72,000–76,500 ozMaintainedCan grade and recovery rates in H2 support the target?
2026 AISC Guidance$2,500–$2,700/ozRaisedWill current high costs become the norm for the full year?
2026 Capital Expenditure$103.3 millionBlanket $51.5M, Motapa $3.8M, Bilboes $48MOperating cash flow must serve both maintenance and growth capital simultaneously
Bilboes Peak Funding Requirement$484 millionFeasibility Study BaselineSignificantly exceeds single-quarter operating cash flow, requiring layered financing
Convertible Bond FinancingGross $150M, Net approx. $130MCoupon 5.875%, includes capped call arrangementPartially resolved, but does not constitute complete project financing
Bilboes First Full Year ProductionApprox. 200,000 ozPlanned for 2029Construction, BIOX process, talent acquisition, and power supply must align simultaneously
WGC Q2 Average Price$4,506.29/ozYoY +37%, MoM -8%High prices remain strong, but volatility has emerged

The key insight from the data lies not in “profits are still growing,” but in the structure of those profits. For every ounce of gold sold, the price provided an additional $1,073 year-on-year, while AISC absorbed an additional $873, leaving a net incremental buffer of only about $200. If gold prices were to drop by 10%, while AISC remained at current levels, unit profitability would contract significantly. Conversely, if Blanket restores its grade and recovery rates, even without further gold price increases, declining unit costs could generate new cash increments.

Full-year guidance can also be used to test production pressures. Assuming the lower bound of the 72,000 oz guidance, H2 output needs to accelerate significantly compared to H1. Mines cannot rely on a single good month to make up the shortfall; instead, they must maintain stability in mining faces, hoisting, grinding, and recovery across multiple consecutive months. For underground mines, grade fluctuations may stem from stope sequencing, ore dilution, and development preparation. Even if production increases in Q3, it is crucial to observe whether this comes at the expense of over-mining high-grade zones. Sustainable recovery should manifest simultaneously as increased production, declining unit costs, and steady progress in mine development advance footage.

The fact that operating cash flow remained essentially flat while profits rose also warrants separate explanation. Working capital dynamics, inventory changes, receivables settlement, and tax payments create timing differences between profit and cash; prepayments for long-cycle Bilboes equipment will also cause free cash flow to lag behind operating cash flow. With the company’s 2026 capital expenditure plan totaling $103.3 million, sustained operating cash inflows over multiple quarters are required. When analyzing whether “high gold prices can self-fund construction,” one cannot simply multiply single-quarter gross profit by four; deductions must be made for Blanket’s own maintenance capital, taxes, interest, dividends, and changes in working capital.

Hedging protections also have boundaries. The monthly 3,000 oz put option at $3,500 covers only a portion of Blanket’s sales, not the entire group’s future production; it provides minimum price insurance but does not cover shortfalls in production, AISC overruns, power outages, or project delays. Lenders value hedging because it mitigates tail-price risk; however, industry practitioners must still match hedge volumes against actual monthly sales volumes and align tenors with Bilboes’ expenditure peaks to determine whether the protection is sufficient.

◌ ◎ V. Bilboes Amplifies the Challenge: One Mine’s Cash Flow Must Fund Two Mines’ Futures

Caledonia is transitioning from a single-asset producer centered on Blanket to a multi-asset gold company with Bilboes as its growth platform. The Bilboes feasibility study indicates that the project plans to develop an open-pit mine and BIOX processing system with an expected mine life of approximately 10.8 years, targeting around 200,000 oz in its first full year of production—significantly higher than Blanket’s current annual scale of roughly 70,000+ oz. If commissioned as planned, the company’s production mix will be reshaped.

However, the larger the growth scale, the stronger the dependency on Blanket’s cash flow during the transition period. Total capital expenditure for 2026 is estimated at $103.3 million, of which Bilboes accounts for approximately $48 million, while Blanket still requires $51.5 million. Blanket must fund both its own deep-level mining, equipment, and development activities, as well as provide cash for Bilboes’ early procurement and engineering. The stability of $28.44 million in Q2 operating cash flow is a positive signal, yet it is insufficient to cover the entire project construction cycle.

Bilboes’ peak funding requirement is approximately $484 million. The company has completed the issuance of $150 million in convertible bonds, netting approximately $130 million after deducting fees and capped call structure costs; it also plans mid-term financing instruments and project finance. Equating the $150 million financing announcement directly with “Bilboes’ funding needs are resolved” ignores at least three gaps: the net financing amount is lower than the nominal amount; project funds are required in tranches according to construction progress; and lenders will conduct independent reviews regarding reserves, feasibility studies, hedging, construction contracts, country risk, and cash flow coverage.

BIOX represents another long-cycle variable for Bilboes. Refractory gold ores require bio-oxidation pretreatment to improve gold recovery. While this technology has commercial precedents globally, each ore body requires site-specific adaptation regarding mineralogy, bacterial cultivation, temperature, oxygen levels, nutrients, and continuous operational control. If any single link in the mine, plant, BIOX circuit, tailings, or power system is delayed, the first gold pour by late 2028 and full-year production in 2029 may be impacted.

◌ ◎ Six, Thinking One Step Further: What Is Overlooked Are Not Gold Prices, But Operating Conditions Composed of Power, Foreign Exchange, and Talent

The implicit constraints on Zimbabwean gold mining projects often lie outside the resource model. Power reliability, foreign exchange settlement, and technical talent determine whether book reserves can be stably converted into US dollar cash flows. Caledonia has consistently emphasized reliable and affordable power, as well as foreign exchange arrangements for gold sales proceeds, during the acquisition and development of Bilboes; these conditions are not ancillary clauses but integral components of project finance viability.

Zimbabwe’s Reserve Bank foreign exchange guidelines issued in March 2026 stipulate that payments for large-scale gold miners surrendering gold will continue to be executed at 70% foreign currency and 30% local currency. For miners, wages and a portion of local procurement can be paid in local currency, but significant amounts of equipment, spare parts, reagents, and debt servicing require foreign currency. The degree of match between revenue currencies and cost currencies affects actual available cash. If the local currency depreciates or exchange is restricted, US dollar income on the income statement does not necessarily equate to US dollars freely usable for importing equipment and servicing debt.

Power similarly permeates the entire cost chain. Power outages not only reduce throughput tonnage but also interrupt underground hoisting, ventilation, dewatering, and plant operations, lowering recovery rates and increasing diesel self-generation costs. Blanket already has solar facilities to buffer some risks, but Bilboes is larger in scale and BIOX requires continuous, stable operation, demanding higher standards for power quality and backup systems.

Talent is an even slower variable. Advancing Bilboes from open-pit mining, grinding, and flotation through to BIOX and final gold recovery requires teams in construction management, metallurgy, bio-oxidation, maintenance, and safety. Equipment can arrive within months via procurement contracts, but cultivating a mature operating team may span multiple production cycles. High gold prices can pay higher wages, but they cannot instantly create experience.

The base case assumes Blanket continues its sequential improvement in the second half of the year, achieving annual production at the low end of guidance, with AISC maintained at $2,500–$2,700/oz; Bilboes procures long-lead equipment as planned and gradually secures mid-term and project financing. Strengthening conditions include significant improvements in Blanket’s grade and recovery rates, sustained high gold prices, stable power, and declining financing costs. Failure conditions include AISC breaking above the upper end of guidance, operating cash flow contracting due to declining production, worsening foreign exchange availability, delays in project financing, or BIOX critical engineering and staffing falling behind schedule.

Next, observers should monitor: Blanket’s grade and recovery rates in Q3; whether quarterly production is sufficient to meet the annual lower bound; whether AISC returns to the center of guidance; the alignment between monthly 3,000 oz hedging protection and actual sales; whether Bilboes’ long-lead equipment procurement starts on schedule; whether mid-term and project financing provide specific amounts, tenors, and conditions precedent; and whether power and foreign exchange arrangements receive executable confirmation.

These milestones must be correlated with the timeline. If key long-lead equipment procurement does not start by Q3 2026, the buffer for the first gold pour by late 2028 will shrink; if project financing continues to delay following independent technical review, the company may need to utilize more parent company funds to maintain engineering momentum; if Blanket’s production recovers yet AISC remains close to $2,700/oz, it indicates that cost issues are not solely a function of the low-production denominator, but may also reflect structural changes in mining depth, maintenance, and input prices.

Industry personnel should also observe whether the company’s disclosures further refine “total funding requirements” into annual and quarterly funding curves. A peak of $484 million does not imply all funds are needed on a single day, but each stage may carry loan conditions precedent and shareholder contribution ratios. The clearer the funding curve, the better one can assess whether Blanket’s cash flow, convertible bonds, mid-term financing, and project loans are truly interconnected.

Ultimately, whether Bilboes comes online on schedule depends on multiple consecutive conditions, not just a high-gold-price condition. High gold prices provide a window for repair and financing; management must lock in Blanket operations, engineering procurement, lender reviews, foreign exchange, and power sequentially within that window. The existence of a window does not mean the work is done.

—— ◆ ——

High gold prices can buy time for a mine with rising costs, but only grade recovery, stable power, and complete financing can transform that time into the next mine.

— END —