Serabi Gold plc (SBI.TO / SRB.L) - Deep Dive Research Report

Prepared 2026-06-15. All figures in USD unless noted. Serabi reports operations in USD, is incorporated in the UK, and pays dividends in GBP/CAD.

A note on "concalls": Serabi is an AIM/TSX micro-cap that does not hold traditional quarterly earnings calls with Q&A transcripts. Its equivalent disclosure cadence is a detailed quarterly production-and-operations RNS plus periodic interim/audited financial results, each carrying signed management commentary. The five reporting periods used throughout this report are:

1. Q1 2026 - production RNS 14 Apr 2026; unaudited interim results 29 May 2026
2. FY/Q4 2025 - audited full-year results 1 May 2026
3. Q3 2025 - production RNS 14 Oct 2025; interim results 28 Nov 2025
4. Q2 2025 - production RNS 14 Jul 2025
5. Q1 2025 - production RNS 14 Apr 2025 (with board change)

The most recent period (Q1 2026, reported within the last ~60 days) is included as required.

1. What the company does

Serabi Gold digs high-grade gold out of narrow underground veins in one of the most remote corners of the Amazon and turns it into doré bars. It is a small but genuinely profitable producer - not an explorer burning cash, not a major. It runs two hard-rock underground operations in the Tapajós region of Pará State in northern Brazil, and in 2025 it produced 44,169 ounces of gold, a number it is now pushing toward 55,000-plus.

The plain version: the company finds quartz-sulphide veins that carry gold at grades (roughly 6-8 grams per tonne) far richer than the disseminated, low-grade ore that big open-pit miners process. Because the gold is concentrated in narrow structures, Serabi mines underground with small trackless equipment, brings the ore to a central processing plant, crushes it, and recovers the gold using gravity separation and cyanidation. The output is gold doré, which is refined and sold at the spot gold price. There is no marketing, no customer relationship, no pricing power and no product differentiation: an ounce of Serabi gold is identical to an ounce of anyone else's gold. The entire business is a contest of geology, operating cost, and capital discipline.

The Tapajós context matters. Serabi operates inside the Tapajós gold province, an area that has historically produced more than 30 million ounces, almost all of it by informal artisanal miners (garimpeiros) panning surface and alluvial gold over decades. Very little of that province has ever been mined by a formal, mechanized, permitted underground operation. Serabi is one of the few companies that has built the permits, the camp, the plant, and the social licence to mine it industrially. That is the company's real edge: not a patent or a brand, but a hard-won operating footprint in a jurisdiction where logistics and permitting defeat most newcomers.

The company has two production hubs:

• The Palito Complex - the mature heart of the business, comprising the Palito and São Chico underground mines plus the only fully built processing plant Serabi owns (around 650 tonnes per day). This complex has produced steadily for over a decade.
• The Coringa Gold Project - a newer underground mine 200 km south that poured first gold in July 2022 and is the growth engine. Coringa has no plant of its own. Its ore is crushed and run through an ore sorter on site, then the upgraded material is trucked north to the Palito plant. This "satellite mine, central mill" design is the single most important structural decision in the company and the reason Coringa was capital-light to build.

The founding arc explains the present. Serabi has been listed for years (AIM since the mid-2000s, with a TSX listing under SBI giving it a Canadian audience for mining capital). For most of its life it was a sub-40,000-ounce, balance-sheet-constrained Palito-only producer that periodically had to raise money. The pivotal decision was the December 2017 acquisition of Coringa for $22 million, and then the patient, years-long grind of permitting and building it during a period that included a global pandemic and a hostile permitting environment. The second pivotal decision was choosing not to build a second plant at Coringa - instead installing an ore sorter and classification plant to shrink the trucked tonnage by roughly half and feed the existing Palito mill. That kept capital low and turned Coringa into a high-return bolt-on rather than a balance-sheet-breaking new build.

The result, by 2025-2026, is a company that has crossed a threshold most juniors never reach: it self-funds its own growth, carries no debt, and pays a dividend. CEO Mike Hodgson framed the rhythm bluntly after Q1 2025:

"We've produced 10,000 ounces for Q1, which followed the 10,000 ounces produced in Q4. So we're certainly maintaining that rhythm." (Q1 2025 RNS, 14 Apr 2025)

By Q3 2025 the rhythm had stepped up, and by Q1 2026 the company was generating $21 million of post-tax profit in a single quarter against a $64 million cash pile, on a gold price near $5,000 an ounce. The thesis is operational leverage: a fixed cost base, rising grade, rising volume, and a soaring gold price compounding on top of each other.

2. Business segments

Serabi does not report financial segments in the conventional sense; it consolidates one product (gold doré) sold at one price. But operationally it runs two distinct mining hubs with different histories, geologies, and roles in the group. They are best understood as two segments.

Segment A - The Palito Complex (the cash cow and the mill)

Palito is the original business and the only place Serabi can actually process ore. It comprises two underground mines, Palito and São Chico, feeding a central plant rated around 650 tonnes per day that recovers gold by gravity concentration plus conventional cyanidation. For more than a decade this complex did 30,000-40,000 ounces a year on its own, and through 2024-2025 its plant-feed grades improved markedly (Palito plant-feed grade up roughly 27% versus the 2024 average, per the Q2 2025 RNS).

Its core capability is the processing plant itself - the permitted, built, paid-for mill with cyanidation circuits, a tailings facility, and the environmental licences that go with them. In the Tapajós, that plant is the scarcest asset in the company. Anyone can find a vein; almost nobody can permit and build a cyanidation plant in the Amazon. The plant is what makes Coringa viable and is the reason Serabi's growth capital is so low.

Strategically, Palito is both the cash cow and the bottleneck. It generates reliable ounces, but its 650 tpd capacity caps how much ore the whole company can process. That cap is exactly why management is installing a fourth ball mill - to lift throughput to roughly 900 tpd (about 330,000 tonnes per annum by 2027) and let both mines run harder. Palito is mature, predictable, and the foundation everything else sits on.

Segment B - The Coringa Gold Project (the growth engine)

Coringa sits 200 km south of Palito and about 70 km southeast of Novo Progresso. It is 100%-owned, bought for $22 million in 2017, and poured first gold in July 2022. Geologically it is a set of northwest-trending quartz-sulphide veins (Meio, Serra, Galena and newer discoveries like Serra South), with measured-and-indicated resources around 179,000 ounces at roughly 7 g/t and inferred around 271,000 ounces at roughly 5.8 g/t as of the April 2024 estimate, before the 2025 drilling that materially expanded the picture.

Its core capability is the ore sorter and classification plant. Rather than build a $100m+ second mill, Serabi crushes Coringa ore, runs it through an X-ray/sensor-based ore sorter that strips out roughly 45-50% of the barren rock, and trucks only the upgraded fraction the 200 km to Palito. Hodgson has repeatedly called the ore sorter "a game changer." It cuts trucking costs and tonnage, raises the grade arriving at the mill, and let the company sequence Coringa as a low-capital, high-return ramp-up. Through 2025 Coringa progressed from one producing zone (Meio) into commercial production, with grades stepping up sharply (the Q1 2026 RNS cited a +39% grade improvement at Coringa as the Meio zone hit commercial production and the classification plant did its work).

Why it exists separately: different deposit, different location, acquisition history, its own permits, and its own ramp curve. Competitively, Coringa is where the resource growth is - the 2025 exploration program (38,400 metres drilled) found the Serra South zone and extended the Meio trend, feeding the company's stated ambition to grow consolidated resources toward 1.5-2.0 million ounces. Coringa is the strategic growth bet: it is what takes the company from ~44,000 ounces to 55,000-plus and, on management's longer vision, toward a 100,000-ounce combined operation.

3. Products and business detail

The product is gold doré - semi-pure gold-silver bars poured on site - sold into the global bullion market at the prevailing spot gold price. There is one product. There is no catalogue, no branding, no specification a customer cares about beyond purity, which a refiner verifies. The only variables that matter to the income statement are ounces produced, the cost per ounce, and the gold price, and Serabi controls only the first two.

The mining and processing flow is where the actual business detail lives:

1. Mining. Both complexes use underground, trackless mechanized mining of narrow veins. Historically the method was selective open stoping (shrink stoping), with mining levels spaced 30-40 metres apart and ramp access that lets ordinary road trucks drive underground. In 2025-2026, helped by the ore sorter and by better-than-expected geological continuity of the Coringa veins, Serabi began converting Coringa from selective open stoping to mechanized sublevel open stoping - a method that is faster, cheaper per tonne, and safer because it keeps people out of the stope. Management targets completion of Coringa's mechanization by Q3 2026.

2. Ore sorting (Coringa only). Crushed Coringa ore passes through a sensor-based ore sorter that rejects roughly 45-50% of the rock as barren before transport. This is the linchpin that makes the satellite-mine model economic.

3. Haulage. Upgraded Coringa material is trucked ~200 km north to the Palito plant on existing roads. Palito ore goes straight to the mill.

4. Processing. At Palito the ore is crushed, milled, and run through gravity concentration plus cyanidation to recover gold. The plant currently runs around 650 tpd. A fourth ball mill is being installed (estimated $5.0 million, fully funded from cash) to lift throughput toward 900 tpd / ~330,000 tpa, targeted operational Q4 2026 and feeding 2027 volumes.

5. Doré and sale. Gold is poured into doré and sold; revenue is recognized at the realized gold price (which in Q1 2026 averaged $4,926 per ounce, against $2,908 a year earlier - a vivid illustration of how much of this business is simply the gold price).

Cost structure. FY2025 cash cost was $1,437/oz and all-in sustaining cost (AISC) $1,816/oz. In Q1 2026 cash cost rose to $1,863/oz and AISC to $2,293/oz, which management attributed to the Coringa ramp-up and the Meio zone's transition into commercial production. Costs are denominated heavily in Brazilian real for labour and consumables while revenue is in dollars, so a weak real flatters margins and a strong real squeezes them.

Geography. Production is 100% Brazilian and 100% in Pará State. There is no export-market diversification because the product is sold into a global commodity market - geography here means operating geography, not sales geography. Everything hinges on two sites in a logistically difficult part of the Amazon, with Palito as the processing chokepoint.

Resources and milestones. Following 38,400 metres of drilling in 2025, the consolidated resource stood at roughly 731,000 ounces measured-and-indicated and 653,000 ounces inferred, with a stated ambition to reach 1.5-2.0 million ounces. The company filed updated NI 43-101 technical reports for both the Palito Complex (effective 30 Jan 2026) and Coringa (effective 30 Jan 2026), filed June 2026, prepared by NCL Ingeniería of Chile - the formal third-party validation of the reserve and resource base underpinning the growth plan. Key historical milestones: Palito plant built and operating for over a decade; Coringa acquired Dec 2017 for $22m; Coringa first gold July 2022; ore sorter commissioned and proving out through 2024-2025; debt fully repaid Jan 2026; inaugural dividend declared for FY2025.

4. Customers

There is, functionally, no customer relationship to analyze - and that is itself the most important fact about the demand side of this business.

Who buys: Serabi sells gold doré to refiners and bullion buyers at the spot price. The "customer" is the global gold market. There is no named account whose loss would dent revenue, no procurement officer to win over, no sales cycle, no negotiation over price, and no qualification testing. Gold is the most fungible commodity in existence; a buyer pays the LBMA-referenced spot price minus a small refining charge regardless of which mine the metal came from.

Why "they" buy / switching costs / concentration: None of the usual frameworks apply. The company has no pricing power and no customer concentration risk in the normal sense, because it can sell every ounce it produces instantly to someone at a transparent world price. The flip side is that it has zero ability to raise prices, zero ability to differentiate, and total exposure to the gold price set by macro forces (real rates, central-bank buying, the dollar, safe-haven flows) entirely outside its control.

Contract structure: Serabi sells at spot. It does not, on the evidence of its disclosures, run a material gold-hedging book - meaning revenue per ounce moves one-for-one with the gold price, quarter to quarter. This makes earnings highly predictable in volume but highly variable in value: the company can tell you almost exactly how many ounces it will pour, and almost nothing about what they will fetch. Revenue predictability therefore comes entirely from production guidance (53,000-57,000 oz for 2026), not from any contractual arrangement.

The practical implication for an investor: do not analyze Serabi's customers. Analyze its ounces and its costs, and form your own view on the gold price separately.

5. Competitive landscape

Gold mining has no "competition" in the product sense - miners do not take share from each other, and a Serabi ounce competes with no one. The real competition is for the things that are scarce: capital, skilled people, drill rigs, and good ground in a workable jurisdiction. The relevant peer set is therefore other Brazil-focused gold producers and the broader universe of junior-to-mid-tier gold miners that compete for investor capital and acquisition targets.

Where Serabi sits: It is a small, high-grade, underground, low-cost, debt-free producer with a single processing hub - distinctive within the Brazilian peer group precisely because it is profitable and self-funding at a size where most peers are still diluting shareholders. Its strengths are grade (6-8 g/t underground vein ore is far richer than the 1 g/t open-pit ore most peers run), a paid-for plant, no debt, and a long exploration runway in an underexplored province. Its exposures are scale (it is an order of magnitude smaller than the Brazilian leaders), single-plant concentration risk, jurisdiction (Amazon permitting and the artisanal-mining backdrop), and total gold-price dependence.

Barriers to entry in this specific niche are real but unusual. They are not patents or capital intensity in the conventional sense; they are permitting and social licence. Getting an environmental licence to operate a cyanidation plant in the Brazilian Amazon, building a camp and roads in a remote region, and maintaining a relationship with surrounding artisanal-mining communities took Serabi years and is the thing a would-be entrant cannot simply buy. Against that, the barriers are low in another sense: there is no technology moat, the gold sells itself, and a better-capitalized major could replicate or acquire the position. This is not a wide-moat business; it is a well-positioned small operator in a hard jurisdiction.

Structural shifts: Brazil is in the middle of a gold-mining upcycle (the Tocantinzinho build by G Mining, Aura's Borborema expansion, renewed exploration interest). Higher gold prices are pulling capital into the country and raising the value of permitted, producing assets like Serabi's - which cuts both ways: it makes Serabi a more attractive acquisition target and a more credible acquirer, and management has said it is "amenable to inorganic growth opportunities" with a "disciplined approach to M&A."

The cleanest comparable is Jaguar Mining - similar size, similar Brazilian underground narrow-vein model, similar ~50,000-60,000-ounce production scale. Serabi's edge over Jaguar in 2025-2026 has been its cleaner balance sheet (debt-free) and its self-funded Coringa growth. Against Aura and G Mining, Serabi simply isn't in the same weight class on scale or diversification, but it wins on grade and on the fact that its growth is paid for from cash flow rather than equity issuance.

Do not over-read a moat here. The "moat" is a permitted plant in a hard place plus high grade. It is durable enough to keep margins healthy at a strong gold price, but it would not protect the company from a gold-price collapse or an operating misstep at its single plant.

6. Industry

What drives demand: Demand for Serabi's product is demand for gold, full stop, and gold demand is macro, not industrial. The price is driven by real interest rates (gold competes with cash and bonds, so falling/negative real rates help), central-bank buying (a major structural tailwind through 2023-2026 as central banks diversified away from the dollar), the US dollar, inflation expectations, and safe-haven flows during geopolitical stress. None of this has anything to do with Serabi's operations - the company is a price-taker on a macro asset.

Industry size and trajectory: The global gold market is enormous and liquid; annual mine supply is roughly 3,500+ tonnes (over 110 million ounces). Serabi's ~44,000-55,000 ounces is a rounding error in that market, which is precisely why it can sell everything it produces at spot with no market-impact concern. The relevant industry trend for Serabi is the gold price itself, which ran from roughly $2,400/oz average in 2024 to $3,481/oz average in 2025 and into the high-$4,000s/near-$5,000 by Q1 2026 - a move that is the single biggest reason Serabi's profit nearly doubled in 2025 and again in Q1 2026.

Where Serabi sits in the supply chain: At the very upstream end - primary mine production. It sells doré to refiners, who turn it into investment bars, jewellery feedstock, and central-bank-deliverable metal. Serabi has no downstream presence.

Regulation: This is the industry's defining feature for a company like Serabi. Brazilian mining is governed by the national mining agency (ANM) and environmental licensing at federal and Pará-state level (IBAMA / SEMAS). Operating in the Amazon adds layers: environmental impact assessments, water and tailings permits, indigenous and community consultation, and a politically charged backdrop because the Tapajós is also home to widespread illegal artisanal mining (garimpo) that draws regulatory and NGO scrutiny. Serabi's status as a permitted, formal, mechanized operator is both its licence to exist and a constant compliance burden. A permitting delay or revocation is an existential category of risk for any single asset.

Cyclicality: Gold is counter-cyclical to risk appetite - it tends to do well when equities and growth are under stress and real rates fall. But gold mining equities are operationally geared to the gold price and can be more volatile than the metal. The industry also faces persistent cost inflation (labour, diesel, reagents, steel) that compresses margins when the gold price is flat - which is why Serabi's rising AISC ($1,816/oz in 2025, $2,293/oz in Q1 2026) bears watching even as revenue soars.

Tailwinds: A multi-year strong gold price, central-bank buying, and a Brazilian mining upcycle pulling capital and validating asset values. Headwinds: Industry-wide cost inflation, a strengthening Brazilian real risk to dollar-margin, and rising environmental/permitting stringency in the Amazon.

7. Growth triggers

All triggers below are drawn directly from the five reporting-period disclosures, with dates cited.

• Fourth ball mill at Palito - throughput from ~650 tpd toward ~900 tpd (~330ktpa in 2027). Installation began in Q1 2026, estimated cost $5.0m funded entirely from cash, targeted operational Q4 2026. (Q1 2026 RNS, 14 Apr 2026; FY2025 results, 1 May 2026)

  "Commenced installation of a 4th ball mill at Palito Complex, expected to be operational by Q4-2026 and will increase annual processing throughput to 330ktpa in 2027." (Q1 2026 RNS, 14 Apr 2026)

• Coringa mechanization (transition to sublevel open stoping) complete by Q3 2026. Better-than-expected vein continuity plus the ore sorter allowed a shift from selective open stoping to a faster, cheaper, safer mechanized method. (Q1 2026 RNS, 14 Apr 2026; repeated theme from Q3 2025 RNS, 14 Oct 2025)

  "Mechanised mining is safer, reduces mining costs, and speeds up production." (Mike Hodgson, Q3 2025 RNS, 14 Oct 2025)

• FY2026 production guidance of 53,000-57,000 ounces (versus 44,169 oz in 2025), reaffirmed at Q1 2026. This is the headline near-term volume ramp, driven by Coringa's Meio/Galena zones reaching commercial output. (FY2025 results, 1 May 2026; reaffirmed Q1 2026 RNS, 14 Apr 2026)

• Coringa grade uplift as Meio zone reaches commercial production (+39% grade improvement cited), with development advancing into the Galena vein. (Q1 2026 RNS, 14 Apr 2026)

• Exploration program of 30,000+ metres in 2026, targeting a consolidated resource of 1.5-2.0 million ounces (from ~731koz M&I + ~653koz inferred). New discoveries (Serra South at Coringa) and extension of the Meio trend feed this. (FY2025 results, 1 May 2026; Q3 2025 RNS, 14 Oct 2025)

  "The standout results have been the discovery of the Serra South zone at Coringa... We've also successfully extended the Meio zone trend." (Mike Hodgson, Q3 2025 RNS, 14 Oct 2025)

• Longer-term vision of a 100,000+ ounce-per-annum combined operation, with 60,000+ oz/annum from 2027 as the intermediate step once the fourth ball mill and Coringa mechanization are complete. (FY2025 results, 1 May 2026)

• Capital returns scaling with cash flow - inaugural dividend established and a stated policy to return 20-30% of free cash flow via dividends or buybacks, a forward commitment that grows mechanically as production and the gold price lift free cash flow. (FY2025 results, 1 May 2026)

• Optionality on M&A - management stated it is "amenable to inorganic growth opportunities" with a "disciplined approach," using the debt-free balance sheet as a platform. (FY2025 results, 1 May 2026)

8. Key risks

• Single-plant concentration (the Palito mill). Every ounce the company produces - from both Palito and Coringa - flows through one processing plant. A major mechanical failure, a tailings-facility problem, a fire, or a permit suspension at Palito would halt the entire company, not one mine. This is the structural risk that no amount of cash on the balance sheet fully neutralizes. The fourth ball mill adds capacity but not redundancy.

• Health and safety - two fatalities in early 2026. This is not a generic risk; it materialized. The FY2025 results disclosed two fatalities in early 2026, an external H&S audit, and the hiring of additional safety personnel. Underground mining is inherently dangerous, and the move toward mechanized sublevel stoping is partly a safety response. Beyond the human cost, fatalities can trigger regulatory stoppages, investigations, and reputational and social-licence damage in a jurisdiction already under environmental scrutiny.

  Management acknowledged the issue directly, citing "renewed Health & Safety focus following two fatalities in early 2026; external audit initiated; additional H&S personnel hired." (FY2025 results, 1 May 2026)

• Total gold-price dependence with no hedge. Serabi sells at spot and appears unhedged. The near-doubling of profit in 2025 and Q1 2026 was overwhelmingly a gold-price story (realized price $2,908/oz in Q1 2025 vs $4,926/oz in Q1 2026). The same leverage runs in reverse: a sharp gold-price fall would compress margins fast, especially with AISC having risen to $2,293/oz in Q1 2026. The business has no contractual cushion against this.

• Cost inflation and the Brazilian real. Costs are largely real-denominated while revenue is dollar-denominated. AISC rose from $1,816/oz (FY2025) to $2,293/oz (Q1 2026), partly Coringa ramp-up but also input-cost pressure. A strengthening real or persistent diesel/labour/reagent inflation would erode the margin even if the gold price holds, and the ramp-related cost bump needs to prove temporary rather than structural.

• Amazon permitting and social licence. Operating a cyanidation plant in the Brazilian Amazon, surrounded by illegal artisanal mining, exposes the company to environmental licensing risk, NGO and political pressure, and community-relations friction. A licence delay or suspension on either the mining or processing side is a low-probability but high-severity event for a single-jurisdiction, single-plant company.

• Execution risk on the ramp. The investment case depends on hitting 53,000-57,000 oz in 2026 and stepping toward 60,000+ in 2027, which requires Coringa mechanization completing on schedule (Q3 2026), the fourth ball mill commissioning cleanly (Q4 2026), and grades holding up. Any of these slipping pushes out the volume story and would test management credibility (see Section 9).

• Loss of cornerstone-shareholder stability / overhang. The 2025 reshuffle of the register (Greenstone fully exited, Fratelli cut to ~10%, Starboard Asset took ~25% via the Classe Roca Magma fund) changed the ownership structure materially. A concentrated holder deciding to exit could create a share overhang, and the termination of the legacy relationship agreements removes some governance scaffolding.

• Resource depth and reserve life. The company is mining high-grade but narrow veins, and the durability of the story depends on exploration continuing to replace and grow reserves toward the 1.5-2.0 Moz target. If drilling disappoints, the multi-year growth-and-life narrative weakens. The newly filed NI 43-101 reports (effective Jan 2026) are the formal anchor here and should be read before underwriting the long-term resource ambition.

9. Walk the talk

Concalls/reporting periods used: Q1 2025 (14 Apr 2025), Q2 2025 (14 Jul 2025), Q3 2025 (14 Oct 2025), FY2025 audited results (1 May 2026), Q1 2026 (14 Apr / 29 May 2026). The most recent is well within 90 days of today.

Serabi's management - CEO Mike Hodgson and CFO Colm Howlin, under Chair Michael Lynch-Bell - has, across these five periods, been notably accurate and slightly conservative. The pattern is one of guiding to a range and then landing in or above it, which is the opposite of the over-promising that plagues most junior miners.

Start at the beginning. In Q1 2025 the company set out a full-year guidance range of 44,000-47,000 ounces and described a steady "rhythm" of ~10,000 ounces a quarter. Hodgson's framing was deliberately unflashy: "We've produced 10,000 ounces for Q1, which followed the 10,000 ounces produced in Q4. So we're certainly maintaining that rhythm." (Q1 2025, 14 Apr 2025). That same quarter brought a board change, handled without drama.

By Q2 2025 the rhythm accelerated rather than slipped: 10,532 ounces, the highest quarterly output since operations recommenced in 2013, with first-half production of 20,545 ounces tracking comfortably toward the 44,000-47,000 range. Development metres hit a record, and grades at both Palito (+27% vs 2024 average) and Coringa (+12%) were improving - exactly the lever management had said would drive the year. Nothing was quietly dropped; the guidance was reaffirmed.

Q3 2025 is the cleanest "did what they said" data point. Production hit 12,090 ounces, a record quarter, +27% year on year, and Hodgson said plainly: "We are pleased to report an excellent third quarter, producing 12,090 ounces, by far the best quarter of gold production in Serabi's history." (14 Oct 2025). Year-to-date production of 32,635 ounces left the company on track for the top of its range, and the ore sorter that management had been talking up for a year was now delivering, "operating for 9 consecutive months with excellent performance." The earlier promise about the ore sorter being a difference-maker had become a demonstrated fact.

The FY2025 results (1 May 2026) confirmed the year landed at 44,169 ounces - inside guidance - with revenue up 65% and post-tax profit up 94%. Crucially, management did two things it had signaled it would do once the balance sheet allowed: it declared an inaugural dividend (5 pence/share) and it set a formal 20-30%-of-free-cash-flow capital-return policy. For a company that spent most of its history capital-constrained and periodically issuing equity, turning cash-generative and initiating a dividend is the most concrete "walk" against years of "talk" about reaching self-funding scale. It also raised 2026 guidance to 53,000-57,000 ounces and laid out the fourth-ball-mill and Coringa-mechanization plans.

Q1 2026 then delivered 12,043 ounces (+20% YoY), reaffirmed the 53,000-57,000 range, confirmed the company was now debt-free (Santander loan repaid January 2026), and showed the fourth ball mill installation underway on the stated Q4 2026 timeline. The one honest blemish: AISC jumped to $2,293/oz on the Coringa ramp - management flagged it rather than buried it, attributing it to the Meio commercial-production transition.

The balance sheet narrative tracks the operational one. Cash climbed quarter after quarter through 2025 ($26.5m Q1 → $30.4m Q2 → $38.8m Q3 → $49.2m year-end → $64.4m Q1 2026) exactly as a self-funding ramp should, and the debt-elimination promise was kept on schedule.

Trackable commitments:

Assessment: This is a management team that does what it says, and tends to set guidance it can clear. Across five periods it hit its production range, accelerated rather than slipped, delivered on the ore sorter, eliminated debt on schedule, and converted years of "we will become self-funding" talk into a real dividend and a formal payout policy. The credibility caveats are forward-looking: the 2026-2027 ramp (ball mill, mechanization, 60koz step-up) is still unproven, AISC is rising, and the two early-2026 fatalities are a serious operational stain that management is addressing but did not prevent. On the financial and production promises, though, the track record is strong and conservative rather than promotional.

10. Shareholder friendliness index

Dividends. Serabi paid no dividend in 2023 or 2024 - it was reinvesting in the Coringa build and managing a constrained balance sheet. For FY2025 it declared its first-ever dividend: 5.0 pence per share (≈7 US cents), roughly 20% of free cash flow, with the final dividend going ex on 25 June 2026 and payable 10 July 2026 (in GBP for UK holders, CAD for Canadian holders), subject to AGM approval on 18 June 2026. So the three-year dividend trend is: nil, nil, then initiation. The company adopted a formal policy to return 20-30% of free cash flow via dividends or buybacks going forward - a meaningful signal that the board now views the company as past its heavy-investment phase and capable of returning capital while still self-funding growth.

Buybacks and dilution. No share buyback program was announced or executed over the last three years; the 20-30% policy permits buybacks as an alternative to dividends, but for FY2025 the return was taken entirely as a dividend (no repurchases). On dilution, the share count has been broadly stable: weighted-average shares outstanding were ~75.7 million across 2024 and 2025, with modest potential future dilution from ~2.2-2.7 million conditional share awards (about 2.9% of issued capital) that vest only on three-year performance conditions. There is no evidence of serial equity issuance over this window - notable for a junior miner, and a direct result of self-funding Coringa. (Sources: FY2025 audited results, 1 May 2026; Dividend Information RNS, 22 May 2026; Director/PDMR RNS, 13 May 2026. No buyback program located in three years of filings or exchange announcements - searched annual results, RNS, and financial press.)

Verdict: Returns Capital (newly). Serabi has just crossed from a reinvest-everything junior into a dividend-payer with a formal capital-return policy and a stable, undiluted share count - the single most important reason being that it now self-funds its growth from cash flow rather than from shareholders' pockets.

11. Insider activities

Serabi is dual-listed (AIM SRB.L + TSX SBI.TO), so insider disclosure flows through UK MAR Article 19 PDMR notices via RNS and Canadian SEDI filings. The primary-source PDMR RNS announcements and the aggregated director-deals record were both accessible.

Recent transactions (most recent first):

Buys - read the signal. The standout is the CEO's open-market purchase of ~45,000 shares at ~£1.69 on 5 June 2025. This was Mike Hodgson's only on-market trade in the trailing twelve months and, per the aggregated record, increased his direct individual holding by roughly 64% in a single transaction. An open-market purchase - cash out of his own pocket at the prevailing price, not an award - by a CEO who otherwise rarely trades, made shortly before two record production quarters and the dividend initiation, is a genuine conviction signal. This is a bullish insider signal: the CEO putting real money into the stock with no recent buying history. The fact that he bought at £1.69 and the shares later traded above £3.40 means the conviction was, in hindsight, well-placed.

Sells - work out the why. The May 2026 sells look superficially large but decode benignly. Hodgson's and Howlin's sales on 12 May 2026 are explicitly tax-related - the routine mechanics of selling a portion of vested shares to cover the tax bill when share awards vest, the standard "sell-to-cover" pattern, not a discretionary exit. Clive Line's sale is flagged option/award-related and he is a former insider (his name was associated with the Q1 2025 board change), so his disposal reflects a departing executive realizing vested awards rather than a view on the business. Critically, these sells happened the same week the company granted new conditional awards and introduced shareholding requirements forcing Hodgson (150% of salary) and Howlin (135% of salary) to build and retain stakes within five years - the opposite of executives heading for the exit.

Net assessment. Over the last twelve months the only discretionary, cash-on-the-table insider trade was a buy - the CEO's open-market purchase. The only sells were tax/award-driven housekeeping on vesting, against a backdrop of newly imposed minimum-shareholding rules that compel management to hold more, not less. The activity is concentrated in two people (Hodgson and Howlin) but points the same direction. Read plainly: mildly bullish - one genuine conviction buy by the CEO with no offsetting discretionary selling, and a governance change that aligns management's required ownership upward. (Sources: Director/PDMR RNS, 13 May 2026; aggregated director-deals record; Hodgson on-market purchase, 5 Jun 2025.)

12. Scenarios

Bull case. The gold price stays strong, and Serabi's operational leverage does the rest. Coringa mechanization completes on schedule by Q3 2026, the fourth ball mill commissions cleanly in Q4 2026, and through 2027 the combined operation runs at 900 tonnes per day, pushing production past 57,000 ounces and toward the 60,000-ounce step. The ore sorter keeps feeding high-grade material to a now-debottlenecked plant, AISC settles back as the Coringa ramp costs normalize, and free cash flow compounds on a high gold price. The 2026 drilling program keeps hitting - Serra South and the Meio trend extensions push the consolidated resource toward the 1.5-2.0 million-ounce target, lengthening mine life and de-risking the long vision of a 100,000-ounce producer. The dividend grows under the 20-30%-of-free-cash-flow policy, the debt-free balance sheet funds either a buyback or an opportunistic, accretive acquisition in the Brazilian upcycle, and a company that the market still treats as a junior gets re-rated as a sustainable, dividend-paying mid-tier. The CEO's June 2025 open-market buy looks prescient.

Base case. Management does roughly what it has guided. Production lands in the 53,000-57,000-ounce range for 2026, the ball mill and Coringa mechanization arrive close to schedule with the usual minor slippage, and 2027 brings the move toward 60,000 ounces. AISC stays elevated for a few quarters on the Coringa ramp before easing, so margins are healthy but not spectacular, and earnings track the gold price more than any operational surprise. The dividend is maintained and modestly grown, the share count stays stable, and exploration steadily replaces reserves without a transformational discovery. The two early-2026 fatalities prompt a real H&S overhaul that costs some money and management attention but does not trigger a regulatory stoppage. Serabi remains what it is: a small, well-run, debt-free, high-grade Brazilian gold producer whose fortunes rise and fall mostly with the gold price, executing competently against a credible plan.

Bear case. The gold price rolls over, and the operational leverage that powered 2025-2026 runs in reverse - with no hedge and AISC already near $2,300/oz, margins compress fast and the dividend story stalls almost as quickly as it began. Worse, the single-plant concentration risk bites: a mechanical failure, a tailings issue, or a permit suspension at the Palito mill halts the entire company because every ounce from both mines flows through that one facility. The early-2026 fatalities escalate into a regulatory stoppage or a social-licence crisis in an Amazon jurisdiction already hostile to mining, and the Coringa ramp disappoints - mechanization slips, grades at depth fade, or the ball mill commissioning runs into problems - pushing the 2027 volume step out by a year and denting the management credibility that has been the stock's anchor. A cornerstone shareholder from the reshuffled register decides to exit, creating an overhang into a falling market. None of these alone is fatal given the cash and zero debt, but in combination they turn a self-funding growth story back into a constrained single-asset junior at the mercy of the gold price.

Sources

Company filings & RNS

• Q1 2026 Production & Operational Highlights (GlobeNewswire, 14 Apr 2026)
• Unaudited interim results, three months ended 31 March 2026 (GlobeNewswire, 29 May 2026)
• Audited Results for the year ended 31 December 2025 (GlobeNewswire, 1 May 2026)
• Record Q3-2025 Production of 12,090 oz (GlobeNewswire, 14 Oct 2025)
• Interim results, three and nine months ended 30 September 2025 (GlobeNewswire, 28 Nov 2025)
• Q2-2025 Production Results (GlobeNewswire, 14 Jul 2025)
• Q1-2025 Production Results and Board Change (GlobeNewswire, 14 Apr 2025)
• NI 43-101 Technical Reports for Palito Complex and Coringa (GlobeNewswire, 11 Jun 2026)
• Dividend Information (GlobeNewswire, 22 May 2026)
• Director / PDMR Shareholding (GlobeNewswire, 13 May 2026)
• Holdings in the Company - shareholder reshuffle (GlobeNewswire, 23 Apr 2025)
• Coringa Gold Project (Serabi Gold website)

Insider / director deals

• Serabi Gold director deals (Hargreaves Lansdown)
• Off-market/on-market insider buying at Serabi Gold (Globe and Mail)
• Serabi Gold shares analysis incl. CEO purchase (Simply Wall St)

Industry & competitors

• Gold Explorers Gain Ground as Brazil's Mining Boom Picks Up Speed (GlobeNewswire, 4 May 2026)
• Brazil Mining 2024 - Gold (GBReports)
• Aura Minerals market cap (CompaniesMarketCap)
• Jaguar Mining market cap (StockAnalysis)
• G Mining Ventures market cap (CompaniesMarketCap)

A quick caveat on two numbers I estimated for charting continuity: the Q4 2025 standalone production (~11,534 oz) is derived as FY2025 total (44,169) minus the three reported quarters, and the 2026E bar uses the guidance midpoint (55,000) - both are clearly labelled as derived/guidance rather than reported figures. Everything else in the charts is taken directly from company filings.