Managed mandates
Owner keeps title, hires us to run the forest. No tax event, no acquisition capital, outside AIFMD. This is the part that scales — most hectares arrive here.
Private & confidential — for founder + investor only
Latvian forest fund — concept v2
Own it. Manage it. Grow it. Invest.
A vertically-integrated Latvian forest business: we buy the right parcels, manage them — ours and owners' — grow their value with science and AI, and pool them into an investment anyone can back. Land trades at ~€4,000/ha, a fraction of the €17–23k a mature stand is worth, and it's fragmented across 110,000+ owners. We consolidate, optimise and compound — an honest ~5–7% net for investors, venture-grade upside in the manager.
~€4,000/hawhat we pay — vs €16.8k–23k/ha a mature stand is worth
5–7%honest net return, nominal (≈2–4% real) — not double-digit
1.55M haprivate forest · 110–133k owners · 92% hold <20 ha
€165–200Mforest land changes hands per year — fragmented, thin
The whole case in one screen
Why this wins.
Faster & sharper
- AI decides in minutes — faster and more precise than every incumbent — better data, better assumptions.
- So we can pay a premium — and still win the parcel, because we underwrite sharper than anyone.
- Consolidation creates value — scattered ~9 ha parcels are worth more as one clean, managed block.
- mezatirgus.lv already has traction — owners, sellers and buyers reach us online first.
Cheaper to run
- A fraction of the cost — modern monitoring + maintenance vs the old players' manual fieldwork.
- Lower expenses everywhere — our own software does the work — that's the margin.
- Smarter selective cutting — generate value, not just volume.
- A data flywheel they can't rebuild — VZD + free LiDAR + a live timber feed, all ours.
A better deal for owners
- Don't sell — pledge (ieķīlāt) — fully covered, and wait for the price to grow.
- We grow it while you wait — well-maintained, compounding forest — not a one-off cut.
- Funding + EU grants — to grow new forest on land that never had it.
- No upfront cost — you pay only when we harvest.
A safe, scalable asset
- Low risk on what we buy — land is a hard, appreciating asset.
- Built to scale — a poolable investment object, not a one-off deal.
- Quick to start — first results you can feel fast.
- Processing optionality — move up into higher-value timber where we spot it.
- Founder edge — forest-industry know-how + the contacts to hire.
The model · in one screen
Two ways owners join, one fund, and the venture sitting on top.
The original idea — “everyone pools their forest, others invest” — is half right. Pooling land for units is a taxable disposal, so it adversely-selects the worst parcels; it stays a small, gated lane. The scalable business is managing forest for a fee. Value lives in the standing timber, not the dirt — and the equity lives in the manager, not the land.
Pool land for units
Owner contributes land for fund 'land units'. Honestly limited: it's a taxable disposal, so reserved for 60-month-exempt / low-basis owners — kept under 10% of the pool.
Closed-end land fund
Owns + pools land for EU/EEA/OECD cash investors. NAV-priced, honestly illiquid (10–15yr), ~5–7% net. A sensible holding — not where the venture upside sits.
OpCo + mezatirgus moat
Asset-light manager: 1% NAV fee + 10% harvest commission + €8/ha retainer + sourcing fees + the mezatirgus data layer. A compounding fee annuity. You own the manager, not the trees.
The opening
A fragmented, thinly-traded market — with a data moat over it.
~1.55M ha of private forest, 110–133k owners, 92% holding under 20 ha. ~3,640 arms-length deals a year averaging ~9 ha. Assembling and managing scattered parcels on a market this thin is hard for anyone without origination + valuation infrastructure. We already built it.
Deal-size distribution · 2023–26, n=12,458
Three-quarters of deals are under 10 ha. Bigger blocks are scarce and trade at a slight premium — fragmentation is the structural opening.
Regional median · €/ha, 2022–26
€4,148Cēsis€4,000Madona€4,000Limbaži€3,958Valmiera€3,968Rēzekne€3,576Krāslava
82,201 forest parcels mapped (≈2.38M ha)
51,865 arms-length deals priced
18 buyers live timber feed, refreshed daily
Institutional appetite is real: Ingka (IKEA) paid €720M for ~153k ha (≈€4,700/ha all-in), closing Jan 2026 to become Latvia's largest private owner (~245k ha). We play the small-parcel lane they can't reach efficiently.
If you own forest — what you'd get
Keep your forest. We grow what it's worth — and you pay only when it pays you.
Written for a small or mid-size forest owner. You don't sell, you don't hand over control, and you don't pay us to wait. You get a living picture of what you own, the fastest science-backed path to grow its value, cover against the bad years — and, if you want, cash today that your forest's own growth pays back.
You pay only when your forest pays you
No retainer, no monthly fees, nothing out of pocket while the forest grows. We earn a share only at a cutting event — when wood is actually sold. Between harvests your cost is €0, and our only way to get paid is to grow what your forest is worth.
A live picture of what you own
Plain-language reports per parcel: standing volume, growth, health, and today's value at live timber prices — with the recommended next move and when. Own several forests? One accumulated portfolio view across all of them.
LiDAR & remote-sensing, shared with you
The same data we manage with: LiDAR-derived height, volume and density, aerial/drone imagery, maps, and change-detection that flags storm or beetle damage early. A measured picture of your stand — not a guess.
The fastest path to grow value
An up-to-date, science-based plan tuned to your goal: where and when to thin, what to regenerate with (improved genetics), and how to shift your future cut out of €53 pulpwood into €100+ sawlog. Refreshed as the forest grows and prices move.
Insured against the bad years
Cover arranged against windthrow/storm, fire, pests & disease (bark beetle) and illegal felling — so one bad season doesn't erase a decade of growth. You see exactly what's covered.
You stay the owner
You keep the title and approve every harvest. Mark stands you never want cut. Leave with notice. We manage; you decide — and because it's a management contract, joining is not a taxable event.
Liquidity · ieķīlāšana, without the interest treadmill
Cash now — repaid by growth, not by you.
Want money in hand today? Unlock it against your forest with no monthly interest payments. Instead of a classic mortgage that bills you every month, the advance is settled later — from a future harvest and the value your forest gains while we grow it. Nothing leaves your pocket in the meantime; the forest covers it. We provide the advance directly or arrange it through a financing partner, secured against the standing timber we value and manage — so your cost is a pre-agreed amount at settlement, not a running rate.
1 · ValueWe value your standing forest from LiDAR + live timber prices.
2 · AdvanceYou take cash today against it — no monthly interest, no payments out of pocket.
3 · GrowWe keep managing the forest so its value compounds.
4 · SettleA pre-agreed amount is settled from a future harvest — the forest covers it, not your wallet.
How we start it legally: as a forward timber purchase (buying a share of the future harvest — a sale, not a loan), so no consumer-credit licence is needed to launch — while we apply for the PTAC lending licence in parallel to scale. Subject to independent valuation + clean title. For the investor it's also a new OpCo line: a financing spread (€0 in the base case until launched).
For the investor: this owner-friendly design — pay-at-harvest, zero upfront cost, optional growth-covered liquidity, real data + insurance — is precisely why owners choose the mandate lane. It's what turns the 9,500-ha managed ramp into credible demand, not a hope.
The edge — how we outcompete
An AI-first operator on a market that still runs on seminars and single-buyer handshakes.
Latvia's forest-owner market is served by slow state advisors, mill procurement arms whose “advisor” is also the buyer, and co-ops that signed ~1,000 of 110,000+ owners in fifteen years. We win on three things they can't copy quickly: speed (AI, not fieldwork), alignment (buyer-agnostic, paid only at harvest), and a data moat (mezatirgus + VZD + free national LiDAR).
AI runs the company — valuation to compliance, in minutes
AI fuses free national LiDAR + VZD cadastre + live mezatirgus comparables to value, score and growth-model any parcel — replacing a taksators' day in the field.
An LLM writes live per-parcel + portfolio reports and bilingual owner messages, on schedule and on event (price move, windthrow, permit) — not grant-cycle seminars.
AI matches each stand to the right buyers and runs an open auction with published comparables as the floor — killing the single-buyer lowball.
Automated canopy-height, volume and clear-cut/windfall/fire detection from the national point cloud; drone + RTK only where the value justifies it.
Geolocation + due-diligence dossiers, FSC/PEFC paperwork and felling notifications, auto-generated — compliance as a feature, not a cost centre.
AI ranks inbound owners and parcels by value and effort-to-serve, so a 5–6-person team works only the highest-leverage long-tail cases.
Foresta · the software wedge
Free to our owners. Rented to everyone else.
Foresta — the first consumer-grade, mobile-first forest app for the 92% of owners holding under 20 ha: instant AI valuation, a live multi-parcel portfolio, a harvest planner, open-auction access and an EUDR vault. Confirmed white space — every Baltic/Nordic tool is enterprise-grade and quote-priced; nothing owner-facing exists.
Free foreverto managed owners + partner foresters/co-ops — maximises data capture and creates lock-in + a network effect
Rented (SaaS)to independent foresters, firms, municipalities, valuers — even competitors, who then feed our comparables
~€5–15 / ha / yrtransparent pricing in an all-quote market; a roadmap line on top of mezatirgus revenue, not a Y1–3 dependency
Guaranteed-price auctions · the flywheel
The seller magnet that quietly feeds our buying pipeline.
On the independent mezatirgus.lv we run real timber and forest auctions with a guaranteed floor price the seller is promised up front. They get certainty no single-buyer mill offers — a known worst case, and the bidding can only go up. That pulls deal flow to a platform sellers trust — and the guarantee is backed by our own readiness to buy, so the marketplace and our land-acquisition pipeline are one and the same funnel.
1 · Guaranteed floorWe underwrite a minimum price from our buyer network + live data; the seller signs knowing the worst case — certainty no single-buyer mill offers.
2 · Deal flow concentratesSellers bring timber and land to a neutral, independent platform they trust — so we see the whole supply first.
3 · We win three waysEarn the auction fee, convert the seller into a managed-mandate owner — and if the lot doesn't clear above the floor, the fund buys it at the floor: a disciplined acquisition at a price we set.
How we beat each incumbent
CompetitorTheir gapHow we win
Mill arms — Finieris · Stora Enso · MetsäBuy timber & felling rights as mill feedstock; bundle “advice”.The advisor IS the buyer — opaque single-buyer pricing, no benchmark.Owner keeps title; we run a multi-buyer auction on aligned commission, paid only at harvest.
MKPC / LLKC — state advisoryFree, grant-funded advice, plans and seminars.Grant-cycle, document-and-seminar; never transacts or manages the asset.AI delivers the advice instantly — then we capture the auction/harvest MKPC never touches.
Co-ops — Mežsaimnieks et al.Members pool timber to reach enterprise prices.Manual, membership-bound — ~1,000 of 110–133k owners in 15 years.AI onboarding at 10–100× their speed; a managed mandate, not a membership.
TimberMarket.lvSubscription RFQ board for ~1,000 companies.B2B, RFQ not open auction; no managed owner journey.Owner-facing open auction + the full managed journey + a free portfolio app.
Latio — broker / valuerOne-off appraisals from ~€300.Transactional; no ongoing management, manual turnaround.Instant AI valuation bundled into an ongoing managed mandate.
LFDF — buy-low-and-flipBuys 10–50 ha, flips blocks to institutions; ~35% margin on ~14.5% P2P capital.Owner loses title AND the upside; expensive capital.Owner keeps title + the appreciation; we earn aligned fees, not flip margins.
Banks — Swedbank · SEB · LuminorNo forest-collateral lending product.Won't underwrite a fragmented absentee owner against future harvest.Growth-covered liquidity — we hold the stand data + harvest schedule, so we can.
Institutions — Ingka · INVL · SCABuy large blocks in single deals.Can't source clean blocks from 110–133k tiny owners.Not a rival — our exit: aggregated managed hectares + an AI data room = premium blocks they buy.
Per-hectare unit economics
Built from real timber prices — and it stands without appreciation.
The income case is deliberately conservative: ~3% cash yield on the land, derived from live delivered prices minus real harvesting + haulage. Total return adds biological growth and a modest land move on top — but the floor doesn't depend on land going up, which matters in a drawdown.
Stumpage = delivered price − (harvest + forwarding + haul). Sawlog nets €70–94/m³; pulpwood only €23–35.
Acquire€4,000/ha
≈9 ha parcel ≈ €36k · just above the €2,200 bare-land floor
Sustainable harvest4.5 m³/ha/yr
65–85% of the ~6.3 m³ increment floor (ramps from young)
Net stumpage~€45/m³
delivered ~€78 − harvest+haul ~€32 · sawlog €70–94, pulp €23–35
Gross timber cash~€200/ha/yr
4.5 × €45 · band €160–250
− management~€75/ha/yr
€40 fee (1% of NAV) + €35 silviculture/admin
Net income yield~€125/ha = 3%
stands WITHOUT appreciation — that's the point
The return stack — honest
~7–8% gross, ~5–7% net nominal, ~2–4% real. Not double-digit.
Institutional timberland is a mid-single-digit real asset; Nordic/Baltic sits at the bottom of the global band. Anything double-digit needs leverage or a repeat land boom — both are risk, not return. So we model it straight.
volume in-grows pulpwood → sawlog · price-independent · most reliable
conservative — explicitly NOT the 2021–23 boom
thinning + assortment · CAPPED, must be proven
land flat/falling, weak EU timber demand persists, alpha never shows. You'd have done better buying parcels directly.
biology + 2–3% land + modest proven alpha, net of 1% fee. The underwrite that survives diligence.
land re-rates + assortment alpha delivers + timber cycle recovers. Upside, not plan. Double-digit needs leverage or a repeat boom.
Underwritten on a drawdown, not a boom: land is −22.7% off the 2023 peak and turnover fell €168M → €122M. Land appreciation is held at 2–3%, not the 8.4%/yr full-period CAGR.
5-year portfolio ramp
The managed lane outgrows the owned lane — on purpose.
Owned + in-kind hectares carry the fund NAV; managed mandates carry the OpCo and need zero acquisition capital. In-kind stays tiny (350 ha by Y5) because the tax makes it self-limiting — that's honest, not pessimistic.
YearHectares (owned · in-kind · managed)Fund NAVTimber turnover
Y1€2.0M€0.15M
Y2€4.4M€0.54M
Y3€7.2M€1.30M
Y4€10.8M€2.42M
Y5€15.0M€4.0M
Owned In-kind Managed mandateY5: 13,250 ha operated ≈ 0.85% of LV private forest · NAV held at cost, not marked up
The OpCo — the actual venture
You own a compounding fee annuity, not a pile of trees.
The forest is the investors' asset — a ~5–7% holding no rational LP overpays a GP for. The venture-grade equity is the management company: light capital, recurring fees that scale with AUM and managed hectares, and a data moat a DIY buyer can't rebuild.
~€700kOpCo revenue, Year 5
~€2–3M+by Y10 (€50M+ NAV, 30k+ managed ha)
Carry is ≈€0 in the base case. We say so — and turn rarely-bites into an alignment selling point.
Management fee
1.0% of fund NAV / yrTIMO norm 0.85–1.0% · on NAV, not committed capital
Harvest commission
10% of timber revenueon third-party managed land — the asset-light engine
Stewardship retainer
~€8 / ha / yraccrues + settles at harvest — no cash from the owner between cuts
Sourcing fee
1.5% of parcel valueone-time per closed acquisition
mezatirgus platform
~€120k / yr by Y5featured listings, profiles, valuation & cruise, data API
Carried interest
20% over 6% hurdlehonest: ≈€0 in base case — alignment, not income
The 10-year build
How the portfolio — and the business — compound.
Every figure ties to the verified per-ha economics (~€4,000/ha, ~€45/m³ net stumpage, 4–5 m³/ha/yr) — no hockey sticks. The forest return stays a mid-single-digit 5–7% net; the wealth compounds from scale — fees on a growing AUM plus 32,000 managed hectares — not from a forest-price boom.
How the forest properties grow
~110–160 parcels/yr at ~€4,000/ha — the fund buys the mispriced, well-located stands the data flags. Owned/pooled grows 500 → 10,700 ha.
Mandates scale with zero acquisition capital and no owner tax event: 500 → 32,000 ha. This is where the hectares really compound.
Kept under 10% — a small, gated lane for tax-exempt / diversifying owners; adds hectares without cash.
The book grows ~4.5%/yr above income — biological in-growth + modest appreciation, pulpwood maturing into €100+ sawlog.
Fund NAV · €2.0M → €50M
Y1Y2Y3Y4Y5Y6Y7Y8Y9Y10
The numbers, year by year
YearOperated haFund NAVTimber turnoverOpCo revenueOpCo EBITDA
Y11,000€2.0M€0.15M€62k−€258k
Y22,900€4.5M€0.54M€129k−€301k
Y35,800€7.5M€1.3M€259k−€281k
Y49,200€11.4M€2.4M€451k−€229k
Y513,250€16.1M€4.0M€705k−€115k
Y617,900€21.5M€6.0M€1.06M+€76k
Y723,200€27.6M€7.8M€1.42M+€274k
Y829,100€34.5M€9.8M€1.85M+€526k
Y935,600€42.0M€12.0M€2.33M+€847k
Y1042,700€50.3M€14.4M€2.87M+€1.22M
What it builds
€50Mfund NAV by Y10 — LP wealth at an honest 5–7% net (€42.8M deployed + €7.5M growth)
€2.9M / €1.2MOpCo revenue / EBITDA by Y10 — EBITDA-positive from Year 6, ~43% margin
~€10MOpCo enterprise value (≈8× EBITDA) — the venture equity, off a ~€0.75–1.4M seed
AsymmetricForesta SaaS + the data moat — the breakout upside, not even in the base case
Honest: break-even is ~Year 6, so the OpCo needs ~€1.2M of cumulative funding to get there — the €750k base seed reaches proof, then a small raise (or the €1.4M aggressive seed) covers the gap. The 9,500 → 32,000 managed hectares need no acquisition capital at all.
Structure & how to join
One Latvian SIA manages; one closed-end SIA holds.
Split the venture equity (management contracts + data) from the illiquid land AUM — the classic GP/ManCo-vs-LP shape. The mandate lane is a genuine operating business, so it falls outside fund regulation.
OpCo
Latvian SIA. Actively manages forest + runs mezatirgus. Earns the fee stack above. Outside AIFMD as an operator/manager.
The Fund
Closed-end holding SIA, two share classes (cash + in-kind land units). Light 'registered AIFM' regime — ≤€500M, €15k capital, Bank of Latvia.
Mandates
Third-party owners keep title, hire OpCo. No tax event, no AIFMD, zero acquisition capital.
mezatirgus
782k VZD transactions, 51,865 priced deals, live timber feed, 3,970 company shells — origination funnel + valuation a DIY buyer can't replicate.
Owner — mandaterecommended
Keep title, sign a management contract. No tax event. We run harvest/silviculture; you keep the timber uplift less a 10% commission + €8/ha. The default path.
Owner — in-kind
Contribute land for fund units. Taxable disposal (CGT 20–25.5% + 1% duty capped €50k). Only for 60-month-exempt / low-basis / diversifying owners. Cruise-valued + clawback.
Investor — cash units
Closed-end NAV-priced units, 10–15yr lock, explicit end-of-life sale. EU/EEA/OECD on par; third-country via a feeder at unit level. Raises <€8M; retail via ECSP ≤€5M.
Legal & tax — corrected facts
The points an investor's counsel will press on, stated straight.
Who can hold landEU/EEA/OECD buy on par — there is NO 51% rule for them; that's only the third-country route.
VehicleRegistered sub-threshold AIFM (≤€500M, €15k min capital, Bank of Latvia). Mandate lane sits outside AIFMD entirely.
In-kind cost~1% capital duty capped €50k/property (not 2.5%). Independent valuation required.
Timber incomeFlat 10% PIT for individuals (≈7.5% after the 25% regeneration deduction). 0% CIT on retained corporate profit — recycle harvest cash tax-free.
Raise limitsKeep each 12-mo raise <€8M prospectus exemption → harmonising to €5M/€12M ~mid-late 2026 (EU Listing Act). ECSP retail ≤€5M/parcel.
Land transfer VATForest-land transfers generally VAT-exempt. Roundwood sales are VATable (domestic reverse-charge).
Structures considered and rejected (5)
- Full licensed AIFM — €125–300k capital is overkill far below the €500M sub-threshold.
- “Everyone pools land for units” as the primary model — dry CGT charge drives adverse selection.
- Listed/closed public vehicle (Phaunos / Latvian Forest Co) — chronic NAV discounts, wind-up risk.
- Open-end / daily-redemption — forest is illiquid; daily NAV is a structural lie that invites a run.
- Pure land-banking — appreciation engine is in drawdown (−22.7% off peak); return collapses to ~2–3%.
Licences & legal barriers
What it really takes — beyond registering a SIA.
The de-risking insight: the management business + the auctions need no financial licence — register the SIA and start. Only two activities are licensed, bolted on later: the fund(a registered AIFM) and the lending product (the PTAC perimeter — which you sidestep at launch via forward purchase). Keep all three in separate entities, and AML on top of everything.
The licences — and when each one bites
LicenceRegulatorWhen it bitesCost · time
Registered AIFMLatvijas BankaWhen the fund pools outside investor money€15k cap · €250 · ~1 mo
≤€500M, unleveraged, no 5-yr redemptions. NO EU marketing passport — market only in Latvia / national private placement; professional + qualified-retail only (retail mis-sale fine €14,200).
Full licensed AIFMLatvijas BankaScale-up only — to passport across the EEA or exceed the caps€125–300k · 3–6 mo
Grants the EEA professional passport but triggers a depositary, independent valuation + Annex IV reporting. Graduate into it; don't start here.
Consumer-credit licencePTAC (→ Latvijas Banka 2027–28)Only to advance real cash credit to consumer owners at scale€250k issue + €55k/yr
A deliberate high wall. Fit-and-proper, creditworthiness assessment, 14-day withdrawal, 0.07%/day total-cost cap. Min capital unconfirmed (~€425k) — confirm. NOT needed if you run forward-purchase first.
Credit-intermediary registrationPTACIf you originate/service for a licensed partner-lender€1,000 + €500/yr + PII
Far cheaper than the full licence — a real scaling route. Being extended to all consumer credit under CCD2.
AML obliged-entityVID / Latvijas Banka · report to FIDBefore the first transaction — runs on top of everythingno big fee · ongoing
Both the AIFM and any financier are 'finanšu iestādes'. KYC, beneficial owner (>25%), sanctions (incl. the third-country chain), goAML reporting. Can't be structured around.
The lending path · forward-purchase now, licence later
Start as a sale, not a loan — then apply for the licence in parallel.
Advancing cash to an individual owner is consumer credit, and the PTAC licence is a wall: €250k to issue + €55k/yr. So we launch as a forward timber purchase — buying the future harvest at a discount today, a true sale that sits outside the licence — and apply for the PTAC licence in parallel to scale into real cash advances. Counsel-essential: a “sale” with fixed repayment + recourse can be recharacterised as a disguised loan, so the deed must put genuine price risk on us, with no recourse beyond the timber.
Route 1 · Forward purchaseBuy the cirsma / a share of proceeds — a true sale, not a loan → no PTAC licence. Start here. Only AML + a felling-permit check. (Advocate papers it so it isn't recharacterised as disguised credit.)
Route 2 · Lend to business ownersCredit to an owner acting as a registered business (saimnieciskā darbība / their SIA) is outside the consumer regime — but most <20 ha owners are consumers, so it needs a borrower-status opinion.
Route 3 · Partner + intermediaryOriginate/service for a licensed lender → only a ~€1k credit-intermediary registration, not the €250k licence.
Route 4 · Own PTAC licenceLend on your own book — most scalable, most expensive (€250k + €55k/yr). The parallel scale-up target once volume justifies it.
Land & operating gates
Nationality chain
EU/EEA/OECD buy on par — but a third-country UBO anywhere in the chain bars the parcel. Ring-fence the cap-table so one never appears.
Forest-dominant escapes the agri regime
If forest is the dominant cadastral category: no municipal consent, no B2 Latvian, no 2,000/4,000 ha caps, no Land-Fund pre-emption — only VMD forestry duties. Favour forest-dominant parcels.
Agri-dominant → municipal consent
Mixed parcels where agriculture predominates need a pre-deed consent from the municipal commission (not LAD), ~1 mo + a ~2-mo pre-emption window.
Pre-emption
On agri land: co-owners → registered tenant → the Latvian Land Fund (ALTUM) can pre-empt before you complete. Price the timing risk in.
Security
Standing trees ride with the land — a hipotēka in the Land Register captures the timber; any non-bank can hold it. The clean route for individual owners (who usually can't grant a komercķīla).
VMD · EUDR · data
Felling permit (€70, 3-yr) + mandatory reforestation before any harvest; EUDR geolocation/DDS from Dec 2026; VZD geometry + LGIA LiDAR are open/commercial, but owner identity is GDPR-restricted — no scraping.
Not legal advice. Load-bearing figures are verified to 30 Jun 2026, but the consumer-credit regime is a moving target (supervisor shifting PTAC → Latvijas Banka 2027–28; CCD2 from Nov 2026). A Latvian advocate must sign off the true-sale deed, the borrower-status test, the AML supervisor and the cap-table before launch.
How we build it — plan · team · capital
Two pools of money, a lean team, and four phases to steady state.
The most important correction to the earlier plan: the money is two separate pools. A small OpCo seed funds the team and platform (the venture); the fund/LP capital buys hectares (the AUM). They never mix — which is exactly what made the old “€1.8–2M buys 450 ha and funds the company” maths impossible.
Pool A · the venture raise
~€750k OpCo seed equity
18–24 mo runway to fund first close + first harvest-commission / SaaS revenue, after which fees self-fund opex. Your return is fees + carry + SaaS — not owning trees.
Pool B · the AUM
€2.0M first close → ~€13.6M by Y5
LP capital that buys/pools title at ~€4,000/ha, drawn in tranches as parcels clear diligence. It never pays OpCo opex.
- €2.0M Y1 first close → 500 owned/pooled ha
- 9,500 ha managed by Y5 need €0 of this — owner keeps title
- ~€15.0M NAV by Y5 · 5–7% net nominal / 2–4% real
- EU/EEA/OECD LPs on par; third-country via a feeder at unit level
Four phases, each with a go/no-go gate
Phase 0Months 0–4
Stand up the OpCo, ship the AI valuation MVP, close the €750k seed.
- Incorporate the OpCo SIA (€2,800, ~1–3 days)
- Raise ~€750k seed (ECSP ≤€5M; raise stays <€8M)
- Hire lead + GIS engineer; get 2–3 AIFM-setup quotes from Riga counsel
- Ingest free LGIA LiDAR + VZD + mezatirgus; ship the AI valuation MVP
- Build Foresta owner app v1 (valuation, live report, document vault)
- Pre-engage Bank of Latvia on the registered-AIFM scope
Gate
AI valuation lands within tolerance of certified taksācija on a 25-parcel test; seed funded.
Phase 1Months 4–10
Sign the first 100–300 owners, run live auctions, prove harvest-commission revenue.
- Onboard pilot owners toward ~1,000 managed ha
- Run 3–5 live timber auctions; book the first 10% commissions
- Ship buyer-matching + EUDR doc generation
- Hire a forester/mežzinis + BD/onboarding lead
- Instrument cost-to-serve per owner and net per harvest
Gate
≥1,000 ha managed · ≥3 auctions beating the prior bilateral offer · positive gross margin per harvest.
Phase 2Months 10–20
Scale mandates via AI onboarding, launch Foresta SaaS, reach fee break-even line of sight.
- Scale toward 500 owned/pooled + 500 managed via automated triage + comms
- Launch the Foresta SaaS paid tier
- Bring drone + RTK capture in-house
- Build the fund data room + acquisition pipeline (1.5% sourcing)
- Open exit talks with Ingka / INVL / SCA
Gate
Managed-ha fees + SaaS MRR on a path to cover OpCo opex; ≥500 ha diligence-ready for first close.
Phase 3Months 18–30
Register the AIFM, close the €2.0M fund, deploy first land, pilot liquidity.
- Register the sub-threshold AIFM (Bank of Latvia, ~1 mo); incorporate the Fund SIA + rules/PPM
- First close ≥€2.0M LP capital; buy/pool 500 ha at ~€4,000/ha
- Stand up the financing entity; settle the first growth-covered advance from a harvest
- Establish statutory audit + outsourced AIFM compliance
Gate
AIFM registered · fund first close ≥€2.0M · first parcels acquired · first advance settled from harvest.
The team — lean by design (~5–6 FT)
Founder / CEORaise, fund, partnerships, early BD; deferred pay stretches runwayPhase 0FT~€30k draw
Lead full-stack / AI engineerThe spine: AI valuation, Foresta, auction engine, automationPhase 0FT~€62k loaded
Mid + GIS / remote-sensingLiDAR/RS pipeline off the free LGIA point cloud + SentinelPhase 0FT~€44k loaded
Forester / mežzinisField trust + auction execution — a brand barrier we must clearPhase 1FT~€25k loaded
BD / owner-onboardingOnboards the fragmented long tail at AI-assisted speedPhase 1FT + comm.~€43k + var.
Certified taksatorsFund-acquisition / audit-grade valuations only; AI self-serves routinePhase 1+Contract€8–25/ha
Counsel · books · audit · complianceRiga fund counsel + outsourced accounting + AIFM adminPhase 0/3Outsourced€15–40k+ setup
Entities
OpCoSIA · €2,800
AI-first management company — holds the Foresta IP, employs the team, runs mandates, earns the fee stack + carry. The venture equity (Pool A).
~1–3 days · ~€220–490 all-inFund holdcoClosed-end SIA + AIFM
Holds title to owned/pooled hectares; takes LP capital. The 9,500 managed ha sit OUTSIDE it (owner keeps title).
AIFM €15k capital · Bank of Latvia ~1 mo · ~2–4 mo totalFinancing entitySIA · Phase 3
Growth-covered liquidity — advances cash settled from a future harvest, direct or via a lending partner. Off the OpCo/fund balance sheets.
~1–3 days, once the product is validatedFast-track — 3× faster
Or we go aggressive: ~10 months to a fund close, not 30.
The base plan is deliberately capital-efficient. With more upfront capital and AI as a genuine force-multiplier, we compress everything compressible and run it all in parallel from day one — ~10 months to first fund close and operating revenue. Honest: the regulator, tree biology and harvest seasons don't 3×; everything else does.
MilestoneBaseAggressive
Time to fund first close~30 months~10 months
Team in placephased — hire as you gofull team, day one
OpCo seed (Pool A)~€750k~€1.4M
Foresta + AI MVP livemonth ~4week ~6
~1,000 ha under mandatemonth ~10month ~4
AIFM / legal track startsPhase 3 · month 18day 1 · in parallel
What makes 3× real
A ~€1.4M seed (vs €750k) hires the whole team on day one, funds paid owner-acquisition + partner channels, and starts the AIFM/legal track immediately instead of in Phase 3. Capital compresses the calendar.
AI-assisted development ships the platform in weeks, not quarters; AI runs valuation, onboarding, reports, auction ops and compliance so 5–6 people output like 20; pipeline triage onboards owners at machine speed.
Platform, mandates and fund-legal advance at once — not sequential phases. Design-partner owners from week one; fail-fast gates replace waterfall. We learn in days, not quarters.
~10 months · four sprints in parallel
Sprint 0Weeks 0–6
Incorporate all three entities, close the ~€1.4M seed, ship the AI valuation MVP + Foresta v1, and start the AIFM/legal track immediately.
Sprint 1Months 1.5–4
10–20 design-partner owners → first live auctions; Foresta live; the full team onboarded; first 10% commissions booked.
Sprint 2Months 4–7
Scale to ~1,000+ managed ha via AI onboarding; launch Foresta SaaS; fund data room + pipeline diligence-ready.
Sprint 3Months 7–10
AIFM registered; fund FIRST CLOSE (≥€2.0M); first land deployed; growth-covered-liquidity pilot settled from a harvest.
What we won't pretend compresses
- Bank of Latvia AIFM decision (~1 month) — can't be compressed, so we file day one and absorb it inside the sprint.
- Harvest seasonality + owner trust — accelerated by a named forester + co-op/partner channels, not waved away.
- Tree biology — speed grows the BUSINESS (AUM, fees, data moat) faster; the per-ha forest return stays a mid-single-digit 5–7% net.
- Higher burn + more dilution — the ~€1.4M must be in the bank before we sprint. That's the honest trade for 3× speed.
What kills this — red-team, owned
The five things a hostile diligence finds — and our answer.
Better we surface them than the investor. Each killer has a designed mitigation, not a hand-wave.
Risk
Adverse selection — in-kind contributors dump their worst forest.
Mitigation
Lead with mandates, not in-kind. Independent cruise valuation (€/m³ standing × verified inventory), quality haircut, right-to-reject, clawback. Keep in-kind <10% of the pool.
Risk
The appreciation engine has stalled — −22.7% off the 2023 peak, still soft in 2026.
Mitigation
Underwrite land at 0–4% nominal; rest the thesis on biological growth + harvest cash. Hold NAV at cost, don't mark up.
Risk
Fee drag vs DIY — a skeptical LP just buys parcels and skips the GP.
Mitigation
Fee on NAV (not committed) at ~1%; honest hurdle carry rarely bites; justify the GP by origination + execution + mezatirgus data the LP can't replicate.
Risk
Active alpha is unproven + key-person / local-execution risk.
Mitigation
Contract a named experienced Latvian forest-management firm with skin in the game + KPIs. Cap alpha at 1–2% and prove it. Asset-light ≠ no local capability.
Risk
Illiquidity + structural NAV discount over a 10–15yr lock.
Mitigation
Match closed-end lock to the asset; be explicit about end-of-life forced sale; price any secondary at an honest discount. Never imply daily liquidity.
The decision
Back the OpCo equity and seed the fund — or don't.
~€750kPool A — OpCo seed equity: the team + AI platform + 18–24mo runway (the venture you back)
€2.0M → €13.6MPool B — fund/LP land capital: €2.0M first close → ~€13.6M deployed by Y5 (the AUM)
OpCo = equityYou back the manager — fees + carry + SaaS — not the trees; the fund holds the land
Next step (Phase 0): incorporate the OpCo SIA, close the ~€750k seed, ship the AI valuation MVP against certified taksācija on a 25-parcel test, and get 2–3 written AIFM-setup quotes from Riga counsel — all before a euro of land capital is raised. The fund (Pool B) only opens once the mandate lane is proven.