Underwriting Methodology
How a deal becomes a VIG project.
No marketing voice. This is the real filter. Four mandatory gates. Three scenarios on every deal. Ten percent first-loss equity from VIG, committed before LP capital. Capital released against evidence, not against the calendar. No blending of realized with projected.
~100
Deals reviewed / yr
5–8
Deals closed / yr
6.5%
Selection rate
22.4%
Realized Gross IRR (3 exits)
What we underwrite
Sofia residential value-add. One market. One product.
VIG does not underwrite retail. We do not underwrite industrial. We do not underwrite outside Sofia. One strategy — Sofia residential value-add — executed by the same operating team in the same market since 2020. Concentration is deliberate. Complexity is the cost of discipline.
Land development
Plots in Sofia with missed potential — partially permitted, fragmented ownership, unoptimised parameters. Exit at visa stage or after consolidation.
Development
Residential buildings, 8–75 units, in Sofia. Executed in-house: VIDESIGN designs, VIBRIX builds. Exit at pre-occupancy or finished-unit sales.
Fix-and-flip / reconstruction
VIFLIP underwrites only when acquisition price + renovation budget sits ≥30% below After-Repair Value, measured against comparable sales in the trailing 6 months.
The four mandatory gates
A deal can fail one of four gates. When it does, it stops.
The gates are not checklists. They are blocking checks. A partner can challenge a judgement call but cannot override a gate. This is why 93.5% of reviewed deals fall out.
Gate 01
Smart Deal Filter — economics
Base-case net profitability must clear 20% on invested equity over the underwriting horizon. Entry priced against Registry Agency comparable sales, not against asking prices. Construction and finishing costs modeled parametrically against current Sofia chamber data, with the +15–18% YoY cost-rise factored into every project before contingency. Exit pricing tested against three independent valuation methods (comparable, income, residual land).
When the gate does not pass
A deal that requires a price assumption above the highest credible Sofia comparable to clear 20% — declined.
Gate 02
Three-Scenario Model — stress test
Every deal carries a Base, Bear, and Best case before it reaches Investment Committee. The Bear case applies: −15% to exit price, +20% to construction cost (within the 15% contingency budget — beyond that is a stop), 6-month delay to first occupancy, financing margin +200 bps. The Bear case must remain capital-preserving — VIG must be able to exit without investor capital loss, even if the project does not achieve target return.
When the gate does not pass
A deal whose Bear case loses LP capital is not underwriteable, regardless of how attractive the Base case looks. Declined.
Gate 03
Fortress Balance Sheet — solvency check
Before any capital is called against a new deal, VIG verifies: (i) the 6-month operating expense reserve is intact post-deployment, (ii) HQ overhead stays under 25% of trailing gross profit, (iii) no single bank holds more than 50% of total project debt across the platform, (iv) the platform-level LTV ceiling of 50% is respected. The check is the same on a €500K deal as on a €5M deal.
When the gate does not pass
A deal that would push the platform above any of the four thresholds is timed-delayed or restructured. Solvency comes before opportunity.
Gate 04
Legal Clearance — title and zoning
No unresolved title chain, encumbrance, zoning dispute, or municipal action against the plot or building. Senior Real Estate Counsel signs off before capital commitment. For development plays, the regulatory pathway (PUP, construction permit, habitability certificate) is mapped end-to-end with realistic timelines from comparable recent Sofia approvals — not the statutory minimums.
When the gate does not pass
A deal with a legal flag that cannot be resolved within the underwriting horizon — declined. Counsel veto is absolute.
Three-scenario model
A single number tells you nothing. Three numbers show you the boundary.
Every active deal is presented to investors in three scenarios simultaneously — Bear, Base, and Best. The point of the Bear case is not to look attractive. It is to show what will happen if the world picks it. A deal whose Bear case loses LP capital never reaches Investment Committee at all.
Bear
What we stress-test
- −15% to exit price/m²
- +20% construction cost (within 15% contingency)
- 6-month delay to first occupancy
- +200 bps financing margin
Capital must survive. Capital preservation before return target.
Base
Management underwriting target
- Exit at mid-market Sofia comparable
- Construction cost at current +15–18% YoY rise
- Timeline realized without slippage
- Rate regime stable
Not a promise. A target. Sits between Bear and Best.
Best
When the market helps
- +5% to exit price/m²
- Construction at the low end of cost band
- 2-month accelerated occupancy
- Rate environment eases
Possible. Not planned. Never the basis for deploying capital.
Every active project page exposes its three scenarios. Realized exits report against bank settlement records, audited independently of internal P&L.
Capital protection
What sits between the model and LP capital.
Four protection layers, each structural rather than discretionary. Discretion can disappear. Structure stays.
SPV isolation
Every project sits in its own Bulgarian special purpose vehicle. No cross-collateralisation between projects. Contagion stops at the SPV wall.
How it works in practice
A loss event on one project does not transmit to capital deployed in another. Investor capital is committed at the SPV level, not at the VIG holding level.
First-loss equity
VIG commits ≥10% of each project’s equity from its own retained earnings, in cash, before LP capital is called. VIG capital absorbs the first slug of any mark-down before LP capital is touched. Same waterfall, different position — VIG is paid last.
How it works in practice
On Bear-case execution, VIG’s 10% layer is written down before LP capital takes any loss. On Base-case execution, LP capital returns first; VIG’s 10% returns last, then carry.
Milestone-controlled capital release
Capital releases tie to project milestones — title deed transfer, construction permit issuance, habitability certificate, exit notarisation — never to the calendar. A draw cannot occur until evidence of milestone completion is in counsel’s hands.
How it works in practice
Investor capital sits in escrow between draws. Money moves when evidence moves. The project administrator (independent of VIG investment decisions under the AIF wrapper) signs off on each release.
Realized vs projected accounting
Realized returns are reported only after exit, audited against bank settlement records. Projected returns are management underwriting targets, disclosed as such. The two are never blended. Active-project performance is shown alongside Bear / Base / Best scenarios so an LP can see both the management target and the downside scenario simultaneously.
How it works in practice
Every page that quotes a number stamps "realized" or "projected" next to it. Every project page on the site exposes the three scenarios. The audit firm verifies realized figures against bank settlement records, not against internal P&L only.
The process
Five stages. Roughly five weeks. Five to eight closes per year.
Discipline lives in the funnel, not in the noise. Slow at the top. Fast at the bottom. Never the other way around.
01 — Source
Origination Desk · ~100 deals / yr
VISTATE (land acquisition · deal origination)
Broker network, off-market intros, distressed lists, developer JV inquiries. Top of funnel: roughly 100 opportunities reviewed per calendar year.
02 — Screen
Initial Screen · 48-hour SLA
Acquisitions
Location, condition, seller motivation, deal economics inside the strategy band. ~30–40 deals per year clear the screen. ~65% volume cut.
03 — Underwrite
Preliminary UW · 5 days
Underwriting + Head of Finance
Two-iteration model: Base and Bear IRR, MoIC, hold period; light tech DD; exit-buyer mapping. ~15–20 deals per year reach this gate.
04 — Diligence
Deep DD + Legal · 3–4 weeks
DD team + Senior Counsel
Full DD: legal title, structural, environmental, planning; market validation; exit scenarios with sensitivity tables. ~8–10 deals per year reach IC pre-read.
05 — Close
Executed Deals · 5–8 / yr
Investment Committee + Closing Counsel
IC approval; SPA negotiation; SPV-isolated closing; KYC and regulatory checks complete. 5–8 deals close per year. Deal selection rate: 6.5%.
Why the rate matters
6.5% selection rate is below the European value-add median of ~9%.
The narrow window is deliberate. Sofia residential, off-market and distressed, with hard exit-buyer mapping in underwriting. Most underperformance in regional value-add comes from saying yes too early. The funnel exists to keep us slow at the top and fast at the bottom.
See the Deal Selection DisciplinePerformance accounting
Realized vs projected. Never blended.
Realized (3 exits)
- Lyulin — Q3 2022 → Q1 2025 · 2.5 yr · Gross IRR 50.2% · Net IRR 41.8% · 2.14×
- My Campus Studios — Q2 2021 → Q1 2024 · 2.7 yr · Gross IRR 34.1% · Net IRR 27.3% · 1.68×
- Reconstruction Portfolio — Q1 2022 → Q4 2024 · 2.8 yr · Gross IRR 52.1% · Net IRR 43.6% · 1.87×
3 exits · 14,050 m² delivered · €2.05M aggregate profit · 100% capital returned. Realized weighted Gross IRR: 22.4%. Audited against bank settlement records.
Projected (7 active)
Arcadia · Odrin 88 · Dragalevtsi · Estera · Detelina · VIGARDEN · Sofia View. Target ROI range 20–42% per project (management underwriting). €10.1M active capital. €7.0M underwritten profit.
Platform-level projected net IRR to LPs: 12–18% after fees, before tax. Projection is not a guarantee.
Active figures are management underwriting targets, not guarantees. Every project page exposes the Bear, Base, and Best scenarios in parallel.
What this methodology does not claim
This does not remove risk. It makes risk visible.
No methodology prevents a market correction that hits Sofia residential alongside the rest of CEE. No discipline can make a closed-end illiquid investment liquid during the term. No audit can guarantee future performance. Three realized exits is a limited sample. What this methodology does guarantee is that every assumption is documented, every scenario is published, every gate can be verified.
Full risk register at /legal/risk-disclosure.