Overview
To: Investment Committee / Private Markets Allocation Team
From: Michael Hartung, Senior Advisor, Fund Structuring
Date: October 2026
Re: Evaluating co-investment and SMA architectures alongside fund commitments
Classification: Commentary — not a term sheet or offering document
Purpose
Large institutional LPs — German pensions, Nordic insurers, Benelux sovereign pools — increasingly treat co-invest access as part of the fund commitment, not a discretionary favour. This memo summarises structural choices, fee norms, governance interfaces, and operational pitfalls we observe on European infrastructure programmes, including co-invest sleeves alongside a €295M Luxembourg core+ programme (first close €175M, six Article 8 LPs) and cross-border assets such as a Southern European port concession SPV (€88M, ECA-covered senior debt, 2024 close).
Use it when negotiating anchor terms or reviewing a GP's co-invest programme for fairness, not when substituting your own legal or tax advice.
Co-invest vs SMA — different tools
Co-investment — LP participates in single assets alongside the fund; typically no management fee on co-invested capital; carry at fund rate or discounted (50–100% discount is common on European infra). Suited to episodic large tickets without building a direct team.
Separately managed account (SMA) — Custom portfolio for one LP; full discretion within guidelines; higher setup and reporting cost. Typical for sovereign wealth and mega-pensions above €500M–1B programme sizes.
Both require allocation policies when deal flow exceeds capacity. Co-invest suits most pensions; SMAs increase operational complexity because discretion spans portfolio construction, not single-asset elections.
Allocation policy — what "fair" looks like in 2026
Written policies should specify:
- Eligibility (fund LPs only vs broader relationships)
- Minimum fund commitment threshold (often €25–50M for infra co-invest)
- Pro-rata vs discretionary allocation
- Capacity when deal flow exceeds co-invest room (max simultaneous syndications, annual cap per LP)
- Right of first offer timing relative to fund investment period
Pro-rata by commitment size remains market standard. Discretionary invites require LPAC-approved criteria to avoid favouritism toward strategic relationships.
Stapled co-invest — committing co-invest capacity alongside fund subscription — accelerates fundraising but concentrates risk. Anchor letters should set minimum and maximum stapling bounds so GP discretion stays within LPAC policy during volatile deployment periods. We have seen unstapled anchors receive no co-invest for eighteen months while staples consumed capacity — both create LPAC friction.
LPAC should review allocation statistics quarterly when discretion applies: deals offered, accepted, declined, average time-to-close. Annual reporting of these metrics supports successor fund marketing and reduces pro-rata disputes.
Fee lite does not mean diligence lite
Co-investors run full technical, legal, and ESG diligence parallel to GP workstreams. Fee lite is an economics choice, not a process shortcut. Pensions with fiduciary documentation requirements reject shortcut memos — wasted syndication effort.
Co-invest memos should mirror fund IC standards: independent technical reports, model audit summaries, taxonomy alignment evidence, red-flag logs. Standardised data room templates and SPV packages pre-approved by counsel shorten time from allocation offer to commitment (four to eight weeks is achievable; twelve+ without templates).
GP revenue modelling matters: fee-free co-invest on large tickets reduces management fee base. Sponsors should model GP economics at maximum co-invest utilisation before marketing — we benchmark carry discounts against peer Munich and Luxembourg GPs to support LP negotiations.
Broken deal cost allocation between fund and co-invest sleeves must be signed before pipeline marketing. Abort-fee disputes destroy anchor relationships.
Governance and conflicts
LPAC approval required for affiliated co-invest where fund LP participation differs. Valuation and exit processes should mirror fund assets to prevent NAV arbitrage. Co-investors may seek board representation via LP corporate vehicles — conflict review when GP appoints directors across cap tables.
Exit coordination: drag-along, tag-along, ROFR aligned with fund exit timelines. Continuation fund and secondary processes need separate co-invest LP consent when direct asset stakes exist outside fund interests.
Tax, regulatory, and cross-border closing
Cross-border co-invest into multi-jurisdiction European infrastructure triggers withholding analysis, permanent establishment review, and AIFMD reporting where discretion is delegated. Luxembourg SPV templates accelerate closing but require per-asset tax review — a leading cause of missed co-invest closes alongside fund closings is withholding clearance delay (two to six weeks per domicile).
European pension and sovereign LPs increasingly require co-invest access letters at fund commitment. Define eligibility, pro-rata mechanics, and discretionary criteria before anchor negotiation — retrofitting after first close generates MFN and fairness disputes.
Budget external counsel hours per co-invest in GP operating plans, not only fund OPEX. Fee-lite sleeves still need full legal, tax, and regulatory review per investor.
Fee structures and GP sustainability
Market practice on European infrastructure co-invest: no management fee on co-invested capital; carried interest at fund rate or 50–100% discount; broken deal costs shared per LPAC-approved policy. Some GPs charge nominal monitoring fees — increasingly resisted by institutional LPs expecting fee-lite co-invest as anchor relationship benefit.
GPs should disclose co-invest carry and fee terms in fund PPM appendices so fund LPs understand economics relative to co-invest participants. When fee-free co-invest consumes 30–40% of deployed capital, fund management fee and carry must compensate adequately — we model this with sponsors before anchor marketing.
Sourcing fees from portfolio companies face heavy LP scrutiny. Transparency in quarterly reporting prevents LPAC escalation that damages successor fund processes.
European LP expectations on access letters
Large German, Nordic, and Benelux institutional LPs increasingly treat co-invest access as a contractual entitlement tied to fund commitment size. Eligibility thresholds, pro-rata mechanics, and discretionary deal criteria should be defined before anchor negotiation — retrofitting after first close generates MFN and fairness disputes.
Co-invest diligence parity with the main fund is non-negotiable for pensions with fiduciary documentation requirements. Operational co-invest teams should mirror fund deal teams on valuation, ESG, and exit processes to prevent NAV arbitrage allegations.
Non-fund LPs accessing co-invest — strategic relationships only in some programmes — typically require LPAC approval as a conflict matter. Disclosure upfront avoids surprise at quarterly review.
SMA-specific note
SMA guidelines need bespoke reporting cadences but should reuse fund-level ESG and risk templates where possible to control cost. Cross-border SMA mandates may trigger AIFMD considerations when discretion is delegated — structure as managed account vs fund commitment per regulatory counsel.
Munich managers serving DACH LPs often batch co-invest closes quarterly to manage operational load — ask about operational capacity, not only allocation policy text. SMA setup costs include separate SPV templates, compliance monitoring, and dedicated reporting — budget €400–800k annually for mid-complexity European SMA programmes in our experience.
Fund-of-one structures differ from SMAs legally: a fund-of-one is still a fund with a single LP — tax and regulatory treatment differs from managed accounts. Counsel should opine before LP signs.
Programme launch — sponsor actions
Draft co-invest memo template mirroring fund IC format before first anchor. Pre-approve SPV and tax structure with counsel. Cap deals in syndication simultaneously. Publish LPAC-reviewed allocation policy. Model GP revenue at maximum fee-free co-invest utilisation. Run mock co-invest close with operations to surface bottlenecks.
Track co-invest conversion rates and time-to-close; LPAC reviews quarterly for allocation fairness. Stapled co-invest percentages in anchor letters should include minimum and maximum bounds.
Diligence questions for GP meetings
- Is allocation policy LPAC-approved and published before first close?
- Do co-invest memos include the same independent reports as fund IC packs?
- How are stapled percentages bounded in anchor letters?
- What are last twelve months' co-invest conversion rate and time-to-close?
- How does the GP model revenue if fee-free co-invest reaches 30–40% of deployed capital?
Bottom line
Co-invest is a fundraising and retention tool — document allocation fairness before anchors sign. SMAs suit the largest allocators but multiply operational load. European LPs treat access as contractual entitlement tied to commitment size; GP programmes that treat co-invest as discretionary hospitality lose re-ups.
We have seen strong fund track records undermined by opaque co-invest allocation during continuation processes. Structure and transparency at programme design prevent reputational damage that no IRR offset clears.
Do co-investors get board seats? Sometimes via LP corporate vehicles; governance negotiated per deal. What is stapled co-invest? LP commits to co-invest a defined percentage alongside fund subscription — common in anchor negotiations; bounds should be in the anchor letter. Do co-invest LPs pay broken deal costs? Often shared per programme policy — disclose in co-invest side letter or PPM appendix before pipeline marketing.
On the €88M Southern European port concession co-invest, a pension co-investor closed six weeks after fund IC because SPV and withholding templates were pre-approved — the same GP's next co-invest without templates took eleven weeks and nearly missed construction milestone covenants. Operational readiness is not secondary to allocation policy text.
When continuation funds acquire assets where co-invest LPs hold direct stakes, consent mechanics must be documented separately from fund LP votes — we have seen secondary sales delayed eight weeks while counsel reconciled ROFR and tag-along across cap tables.
Disclaimer: Commentary only. This article reflects observed market practice and advisory experience from Consultinghouse GWB; it is not legal, tax, or investment advice. Readers should obtain independent professional counsel before acting on any structure described.
