CHRY — CHRYSALIS INVESTMENTS LIMITED — Research Note

Executive summary

Chrysalis is a Guernsey-registered, LSE-listed closed-ended investment company holding late-stage private growth-capital positions — dominated by Starling Bank (56.5% NAV), Smart Pension (21.6%), Klarna (9%) and wefox (5.7%). Since March 2026 the Company has been in explicit orderly-realisation mode with a three-year horizon to February 2029, and on 21 August 2026 it transitioned to a self-managed model with a target run-rate cost of ~£2m/yr, down from £4.6m 2026-06-30 half-year. The single most important valuation fact is the 44% discount of the 73.4p share price to the last published NAV per share of 131.09p, which the Chairman explicitly notes is "more than covered" by the carrying value of Starling alone 2026-06-30 half-year.

Fair value estimate

Methodology: probability-weighted NAV realisation, discounted at a required IRR.

Anchor: NAV/share of 131.09p at 30 June 2026 (£630.5m NAV; 481.0m shares).

Assumptions for a 3-year orderly realisation (to Feb 2029):

Base case (60% weight): realisations achieve ~85% of carried NAV net of costs. Starling exit near current mark (56.5% of NAV) is the swing factor; Smart Pension and Klarna have plausible paths to par or better; wefox and residual assets discounted heavily. → 111p gross of running costs; ~108p net after ~£6m of forecast costs.
Bull case (20%): near-full NAV realisation (~95%), Klarna re-rating and clean Starling exit → ~124p.
Bear case (20%): ~65% realisation, valuation gaps at Starling, further wefox write-down → ~85p.

Weighted terminal proceeds ≈ 105p. Discounted at 12% p.a. over ~2.5 yrs = ~79p. At 10% p.a. = ~83p. At 15% p.a. = ~73p.

Fair value range: 78 – 95p per share (mid ~86p).
Implied market cap range: £375m – £457m (mid ~£414m).
Vs current £380m: 6% – 30% upside; mid ~+17%.

The wide range reflects the fact that this is essentially a levered play on Starling's exit value.

Sector context

Correctly classified as Financials / Financial Services (closed-ended investment company / listed private growth vehicle). Compared with typical listed private-growth peers, Chrysalis's portfolio quality is broadly in line (mature late-stage fintechs), its balance sheet is now cleaner (debt fully repaid June 2026, ~£58m total liquidity dominated by Klarna listed stock) but its concentration risk is materially higher than peers. Closest listed comparables: Molten Ventures (GROW), Augmentum Fintech (AUGM), Schiehallion Fund (MNTN). All trade at material NAV discounts; Chrysalis's ~44% is at the wide end but justifiable given portfolio concentration and asset-liquidation timeline risk.

Investment thesis (3 bullets)

Deep discount to a NAV that has been repeatedly written down and stress-tested. NAV has already dropped from 171.65p (Sep 2025) to 131.09p (Jun 2026), predominantly on peer-multiple contraction rather than operational deterioration — Starling posted its fifth consecutive year of profitability (PBT £217m) and Smart Pension's Master Trust exceeded £10bn AuM 2026-08-03 quarterly. Buying at 73.4p offers ~44% margin to a NAV that already reflects a 22% haircut this year.
Structural forced-return mechanism. The Board is contractually committed to an orderly realisation over three years with net proceeds returned to shareholders "as efficiently as possible" 2026-08-03 quarterly. There is no continuation vote until Feb 2029 and no new investments permitted. This is a defined-outcome situation, not an open-ended growth fund.
Starling optionality. Starling alone is carried at £356m (~56% of NAV, or 74p/share on its own vs. 73.4p share price), and has £525m surplus capital plus a scaling BaaS business (Engine, £70m committed ARR of a targeted £100m by 2027) 2026-08-03 quarterly. If Starling can access capital markets or trade sale near its carrying value, that alone effectively backstops the current share price.

Key risks (3 bullets)

Concentration risk in Starling. With Starling at 56.5% of NAV, a ~20% write-down in Starling would reduce NAV by ~15p (11%). The revised valuation methodology explicitly acknowledges "the potential volatility of a highly concentrated portfolio is uppermost in our minds" 2026-06-30 half-year.
Private-company valuation risk demonstrated repeatedly. The recent history is unfriendly: Deep Instinct fully written down; wefox valuation experienced a NAV-adjustment error of 3.33p/share (a control failure flagged by the auditor); Klarna's shares fell 63% in H1 2026 to $13.09 2026-06-30 half-year. Realisation proceeds could easily undershoot carrying values.
wefox waterfall / capital-structure risk. wefox required €7m of follow-on funding in April 2026 and has a "complex capital structure and outstanding debt" that could impair the Company's economics in an exit scenario 2026-06-30 half-year. The Chairman explicitly warns wefox "will require careful management to ensure that full value is realised".

Operating leverage

At the Company level, operating leverage is essentially irrelevant — this is a closed-ended fund with ~£2m/yr of fixed running costs against ~£630m of NAV. The cost base is effectively fully fixed and the halving of the annual expense from £4.6m to £2m (post 21 August 2026) is a one-off structural change already announced 2026-06-30 half-year. Portfolio-company operating leverage is real but second-order to CHRY holders: Klarna's incremental economics are compelling (44% revenue growth vs. 3% non-transaction opex growth in Q1 2026, and Fair Financing loan cohorts maturing) and Starling's Engine SaaS could add valuation multiple expansion. But the Company itself does not directly compound these — it simply captures value at exit. A 10-20% revenue beat at the portfolio-company level does not translate into "multiples of profit" for CHRY; it translates into a modest uplift in exit proceeds, offset by discount-widening/narrowing dynamics. This is fundamentally the wrong vehicle for an operating-leverage investor.

Value-trap signals

Wind-down structure. By design, the vehicle is shrinking, not compounding. NAV trajectory over three years is bounded above by the carrying value.
Valuation control failure. The wefox waterfall error (3.33p/share adjustment identified by the auditor) is a meaningful red flag on valuation governance.
Track record of NAV disappointment. NAV per share peaked at 251.96p in Sep 2021; it is now 131.09p. Five years of persistent write-downs vs original expectations.
Concentration in one asset whose peer group is de-rating. Starling is being marked against a UK/European banking peer group that has itself been volatile.
wefox capital-structure complexity. Complex waterfall + outstanding debt means the equity waterfall can materially disadvantage minority shareholders.

Earnings vs expectations

Not a conventional earnings company. The relevant metric is NAV per share vs. prior-quarter NAV and market expectations. Over the last eight quarters NAV has: risen (Q1 26 vs Q4 25: 121.20p→105.00p was -13%; actually a fall), then fell sharply Q1-Q2 2026 (from 165.36p to 137.27p at 31 Mar), then was adjusted downward by 3.33p for the wefox error, then declined again to 131.09p at 30 Jun. Pattern: persistent negative NAV surprises through the tightening cycle, driven by peer-multiple compression and specific asset write-downs (Klarna, wefox, Deep Instinct). Not a "beat" story.

Conviction

Conviction: 3 (moderate).

Anchors: (i) NAV is externally-valued with KPMG audit involvement and is now anchored by a listed Klarna position (£56.9m as of Jun 2026); (ii) the orderly-realisation timeline is contractually defined with clear governance; (iii) the ~44% discount provides substantial margin of safety on a range of realistic outcomes.
Caveats: (i) NAV is dominated by Starling, whose exit value is not observable and could vary by ±30%; (ii) the Company's own valuation controls have recently failed (wefox waterfall error), reducing confidence in mark accuracy.

Driver scoring rationale

AI beneficiary (25): Starling deploys AI internally (Starling Assistant, Scam Intelligence) but is a user, not a seller of AI. Klarna similarly. Deep Instinct (the one direct AI-security holding) has been written down to zero. Brandtech has GenAI marketing exposure but is only 4.3% of NAV. This is not an AI-receiver vehicle.
Operating leverage (20): At the CHRY level, ~£2m/yr of fixed cost against ~£630m NAV — effectively zero operating leverage at the fund level. Portfolio has some, but doesn't compound to shareholders.
Earnings surprise trend (30): Sustained NAV disappointments over the last eight quarters vs implicit expectations set by prior valuations.
Cyclicality (55): Portfolio is fintech-heavy (Starling banking sensitive to rates; Klarna consumer credit; wefox insurance). Moderately cyclical.
Moat (40): Individual assets have moats (Starling's tech, Smart Pension's platform) but CHRY as a vehicle is a passive holder with no ability to add value beyond stewardship.
Leverage (10): Net cash after Barclays facility repaid in June 2026. Fortress balance sheet at the Company level.
Earnings quality (40): Level 3 unobservable NAV; recent valuation error acknowledged.
Management quality (55): Board is competent; transition to self-managed model executed cleanly; buyback programme returned £117m at attractive discount. Some concerns on past wefox governance and audit-flagged error.
Growth momentum (30): NAV trajectory is negative; portfolio revenue growth continues but exits are the driver, and pace is uncertain.

Overall score rationale

Score: 340 / 1000.

This is a genuinely cheap vehicle (44% discount, backed by a listed Klarna position and a substantial carrying value in Starling) with a defined-outcome three-year realisation path. But it is fundamentally the wrong strategic fit for this investor: no meaningful AI-receiver exposure, no operating leverage to shareholders (fund-level structure), and the "value" case depends on Starling's private-market exit going well. Attractive as a special-situations play; not a fit for the AI-receiver / operating-leverage / valuation-discipline framework. Rated in the "partial fit" band, near its lower boundary.