RIT Capital Partners plc (RCP) — Research Note

Executive summary

RIT Capital Partners is a FTSE 250-listed, self-managed multi-asset investment trust (est. 1971) that invests globally across quoted equities, private investments and uncorrelated strategies, with a permanent-capital structural advantage. Over the covered period (2021–2026H1) the trust has swung from a difficult 2022 (-8.8% NAV) through subdued 2023 (+3.2%), stronger 2024 (+9.4%), a very strong 2025 (+13.5%) and a further +9.0% in H1 2026, taking NAV per share to a record 3,159p diluted. The single most important valuation point today is the ~17% discount to NAV — the shares (2,615p) trade well below both current NAV (3,159p) and the £26.85 (85% of NAV) July 2026 tender price, which explicitly defined the Board's floor.

Fair value estimate

Methodology: NAV per share less a "warranted" discount, cross-checked against the tender price and buyback programme.
Post-tender diluted NAV: broadly unchanged from 3,159p — the tender at 85% of NAV is accretive to continuing shareholders (~+1.3% per the filing).
Central assumption: a "fair" discount for a diversified multi-asset trust with ~34% Level-3 private book of 8–15% (peers Caledonia, Personal Assets and Ruffer sit in a similar band; RIT's historical average discount is roughly in the mid-teens but recently wider).
Fair value range per share: 2,685p – 2,905p (NAV 3,159p × 85–92%).
Implied market-cap range: £3,325m – £3,595m (using 123.8m post-tender shares).
Mid-point mcap: ~£3,460m vs. current £3,176.7m.
Absolute upside to mid: ~+9% (range: +3% to +11%).
The tender at 85% NAV effectively puts a floor near 2,685p; buybacks continue at wider discounts.

Sector context

Classification: Financials / Financial Services — specifically a closed-end investment trust (AIC "Flexible Investment" peer group).
Quality/growth/leverage vs. peers: Above-average on portfolio quality (permanent capital, access to Anthropic/Databricks/SpaceX/Stripe pre-IPO); in line on growth (~10–11% p.a. NAV since inception); below-average on leverage (5.5% gearing vs. sector norm 5–15%).
Listed peers: Caledonia Investments (CLDN), Personal Assets Trust (PNL), Ruffer Investment Company (RICA); more direct private-equity-tilted comps include Scottish Mortgage (SMT) and HgCapital Trust (HGT).

Investment thesis (3 bullets)

Genuine, differentiated pre-IPO AI exposure: The private book holds Anthropic, Databricks, SpaceX (transferred to Quoted post-IPO), Stripe and Cognition, all added or scaled in 2025–H1 2026, giving one of the very few UK-listed routes into these names — H1 2026 SpaceX alone produced an unrealised gain of £110m (4.0× cost) 2026-08 half-year.
Persistent, structurally wide discount now backstopped by explicit capital return policy: Since 2023 the Board has repurchased >13% of share capital (~£392m) plus a further £300m tender in July 2026 at 85% NAV, with a dividend review promising a step-up from 2027 2026-08 half-year; 2026-03 final.
Realisation track record validates private-book marks: Over 2 years >43% of the private portfolio realised, direct realisations at aggregate 70% above prior carrying value — a strong empirical rebuttal to the "PE valuations are optimistic" concern that has driven the discount 2026-08 half-year.

Key risks (3 bullets)

Discount could stay wide or widen: Despite tender + buybacks the discount was still 27.5% at 30 June 2026 (vs. ~10% target); if AI/private sentiment reverses, share price could underperform NAV materially 2026-08 half-year.
Level 3 valuation risk: £1,940m (45% of NAV) is Level 3, valued using GP marks/models; a 5% haircut to the fund book alone is £73m/1.7% of NAV, and public-market AI multiple compression would flow through with a lag 2026-08 half-year notes.
NAV persistently lags the ACWI in narrow, US mega-cap-led rallies — 5-yr NAV total return 25.7% vs. ACWI 74.3%; if the "narrow tech rally" continues, relative underperformance persists 2026-08 half-year performance history.

Operating leverage

Operating leverage is structurally low for this business model. Costs are dominated by fixed management/operating expenses (~£41m in FY25, OCF 0.73% of avg NAV) plus interest on ~£361m of drawn borrowings. Revenue (investment income + capital gains) is entirely driven by portfolio moves, so "incremental revenue" doesn't scale disproportionately — a 10–20% NAV beat translates roughly 1:1 to shareholder returns, not into a multiple of profit. The one place operating leverage exists is buyback accretion: repurchasing at a 27% discount is mechanically accretive (~0.5–0.9% NAV lift per year historically), and the tender added a further ~1.3% 2026-08 half-year; 2026-03 final. But this is a small tailwind, not a compounding earnings multiplier. There is no fixed-cost inflection point comparable to a software or capacity-constrained business.

Value-trap signals

None identified. NAV is at an all-time high; realisation prices consistently exceed carrying values; balance sheet is strong (£227m cash, £361m drawn debt, only 5.5% gearing); dividend has been raised for 13 consecutive years; buyback + tender demonstrate Board discipline. The wide discount reflects sector sentiment and private-asset scepticism, not deteriorating fundamentals.

Earnings vs. expectations

As a closed-end trust RIT does not issue EPS guidance or attract sell-side consensus in the conventional sense; performance is judged vs. CPI+3% and ACWI (50% £). The pattern across the period: materially beat CPI+3% in H2 2024, 2025, and H1 2026; beat ACWI in 2022 (much smaller loss) and 2023 (bounced back less); lagged ACWI meaningfully in 2024 (+9.4% vs +20.1%), 2025 (+13.5% vs +17.1%) and H1 2026 (+9.0% vs +12.5%), a pattern management attributes to deliberate underweight of US mega-cap tech. Net: beats the inflation hurdle consistently, lags a narrow-tech-led ACWI — this is by design.

Conviction

4 — high. The valuation methodology (NAV less discount) is unambiguous for an investment trust; NAV is audited and disclosed monthly; the tender price provides an independent third-party price point; and realisation prices vs. carrying values give empirical support to the marks. Limiting factor: the 45% Level-3 tail means a ~5–10% haircut to private marks would compress fair-value estimates materially, and the appropriate "warranted" discount is judgement-based rather than mechanically derivable.

Driver commentary

AI beneficiary (55): Real but indirect — Anthropic/Databricks/SpaceX/Stripe/Cognition are genuine AI-value-chain holdings, but wrapped in a diversified trust means <10% of NAV is directly AI-linked.
Operating leverage (25): Trust structure has no meaningful operating leverage; revenue = portfolio returns, costs are near-fixed but small.
Cyclicality (40): Diversified, hedged, historically captured 74% of upside / 40–41% of downside — moderate cyclicality by construction.
Moat (65): Network access to top-tier private deals is a real competitive advantage; permanent capital structure and Rothschild brand support long-run sourcing.
Leverage (18): Fortress balance sheet, 5.5% gearing, ample liquidity.
Earnings quality (60): Clean IFRS, external audit, Valuation Committee — but 45% Level 3 requires judgement; realisations above carrying values are reassuring.
Management quality (70): New CEO Fanari has driven improved communications, active buyback + tender, disciplined asset allocation; Board acted decisively on the discount.
Growth momentum (68): NAV +19.7% over LTM to Jun-26, +9.0% H1 2026 — accelerating, driven by private realisations.

Overall score — 545

A partial fit. The AI exposure is genuine (unusual pre-IPO access to Anthropic/Databricks/SpaceX/Stripe), the valuation is fair-to-cheap (~17% NAV discount, tender price floor), and downside protection is excellent (fortress balance sheet, permanent capital, diversified). But the fundamental mismatch with this investor's profile is the near-total absence of operating leverage — the trust structure cannot convert a revenue beat into a multiple of profit, which is exactly what the investor wants from AI-cycle upside. Score reflects "interesting hybrid but doesn't capture the full thesis."