This is investment research, not personal financial advice.
The market marked the report twice
CLS Holdings plc (LSE:CLI) closed 3.87% lower at £0.4715, or 47.15p, on 12 August after publishing its half-year accounts. The price at writing was £0.4715. The move followed a warning eight days earlier, so the result was not a first shock. It was the market's second chance to price the same problem: EPRA net tangible assets fell 11.5% in six months to 177.7p a share, group vacancy reached 14.5%, and the statutory loss widened to £69.6 million (CLS 2026a; LSE 2026).
The 47.15p close sits at only 26.5% of reported EPRA NTA. That sounds like a wide margin between price and appraisal. Yet the cash record explains why the discount exists. H1 cash generated from operations was £18.4 million. Cash interest was £18.6 million. Property capital expenditure then absorbed £10.3 million, before the £7.3 million dividend paid in the half (CLS 2026a).
The market reaction looks roughly proportionate to the new evidence. A 3.9% fall did not assume that the entire portfolio is impaired. It recognised that another 11.5% NTA decline and rising vacancy made an already large discount harder to dismiss as sentiment. CLS has already completed plenty of disposals. The question is whether the last part of its 2026 programme can validate carrying values and leave a smaller estate that produces cash after interest and refurbishment.
One correction matters before any valuation work. The often-quoted £300 million disposal figure is cumulative. CLS reported or targeted about £67 million of sales in 2024, £144 million in 2025 and £100 million in 2026. By June 2026, roughly £56.8 million of the current-year target had completed. Only about £43 million remained against that target, not £300 million of fresh proceeds (CLS 2024; CLS 2025; CLS 2026a).
Debt fell while leverage refused to follow
The basic economics are those of a leveraged office landlord. CLS owns 76 properties across the UK, Germany and France. Tenants pay contracted rent; CLS pays property costs, staff, interest, tax, refurbishment and leasing incentives. Asset sales can repay debt, but they also remove rent and may crystallise a discount to the valuation book.
That distinction is visible across four years. Net debt fell from £997.3 million at December 2022 to about £808.3 million at June 2026. Balance-sheet loan-to-value still rose from 42.2% to 50.7%. The numerator improved, but the property denominator fell faster. EPRA NTA per share dropped from about 316.5p to 177.7p over the same span, while occupancy fell from 92.6% to 85.5% (CLS 2022; CLS 2026a).
| Period | EPRA earnings | EPRA EPS | EPRA NTA/share | LTV | Occupancy | Net debt | Interest cover |
|---|---|---|---|---|---|---|---|
| FY2022 | £46.8m | 11.8p | 316.5p | 42.2% | 92.6% | £997.3m | 2.98x |
| FY2023 | £36.5m | 9.2p | 253.0p | 48.5% | 89.0% | about £1.0bn | 2.23x |
| FY2024 | £36.7m | 9.2p | 215.0p | 50.7% | 87.3% | £938.7m | 1.91x |
| FY2025 | £8.9m | 2.2p | 200.7p | 50.0% | 87.2% | £852.5m | 1.80x |
| 1H 2026 | £10.9m | 2.7p | 177.7p | 50.7% | 85.5% | about £808.3m | about 1.70x |
The table uses reported EPRA earnings and NTA. Occupancy is the complement of disclosed EPRA vacancy where the reports present vacancy rather than occupancy. The interim row covers six months and is not directly comparable with full-year earnings. FY2025 also carries a sharp earnings reset that should not be smoothed away. The history shows a balance sheet losing asset cover even as management pays debt down (CLS 2022; CLS 2023; CLS 2024; CLS 2025; CLS 2026a).
This is the central operating loop, and at present it runs backwards. Higher vacancy lowers rent. Empty buildings need spending and incentives before rent returns. Lower income and higher yields reduce appraised values. Lower values raise LTV, which then makes sales and refinancing more urgent. Each disposal improves debt in pounds but reduces the rental base available to service what remains.
The £9.5 million rent gap has a price
The portfolio is geographically diversified, but the weak spots are not evenly spread. Germany accounts for £725.2 million, or 46.3%, of H1 property value. The UK contributes £628.2 million and France £212.5 million. Germany also has the longest weighted average unexpired lease term at 6.0 years. France is 5.0 years. The UK is only 3.5 years (CLS 2026a; CLS 2026b).
| Region | Property value | EPRA vacancy | WAULT | H1 valuation change | Contracted rent | ERV |
|---|---|---|---|---|---|---|
| UK | £628.2m | 17.7% | 3.5 years | -7.2% | £46.1m | £53.1m |
| Germany | £725.2m | 12.5% | 6.0 years | -2.5% | £41.0m | £42.8m |
| France | £212.5m | 7.7% | 5.0 years | -3.7% | £13.3m | £14.0m |
| Group | £1,565.9m | 14.5% | 4.7 years | -4.6% | £100.4m | £109.9m |
The £9.5 million difference between estimated rental value and contracted rent is not idle cash waiting to be collected. It is an operating project. Space has to be refurbished, marketed, fitted out and leased. Incentives can defer the cash benefit after a lease is signed. In the UK, where vacancy is highest and leases shortest, the group suffered its largest valuation decline in the half.
Government and major-corporate tenants offer useful covenant quality. They do not eliminate expiry risk or make vacant floors productive. Nor does the three-country footprint create a network effect. Tenants in Munich do not make London space easier to let. The advantage is risk distribution and local expertise, not a self-reinforcing moat. The numbers now classify that advantage as eroding: vacancy has almost doubled from 7.4% in 2022 to 14.5%.
The anti-thesis deserves equal weight. Germany and France have longer leases and lower vacancy than the UK. Contracted rent remains close to ERV in both markets. Large-tenant exposure can protect collections during a weak cycle. If letting costs turn vacant buildings into income-producing space, today's rent gap could lift earnings without an acquisition. But that case requires cash before it produces cash, precisely when interest is consuming the operating inflow.
The owner-cash bridge is below zero
EPRA earnings are designed to remove property revaluation noise and other items from statutory profit. They are useful for comparing recurring property earnings. They are not the same as owner cash after maintaining and leasing the estate.
The H1 bridge begins with £18.4 million generated from operations. Add £0.3 million of interest received, subtract £18.6 million of interest paid and £0.1 million of tax, and net operating cash is approximately zero. Subtract £10.3 million of property capex and post-capex operating cash is about negative £10.3 million. The £7.3 million dividend takes the pre-disposal owner-cash deficit to roughly £17.6 million (CLS 2026a).
| H1 2026 cash item | £m |
|---|---|
| Cash generated from operations | 18.4 |
| Interest received | 0.3 |
| Interest paid | (18.6) |
| Tax paid | (0.1) |
| Net operating cash | about 0.0 |
| Property capex | (10.3) |
| Post-capex operating cash, author calculation | about (10.3) |
| Dividends paid | (7.3) |
| Pre-disposal owner cash, author calculation | about (17.6) |
Asset sales supplied £56.6 million of cash. CLS repaid £113.5 million of debt, drew £62.3 million of new borrowing and ended the half with £35.3 million of cash, down from £49.4 million. The sequence is not evidence that disposals failed. They helped refinance and repay debt. It does show who received the immediate benefit: lenders and balance-sheet liquidity, not equity owners.
FY2025 tells the same story over a full year. Operating cash after interest and tax was £14.6 million. Property capex was £17.3 million. That left a £2.7 million deficit before £15.9 million of dividends. Property-sale proceeds were £136.7 million, while debt repayments reached £186.5 million (CLS 2025).
The relevant formula is therefore wider than EPRA earnings:
owner cash = operating cash - cash interest - tax - maintenance and leasing capex - recurring corporate costs
Then add net sale proceeds and subtract debt repayment, sale taxes, transaction costs and distributions. On that measure, CLS has not yet shown a recurring surplus. For a landlord with 14.5% vacancy, treating all property capex as optional would overstate owner economics.
The refinancing ratchet has not finished
CLS entered 2025 with £373.7 million across 11 facilities due during that year. It refinanced, extended or repaid those facilities. The price was visible: £222.8 million of new loans carried a 5.5% weighted average all-in rate, compared with a 3.8% average cost across group debt at year-end. Only £51.6 million of those new loans was fixed, at 4.1% (CLS 2025).
The group began 2026 with £145.5 million of long-term debt maturing, £11.8 million of amortisation and £42 million of short-term facilities expiring. Weighted average debt maturity was 3.6 years. Undrawn committed facilities were £28 million, with another £10 million uncommitted. Only £64.4 million of property was unencumbered at FY2025 (CLS 2025).
That creates a lag. Falling policy rates would help new funding, but legacy low-cost facilities still refinance against market spreads, property-level collateral and weaker valuations. H1 gross debt was approximately £844 million when the reported net debt and cash are combined. Even after disposal-led repayments, interest cover was around 1.7 times.
The balance-sheet target is 35% to 45% LTV. Reaching it through debt repayment is constructive. Reaching it because the group repeatedly sells the best assets while weaker buildings remain would produce a smaller but not necessarily safer earnings base. The evidence needed is transaction-level: sale price against carrying value, rent lost, cash tax and costs, debt cancelled, and the interest saved.
The ownership structure adds another layer. The Sten and Karin Mortstedt Family & Charity Trust held 55.24% at FY2025, while Bengt Mortstedt held 6.55%. A family-related investment vehicle provided a £40 million revolving facility, of which £20 million was drawn (CLS 2025). Patient control can support the company through a difficult cycle. It can also allow a large market discount to persist without a transaction or capital return that closes it for minority holders.
What 47p says about the appraisal
At 47.15p, the market assigns only 26.5p to every 100p of reported EPRA NTA. Using roughly 398.1 million voting shares for the NTA calculation, H1 EPRA NTA equity is about £707 million. The LSE market page reported a £196.9 million instrument market capitalisation. The difference is roughly £510 million (CLS 2026a; LSE 2026).
If the whole gap were blamed on property overstatement, it would equal about one third of the £1.566 billion portfolio. That is a reverse valuation, not a forecast. Some discount belongs to disposal costs, tax, refinancing, vacancy capex, governance, low liquidity and the risk that a smaller portfolio earns less after sales.
The sensitivity is still stark. A 1% change in portfolio value is £15.7 million, or about 3.9p per voting share before tax and other adjustments. Another 5% portfolio decline mechanically removes about 19.7p from NTA. A 10% decline removes about 39.3p. This explains why a 47p equity price can move hard on modest changes to property yields.
| Incremental portfolio move | Mechanical EPRA NTA/share |
|---|---|
| -20% | about 99p |
| -15% | about 119p |
| -10% | about 138p |
| -5% | about 158p |
| 0% | 177.7p |
| +5% | about 197p |
The reverse case is more demanding than the headline discount suggests. Even after a further 10% portfolio haircut, mechanical NTA remains near 138p. For the price to stay near 47p, the market must also assume that a substantial portion of that residual cannot reach minority holders as cash. Continued capex deficits, expensive debt and sales below book are the obvious mechanisms.
Four paths through the same balance sheet
The scenario ranges below start with H1 EPRA NTA, apply a portfolio-value adjustment and disposal leakage, then apply a market recognition range to adjusted NAV. They are analytical ranges in pounds per share. They are not forecasts from CLS. The disposal leakage sensitivity uses a £300 million gross reference amount only to make cases comparable; it must not be read as £300 million still available after June 2026.
| Case | Main assumptions | Adjusted equity logic | Value/share range |
|---|---|---|---|
| Severe downside | values -15%; 15% sale leakage; vacancy above 15%; recurring cash deficit | market recognises 20%-45% of stressed NAV | 21.5p-48.3p |
| Bear | values -8%; sales 8% below book; interest cover below 2x | market recognises 35%-55% of adjusted NAV | 50.0p-78.5p |
| Base | values -3%; sales within 3%-5% of book; LTV below 45%; positive post-capex cash | market recognises 55%-75% of adjusted NAV | 93.5p-127.4p |
| Bull | values +3%; lower vacancy; easier refinancing; near-book sales | market recognises 75%-95% of adjusted NAV | 148.7p-188.4p |
The severe-downside range contains the current price. It describes a company where more asset value disappears, sales leak heavily against book and the equity remains valued as a thin residual beneath debt. The bear case allows deleveraging but assumes that creditors continue to capture most of the benefit.
The base case requires evidence not yet present in the H1 cash flow: sales near book, LTV inside the stated range, falling vacancy and cash left after interest and property capex. The bull case needs both operating repair and market recognition. Neither case follows automatically from a 73.5% headline discount.
Two variables dominate. First is the external property mark. Each 5% changes mechanical NTA by about 19.7p. Second is sale leakage. On a hypothetical £300 million at book, a 10% discount removes £30 million, or about 7.5p per voting share, before tax and costs. A sale can lower LTV and still destroy NTA if the price is weak enough.
The remaining £43 million is a test, not a solution
By H1, CLS had completed most of its £100 million 2026 sales target. The approximate £43 million still to complete is small next to £808 million of net debt and a £1.566 billion estate. It cannot close the valuation gap by itself. It can answer several important questions.
First, does the market clear assets within 5% of the June carrying value? Second, does at least 80% to 90% of net cash after costs reduce net debt? Third, does the interest saving exceed the net rent sold? Fourth, does year-end LTV fall below 45% because debt falls, rather than stay above 50% because values fall again?
The next letting data matters just as much. Vacancy below 13% would be the first convincing reversal of a four-year deterioration. Vacancy above 15% would mean more foregone rent and more spending. Interest cover above 2 times would show a usable buffer. Remaining near 1.7 times would leave the refinancing ratchet in control.
The FY2026 cash-flow statement is the cleanest test. Positive operating cash after interest and property capex would show that the estate can maintain itself without selling another building. A restored distribution before that point would weaken the bridge. This is an observation about funding, not a prescription.
What would change the market's verdict
The 12 August decline was not a panic response. The shares were already pricing severe scepticism. H1 supplied enough new evidence to justify a little more: another double-digit NTA decline, vacancy at 14.5%, a wider statutory loss and operating cash wholly absorbed by interest.
There is counter-evidence. CLS has completed substantial sales, reduced net debt and extended maturities. Germany and France provide longer leases and lower vacancy than the UK. Contracted rent sits £9.5 million below ERV, leaving income recovery potential. External valuations are a more disciplined starting point than an unsupported management estimate (CLS 2026a; CLS 2026b).
But the first roughly £268 million of cumulative sales did not create a post-capex owner-cash surplus or bring LTV below the target range. That history makes the last £43 million useful mainly as evidence. Near-book completions, lower vacancy and positive post-capex cash would make the appraisal more credible. Repeated sale discounts, LTV above 50% and another cash deficit would show why 177.7p remains an accounting reference rather than money available to owners.
At 47.15p, the market is not merely forecasting lower office values. It is pricing a long, costly path from appraisal to cash. The next accounts will reveal whether that path has finally started to shorten.
Source notes: confidence and missing information
Verification is partial. The H1 report, presentation, FY2025 annual report, prior annual reports, company regulatory feed, LSE page, Companies House record and independent market coverage were fetched and read. The Finance API health check passed, but its resolver returned no exact LSE:CLI match. No fuzzy or alternate-venue identity was used. Identity was instead matched to the LSE company page and Companies House company 02714781.
The market-data share count needs care. CLS reported 438.8 million issued ordinary shares, 40.7 million held in treasury and 398.1 million voting rights. The LSE displayed £196.9 million of instrument market capitalisation, which implies a different capital-count convention. Frontmatter preserves the LSE snapshot for chart consistency; per-voting-share analytical calculations use 398.1 million and say so. Exact future disposal prices, property-level debt cancellation and maintenance versus leasing capex are not yet disclosed in enough detail to turn the scenario ranges into point estimates.
References
- CLS Holdings plc, Half-Yearly Report 2026 and H1 results release (CLS 2026a).
- CLS Holdings plc, H1 2026 investor presentation (CLS 2026b).
- CLS Holdings plc, Trading update, 4 August 2026 (CLS 2026c).
- CLS Holdings plc, Annual Report 2025 (CLS 2025).
- CLS Holdings plc, Annual Report 2024 (CLS 2024).
- CLS Holdings plc, Annual Report 2023 (CLS 2023).
- CLS Holdings plc, Annual Report 2022 (CLS 2022).
- London Stock Exchange, CLS Holdings plc company and market page, 12 August 2026 (LSE 2026).
- Companies House, CLS Holdings plc, company 02714781 (Companies House 2026).
- London South East, reporting on the H1 results and share-price reaction (London South East 2026).
- Office for National Statistics, inflation and price indices, June 2026 (ONS 2026).
- European Public Real Estate Association, monthly listed real-estate reports (EPRA 2026).
- Great Portland Estates plc, FY2026 latest-results and NTA context (GPE 2026).