When EBITDA Lies: Retail & Consumer – IFRS 16, EBITDA(R) and the Inventory Trap

Since IFRS 16, the number a retailer calls EBITDA is really EBITDA(R): earnings before the rent, as well as before interest, tax, depreciation and amortisation. The IFRS 16 impact on EBITDA is largest where margins are thinnest and estates are biggest – and the stock, rebates and like-for-like figures beneath it carry traps of their own.

The retail & consumer paper in the Bloor Finance series: a CFO-perspective reading of JD Sports Fashion plc (LSE: JD.) and ASOS plc (LSE: ASC) against the UK-listed retail peer set – NEXT, Tesco, Marks and Spencer and Debenhams Group – on audited annual reports and the latest interim and trading statements published to 8 October 2026.

The opening paper in this series argued that wherever an organisation changes how it operates, cost leaves the line EBITDA sees and reappears as depreciation, interest, leases and provisions, where it does not. The professional services paper found it in a people business. In retail the cost that moved is rent, and it moved by accounting standard rather than by management choice. The opening paper put it this way:

“Thin margins and lease-heavy estates make retail the sector where IFRS 16 flatters EBITDA most … every point of rent reclassified as depreciation-and-interest is a point EBITDA no longer sees.”

Re-verifying the opening paper’s retail anchors tightened two of them. The IFRS Foundation’s own effects analysis estimated retailers’ then off-balance-sheet leases at 21.4% of total assets, against 5.4% across all the companies it studied – the double-digit lift the opening paper described. Amazon’s million-robot figure is the company’s own. The six-figure hiring the opening paper cited comes from internal documents reported by the New York Times, which Amazon disputes, and is attributed that way here. Since the opening paper went to press, four of the six companies in this paper have reported. NEXT raised its full-year profit guidance to £1,255m on 17 September. JD Sports reported on 23 September that its own IAS 17 proxy for EBITDA had fallen 21%. ASOS guided on 24 September to adjusted EBITDA above the middle of its £150–180m range. Tesco reported its interims on 8 October, with adjusted operating profit up 6.3% to £1,783m.

IFRS 16 took rent out of retail EBITDA without taking it out of the business. Since 2019 a UK retailer’s EBITDA has been a structurally larger number than the one it replaced, and most of the increase is the store estate.

The mechanism is in the standard. A lessee recognises a right-of-use asset and a lease liability, then presents ‘interest expense on the lease liability separately from the depreciation charge for the right-of-use asset’. Both charges sit below EBITDA. Only short-term leases, low-value leases and variable rent that is not linked to an index or rate stay in operating costs.

The standard-setter knew what this would do. Its 2016 effects analysis said that ‘EBITDA applying IFRS 16 excludes all expenses related to leases’ and ‘corresponds to EBITDAR applying IAS 17’. Retail was its most exposed sector. It found that for 36% of the retailers in its sample, leases not yet on the balance sheet were worth more than half of total assets. PwC’s study of the same year estimated a median EBITDA uplift of 41% for retailers and 98% more debt.

Post-2019 retail EBITDA is the old EBITDAR wearing the old name. That is why this paper writes EBITDA(R). It is also why a UK retailer’s EBITDA and a US retailer’s are no longer the same measure. Under US GAAP (ASC 842), operating-lease cost is still recognised as a single straight-line charge within operating costs, so a US GAAP retailer’s EBITDA still carries its rent. Amazon is therefore a contrast in this paper, not a peer.

JD Sports does the honest arithmetic itself. It reports operating cash flow net of lease repayments as ‘a close cash proxy for EBITDA on an IAS 17 basis’ – in effect a pre-IFRS 16 EBITDA. In the latest half that proxy fell 21%.

Start with the year to 31 January 2026, on the audited accounts. Operating profit before adjusting items was £1,035m. Adding back £287m of depreciation on owned assets, £562m on right-of-use assets and £48m of non-adjusting amortisation gives an IFRS 16 EBITDA of about £1.93bn. The leases cost £508m of principal and £149m of interest, £657m in all, and a further £84m of variable rent and £5m of short-term leases was charged in operating costs. Total cash out for leases was £746m.

Take the £657m out and pre-IFRS 16 EBITDA is about £1.28bn. JD’s own cash proxy, reached by a different route, is £1,309m. IFRS 16 adds roughly half again to JD’s EBITDA, and lease cash absorbs about 34% of the EBITDA it reports.

Net cash of £311m before leases is net debt of £2,827m after them. Both numbers are true. Only one of them is the debt.

The balance sheet carries £3,138m of lease liabilities (£3,059m a year earlier), on £3,783m of undiscounted payments. JD’s own alternative measures make the case for this paper. Its ‘operating profit before adjusting items after lease interest’ was £886m against £1,035m, so JD itself treats the £149m of lease interest as a cost of trading.

The half to 1 August 2026 moves the same way. The IAS 17 proxy was £433m against £548m. Lease principal rose to £303m from £254m – partly because the half held seven monthly rent dates rather than six – and lease interest to £76m from £70m. Profit before tax and adjusting items fell to £282m from £351m. Net cash before leases was £168m; net debt including leases was £2,936m.

The pure-play DTC anchor shows how much of this is the estate. ASOS has no shops; its £224.5m of lease liabilities are fulfilment centres and offices. Lease cash of £32.1m was about 24% of its £131.6m adjusted EBITDA. NEXT, whose estate is shorter-dated, comes out at about 14%. NEXT does not report EBITDA at all, and charges lease interest inside its divisional profits.

Inventory valuation is where retail EBITDA is decided before the till opens. Stock is held at the lower of cost and net realisable value, so the provision is a judgement about what unsold product will fetch. That judgement runs through cost of sales, and through EBITDA, when it changes.

ASOS shows the full cycle. At 31 August 2025 it held £523.5m of gross stock against a provision of £121.2m, or 23.2%. A year earlier it held £683.6m against £163.3m, or 23.9%. Between 2023 and 2024 it recognised £228.0m of inventory write-downs as adjusting items, outside adjusted EBITDA. In 2025 it sold that stock through clearance channels and stripped the proceeds out of the adjusted figures too: £13.0m of revenue and £8.3m of cost of sales.

The adjusted line saw neither the write-down nor the recovery. It reports a business that never had the stock.

The gap between the two views is large. ASOS reported adjusted EBITDA of £131.6m for the year and a statutory loss before tax of £281.6m – a difference of £413.2m. The biggest single item was £179.3m of property costs to vacate and mothball its Atlanta fulfilment centre, which is lease debt leaving the business through adjusting items. In the half to 1 March 2026 a further £67.3m technology-asset impairment went the same way. ASOS’s auditor names the inventory provision as a key audit matter.

Store retailers carry the same judgement at a different scale. JD’s provision was £79m at 31 January 2026, against £77m a year earlier and 3.8% of gross stock. It recognised £33m and used £27m during the year, and its cost of sales explicitly includes shrinkage and provision movements. Tesco disclosed £1,503m of ‘inventory losses and provisions’ expensed in the year, against £1,440m – equal to 53% of its year-end inventory. Debenhams Group booked £52.0m of UK stock clearance as an exceptional item in its 2025 year. M&S attributed a lower gross margin to ‘increased markdown and waste’ without giving a provision figure.

Stock turnover – cost of goods sold divided by average inventory – is the retail gauge that EBITDA cannot flatter. It measures how hard the stock works, and it falls before margins do.

On the audited figures, JD turned its stock about 3.3 times in the year to January 2026, or roughly 110 days of cost of sales. ASOS turned its net stock about 2.8 times, roughly 128 days. NEXT turned its stock about 4.2 times, roughly 86 days.

These are not like-for-like comparisons. Each company’s cost of sales has its own definition: JD includes inbound freight, NEXT uses standard cost and had a 53-week year, and ASOS’s stock includes a £37.2m ‘right to recover’ asset for goods expected back as returns. The trend within each company is the reliable signal, not the level across them.

The trend at JD is the one to watch. In the half to 1 August 2026 its inventory was £2,335m, against £2,294m a year earlier – up 2% – while revenue fell to £5,899m from £5,940m. Stock growing faster than sales is the first sign of the markdowns to come.

For a DTC business, returns are the second inventory line. Debenhams Group’s returns rate was 28.2%, and it discloses that each point moves revenue by about £7.0m and operating profit by about £3.0m.

Supplier rebates are a reduction in what stock cost, not a source of revenue, and they can only be recognised once they are earned. The sector’s most expensive lesson in that rule is more than a decade old and still on Tesco’s audit report.

In 2014 Tesco disclosed a £263m overstatement of profit. The FCA’s Final Notice records that it arose principally from accelerated recognition of commercial income and delayed accrual of costs. Twelve years on, Tesco’s auditor still names ‘recognition of commercial income’ as a key audit matter and the area of greatest potential for fraud. Tesco now discloses the balances at 28 February 2026: £105m in trade receivables, £130m in accrued income, £157m netted against trade payables and £(14)m within inventories.

No company in this paper discloses how much rebate income it recognised in the income statement. JD’s policy treats ‘supplier rebates, discounts and marketing contributions’ as a reduction in cost of sales and defers £3m within inventory (£1m a year earlier). ASOS holds £1.6m of accrued rebates and has cited ‘lower volume based rebates from reduced volumes’ as a cost driver. Debenhams Group recognises retrospective rebates in the period to which they relate. NEXT’s figure is n/d.

A rebate booked a quarter early is gross margin, and gross margin is EBITDA. The amount is the one figure in the sector that no one publishes.

Like-for-like (LFL) growth is a definition, not a fact, and definitions change. The honest test is whether the measure is the same one used last year.

JD’s annual report defines LFL as sales from stores that traded through the whole of both periods, ‘excluding the impact of store openings, closures, relocations, and conversions’. On that basis the JD fascia was down 2.8% in the year to January 2026, and the Group down 2.1%. In the latest half JD introduced ‘comparable sales’, which includes relocations and which it says ‘provides a more complete view’. Comparable sales were down 2.0%; LFL was down 2.8%.

Elsewhere the definitions differ again. Tesco measures LFL on stores open at least a year, excluding dunnhumby and mall rental income; its interim LFL was 1.0%. NEXT reports a stores LFL of −3.3% in the half alongside group full-price sales up 7.7%, a measure that excludes Sale and clearance. M&S reported LFL of 6.7% in Food and −7.5% in Fashion, Home & Beauty; its definition is not given in the results release.

ASOS reports no LFL at all. Its honest gauges are active customers (customers who have shopped in the last 12 months), order frequency, average basket value and conversion, each with a published definition. For a DTC brand those are the like-for-like.

Read on one basis – UK-listed, IFRS, sterling – the peer set shows the same pattern. Every store-estate retailer carries more lease debt than net financial debt – the two online businesses do not – and none of the six discloses rebate income or AI spend.

Honest gauges – UK retail peer set (latest audited year; IFRS; £m)

Company (year end)	Lease liabilities (prior year)	Net debt before leases → including leases	Lease cash as share of IFRS 16 EBITDA (Bloor)	Stock provision / losses disclosed	Supplier income disclosed	Like-for-like basis
JD Sports (31 Jan 2026)	3,138 (3,059)	Net cash 311 → net debt 2,827	c.34% (657 of c.1,932)	Provision 79 (3.8% of gross); 33 recognised	Policy; 3 deferred in stock; income n/d	Excludes openings, closures, relocations; H1 ‘comparable’ adds relocations
ASOS (31 Aug 2025)	224.5 (289.6)	184.7 → c.409.2	c.24% (32.1 of 131.6 adjusted)	Provision 121.2 (23.2% of gross); 228.0 written down FY23–24	1.6 accrued; income n/d	No LFL; active customers, frequency, basket, conversion
NEXT (31 Jan 2026)	999.8 (1,014.4)	713 → c.1,713	c.14% (223.0 of c.1,608)	n/d	n/d	Stores LFL excludes openings, closures, refits; full-price sales
Tesco (28 Feb 2026)	7,884 (7,716)	2,679 → 10,563	c.21% (c.1,038 of c.4,880)	Losses and provisions expensed 1,503	Balances: 105 / 130 / 157 / (14); income n/d	Stores open at least one year
M&S (28 Mar 2026)	2,750.0 (2,227.4)	Net funds 338.2 → net debt 2,411.8	n/d (cash lease payments 430.0)	n/d (‘markdown and waste’)	n/d	n/d in results release
Debenhams Group (28 Feb 2026)	74.4 (120.2)	93.2 → c.167.6	c.26% (13.7 of 53.3 adjusted)	n/d; FY25 stock clearance exceptional 52.0	Policy; income n/d	No LFL; GMV, 5.7m active customers

All figures are from each company’s latest audited annual report or full-year results release (ASOS and Debenhams Group EBITDA are the companies’ adjusted measures). IFRS 16 EBITDA and lease-cash shares are Bloor arithmetic: operating profit plus depreciation and amortisation, against lease principal plus lease interest (Tesco: total lease cash outflow less short-term and low-value lease costs). They are indicative, not company figures. NEXT and Tesco reported 53-week years. US GAAP companies are excluded so that every row is on one basis. Amazon, the opening paper’s retail exhibit, reports under ASC 842, where operating-lease cost stays within operating costs, so its EBITDA is not comparable. n/d – not disclosed in the cited document; never treated as zero. No company is ranked.

The spread in the lease-cash column is the point of the table. Rent takes a third of EBITDA at JD, about a fifth at Tesco and about a seventh at NEXT. On reported EBITDA all three look like the same kind of business; on EBITDA(R) they do not. The stock columns diverge further, from a 3.8% provision at a store retailer to 23% at a pure-play online one.

A buyer will put the rent back. Lenders, rating agencies and acquirers price a retailer on lease-adjusted numbers, because the leases are debt whichever line the accounts put them on.

The agencies have said so. Fitch now uses companies’ reported IFRS 16 and ASC 842 lease liabilities in its lease-adjusted leverage, and still quotes EBITDAR-based metrics for retailers. S&P adjusts debt, earnings, cash flow and interest ‘for comparability across accounting regimes’. A buyer’s normalisation of a retail EBITDA usually runs in five steps.

– Restate to EBITDAR and lease-adjusted net debt. JD: c.£1.93bn of IFRS 16 EBITDA, plus £89m of variable and short-term rent, gives about £2.02bn of EBITDAR against £2,827m of net debt including leases – about 1.4 times, not ‘net cash £311m’.

– Test the lease term. Tesco discloses £14.7bn of potential lease payments beyond the reasonably certain term, which are not in its £7,884m liability. NEXT’s weighted average remaining lease is 7.0 years.

– Re-run the stock provision. A buyer will apply its own sell-through and loss rates; the gap between a 3.8% and a 23% provision is a gap in the price.

– Trace the rebates. Rebate receivables, accrued income and amounts netted against payables are checked against supplier confirmations, as Tesco’s auditor does.

– Hold the LFL definition constant across the periods being valued.

The buyer will also ask what the technology spend buys. The automation numbers are real – Amazon runs more than a million warehouse robots, and internal documents reported by the New York Times modelled avoiding six-figure hiring, which Amazon disputes. But the attribution is where the headline inverts: analysts already forecast that half of the layoffs blamed on AI will be quietly reversed.

ASOS says AI now writes more than 15% of its software and makes about 20% of its buying decisions. In the same half it impaired £67.3m of technology assets, outside adjusted EBITDA. MIT NANDA’s 2025 study found 95% of organisations getting zero return from generative AI. No company in this peer set discloses its AI spend, let alone attributable margin.

Not a different number, but the same numbers read with the rent and the stock put back. The retail edition of the honest-gauge dashboard, applied to the two anchors:

– EBITDA(R) both ways – reported IFRS 16 EBITDA and pre-IFRS 16 EBITDA, side by side, with total lease cash outflow. JD: c.£1.93bn and c.£1.28bn; lease cash £746m.

– Lease-adjusted net debt rather than ‘net cash before leases’. JD £2,827m; ASOS c.£409m.

– Lease cash as a share of EBITDA. JD c.34%; ASOS c.24%.

– Inventory provision as a share of gross stock, and its movement. JD 3.8% (£33m recognised); ASOS 23.2% (23.9% a year earlier).

– Stock turnover and stock days, against sales growth. JD c.3.3 times; H1 inventory +2% against revenue −0.7%. ASOS c.2.8 times.

– Shrinkage, waste and markdowns expensed – disclosed by Tesco (£1,503m); n/d at JD and ASOS.

– Supplier income recognised in the income statement, with rebate balances. n/d across the peer set.

– LFL on an unchanged definition, alongside total sales and space change. JD: LFL −2.8% against comparable sales −2.0% in the half. – AI and personalisation spend, with the margin attributed to it. n/d across the peer set.

For a retail CFO the work is specific. Publish pre-IFRS 16 EBITDA next to the reported figure, as JD already does, and lead with lease-adjusted net debt. Disclose the stock provision as a percentage of gross inventory and say what moved it. Publish supplier income recognised in the year, not just the balances. Keep one LFL definition, or restate the history when it changes.

JD and ASOS are not the weak cases. JD is more open about its own IFRS 16 impact on EBITDA than most of the sector, and ASOS’s notes set out its stock cycle in full. That is why they are the right cases. If the most transparent store retailer still spends a third of its reported EBITDA on rent, and the most-written-down online retailer reports an adjusted line that never saw the write-downs, the sector’s EBITDA should be read as EBITDA(R) until the CFO shows otherwise.

Every figure in this paper comes from a named, dated document. The principal ones are listed below with links, and a full Source Note gives the origin, pinpoint and status of every figure. IFRS 16 EBITDA, pre-IFRS 16 EBITDA, EBITDAR, lease-cash shares, stock turnover and provision percentages are Bloor arithmetic on reported figures, with the formula shown in the Source Note. Where a figure is not disclosed in the cited document it is shown as n/d, never as zero. No company is ranked. This is analysis for a CFO audience and is not investment advice.

• JD Sports Fashion plc – Annual Report & Accounts 2026 (Notes 15: Leases, 18: Inventories; Alternative Performance Measures): Annual Report & Accounts 2026
• JD Sports Fashion plc – Half Year Results 2026/27, 23 September 2026: Half Year Results 2026/27
• ASOS plc – Annual Report and Accounts 2025 (Notes 13: Leases, 16: Inventories; APMs): Annual Report 2025
• ASOS plc – Interim Results, 23 April 2026, and Pre-close Trading Update, 24 September 2026: Interim Results 2026
• NEXT plc – Annual Report and Accounts January 2026 and Half-year Results July 2026: Annual Report 2026
• Tesco PLC – Annual Report and Financial Statements 2026 (Notes 13, 17 and 21) and Interim Results 2026/27, 8 October 2026: Annual Report 2026
• Marks and Spencer Group plc – Preliminary Results 2025/26, 20 May 2026: M&S Corporate Website
• Debenhams Group plc – Final Results, 16 June 2026: Debenhams Group Website
• FCA – Final Notice to Tesco plc and Tesco Stores Limited, 28 March 2017: FCA Final Notice
• IASB – IFRS 16 Leases Effects Analysis, January 2016: IFRS 16 Leases Effects Analysis
• Amazon – One millionth robot announcement, 1 July 2025: Amazon Official Website
• The New York Times – Report on internal hiring projections, 21 October 2025 (disputed by Amazon): The New York Times
• MIT NANDA – The GenAI Divide: State of AI in Business 2025: MIT NANDA Research
• Forrester – Predictions 2026: Forrester Research
• PwC – 2016 research: PwC Official Website
• Credit rating agencies – Lease methodologies: Fitch Ratings, S&P Global Ratings, and Moody’s Ratings. All sources are attributed and used directionally.

Note: For references without a supplied direct report URL, the links lead to the publisher’s website rather than the specific publication.

Print Article for this blog Download here