A cheap stock paying for its own growth — and H1 2026 margin is the verdict.

• The bull math. EV/EBITDA sits at 6.1× against a 20-year mean of 7.9× and a 5-year mean of 9.2× — the deepest discount since 2008. ROCE is still 16% with £2.15B of revenue compounded from £811M in 2020.
• The bear math. Underlying PBT fell 9.4% in FY25, ROCE dropped from 20.3% to 16.0% in twelve months, and like-for-like sales slowed to +1.6% in the first nine weeks of 2026 — below the +3–4% the operating leverage needs.
• The single decisive print. H1 FY26 operating margin (due early August). Above 9% with capex normalising and the 20-year multiple reopens; below 8.5% and the £9–10 bear case goes live.

Sales still grew 6.8% — but capex quintupled and FCF collapsed.

The supply-chain build — Derby (live H2 2026) and Kettering (H1 2027) distribution centres — sits in the asset base before it generates revenue. That plus 5.6% wage inflation compressed operating margin by 100bps while revenue kept climbing. Management has guided capex down to £200M in 2026 and £150–170M from 2027; if that lands, FCF rebuilds from £52M toward £200M by 2028 on the same revenue base.

The doubling target quietly disappeared from the CEO letter.

Before: At the October 2021 Capital Markets Day, management committed to doubling sales by 2026 — from £1.23B to ~£2.46B. The phrase appeared verbatim in three straight annual letters through FY24, framed as the valuation-load-bearing promise that justified a 25× multiple.

Pivot: The FY25 preliminary results, signed March 2026, omit the doubling target entirely. It has not been formally retracted; it has simply stopped being said. At £2.15B FY25 revenue plus flat FY26 guidance, the math no longer works.

Today: The new framing is a 3,500-shop estate opportunity and ROCE recovery — a smaller, more falsifiable promise. The question for the next chapter: does the market re-rate on a plausibly achieved goal, or stay suspicious of a company that walked one back without naming it?

Worst FTSE 250 name of 2025 — counter-trend rally inside an unresolved stall.

• The drawdown. From £34.16 in December 2021 to a £14.07 low in July 2025 — a 59% peak-to-trough. Greggs finished 2025 as the worst-performing stock in the FTSE 250, down roughly 43% on the year, on a hot-summer H1 miss and Jefferies' GLP-1 demand-destruction downgrade.
• The bounce. Price at £16.43 sits 1.5% above a flattening 200-day; RSI 61, MACD positive. The near-term tape has turned, but the 50-day is still sub-200, the death cross is unresolved, and four of the five biggest-volume days of the last six months were panic sells.
• The two levels that matter. A weekly close above £17.50 reclaims the 100-day and breaks 18 months of lower highs — targets £20–22. A daily close below £15.00 reopens the £14.07 low and the pandemic-era £14 floor.

Bifurcated sell-side, new structural bear thesis, and a Chair buying at the lows.

• JPM vs Jefferies, 500p apart. J.P. Morgan initiated Overweight in December 2025 at 2,110p (~35% upside, Dec-2027). Jefferies downgraded to Hold in February 2026 and cut its target from 2,500p to 1,610p. Consensus sits at Hold with ~1,993p — the widest ratings split on the name in a decade.
• GLP-1 demand risk is now embedded in the bear case. Jefferies explicitly cited Ozempic/Mounjaro as a structural risk to Greggs' highest-frequency snack customer; the report alone drove a 6% single-day decline. Unprovable from historical data, but now part of how the market prices the stock.
• Insider signal is mixed but bottom-tilted. CFO Hutton sold 7,438 shares at £15.71 in November 2025 — a trim, not an exit. New Chair Matt Davies put £20k of his own money in at £16.00 in August 2025, near the 52-week low. All insiders combined own 0.09% of the company.