The model is bullish. The evidence is narrower.
Lead with the objections: a pre-buy only pays when a supplier raises prices soon after December and by more than the cost of holding the stock. Then weigh the case for.
Four reasons not to stock up on the model's say-so
- 1Carrying is expensive.
- 2One assumption drives everything.
- 3Last winter did not cooperate.
- 4The TRY impact is unknown. No monthly volumes are entered (0 of 41), so every saving is a per-unit percentage. Cash headroom and storage space are also open.
What the workbook rests on
- ASteep 2026 drift.
- BInflation is still running.
- CCheap structures exist. Price locks, staggered call-offs and supplier-held stock capture most of the saving without the financing cost.
Our reading: where the model and last winter agree
Each pre-buy item is tested a second way: replay the same December purchase against the actual Dec-25 to Mar-26 prices in the workbook. This is a cross-check built from the workbook's own data, not a replacement for its calls.
Inflation is slowing. Money still costs more than prices rise.
Year-on-year inflation, official vs independent
TurkStat CPI and the ENAG E-CPI use different methods. Both are falling; the gap between them is narrowing.
Monthly carry vs monthly price creep
What holding a unit of stock costs each month, against how fast prices drift.
Averages hide the story. A pre-buy is a bet on individual price paths, so look at the spread first, then at when each item actually moved.
2026 price change by item
Food, opex and energy: Jan-26 to Sep-26. Beverages: Dec-25 to latest. Bars in colour are on the pre-buy list.
Price journeys, indexed
Pick up to four items. Index: Jan-24 = 100. The shaded band is the Dec-25 to Jan-26 reset window.
Heatmap: annual price change, 2022 to 2026
Each column is January to January. The last column is the eight months to September 2026. Red is a rise, blue a fall.
How one unit of stock earns or loses money
Buy in December at today's price, hold for the months of cover, and compare with buying each month at the new price. The difference, less financing and storage, is the net saving.
Unit economics, step by step
Everything as a share of the December price. Choose an item.
The break-even bar
How big must the January increase be before holding stock pays? It rises with the months of cover.
Forty-one items, three calls
Buy ahead when the net saving clears 3% of spend. Lock the price when it is positive but below the hurdle. Skip when it is negative or the cover outlasts the shelf life.
Net saving per unit, by item
After financing and storage. The dashed line is the 3% hurdle.
Strategy tier vs model call
The qualitative tiers were set first. The model mostly agrees, except where it is more bullish than the tier.
Average net saving by group
Open the full item register (41 items, sortable)
Where the thesis can break
Three tests, each designed to make the pre-buy case fail. Whatever survives is robust.
Test 1 · Replay last winter: when did prices actually move?
Test 2 · How much of the 2026 drift must repeat?
Share of each item's 2026 move that suppliers must pass on in Jan-27 for the item to clear the hurdle. The workbook assumes 100%; it suggests flexing 50–120%.
Test 3 · Is the model sound and complete?
Nine integrity checks pass. Two completeness checks do not, and they matter more than the nine.
Change an assumption. Watch the calls flip.
This is the workbook's model running live in your browser with its own formulas. Every tile is one item. A pulsing tile changed its call versus the workbook base case.
Tripwires at these settings
What if we commit a budget?
With no volumes in the workbook, assume equal TRY spend on each Buy-ahead item. Indicative only.
Sensitivity curves
How many items stay on Buy ahead as one assumption moves and the others hold at the settings on the left.
Capture the saving without carrying the stock
Where the replay and the model disagree, the cheaper answer is usually a commercial one: fix the 2027 price now and take delivery later.
3 months of cover
Steep 2026 increases, stable shelf life, easy storage. Place December orders.
Fix the 2027 price
Negotiate a fixed 2027 price with staggered delivery or buy-now, deliver-later. Buy physically only on promotions.
Buy as needed
Falling or regulated prices, storage limits and all perishables: meat, seafood, dairy, fresh produce, electricity and gas.
Assumptions, open items and data quality
Every number above depends on the inputs below. Confidence labels show how much weight each one can bear.
Assumption ledger
Information needed before committing
Data issues to fix before the meeting
The first seven are listed in the workbook. The last two were found while building this page.
What we need decided, and by whom
A staged commitment: approve the robust core now, gather the facts that decide the rest, and set tripwires that cancel orders automatically.
Approve the robust core
Get supplier facts before ordering
Purchasing asks every Tier 1 and Tier 2 supplier for the date and size of their next list-price change, and for TRY-fixed quotes on EUR/USD-linked imports. No date, no stock.
Fix the money inputs
Finance confirms the hotel's cost of funds after the 22 Oct MPC decision and the Dec-26 cash headroom. Stores confirms dry-store, cellar and chemicals capacity.