Syntha Lab RU

Markdown mathematics: steps, leftovers, timing and liquidation

A quantitative treatment of in-season price reductions: how much extra volume pays for a cut, when to start, which depth and steps to choose, how it ties to sell-through and margin and when leftovers are simpler to liquidate. The decision: the season markdown schedule and the cutoff threshold.

Math and analytics Brand and retail Verified: 2026-11-13

Open interactively →

Why this document

The commercial director and the pricing manager read this when the season is ending and stock is left. It answers when to start a markdown, how deep and in what steps. This matters because a cut started early gives away margin, and one started late cannot sell the leftover in time. The threshold for required volume rise shows whether the cut pays at all at a given elasticity. If it does not, it is simpler to liquidate the leftover at once and free warehouse space and cash for the next season.

At a glance

What is this document and why is it needed?
A quantitative treatment of in-season price reductions: how much extra volume pays for a cut, when to start, which depth and steps to choose, how it ties to sell-through and margin and when leftovers are simpler to liquidate. The decision: the season markdown schedule and the cutoff threshold.
Who owns the document and how often is it rebuilt?
Buyer or category manager with an analyst; the season markdown schedule is drawn up before launch and reviewed weekly; approved by the commercial director, the margin effect is shown to finance.
What mistakes are the most common?
Cuts are set looking only at the discount without computing the threshold u = m/(g − m) → volume does not repay margin → compare expected rise with the threshold. The first step is postponed → leftovers grow while demand fades → trigger it by weeks of cover. A deep single cut → margin lost and customers trained to wait → a ladder with transition conditions.

For whom

Brand
A brand reduces price in its own channels (D2C, marketplaces, own stores) and agrees price support with partners; wholesale leftovers at a partner affect the brand through returns, compensation and the next order, so brand and partner schedules are aligned.
Retail
A retailer runs a markdown schedule by category and store, using a reduction ladder and triggers on weeks of cover; a cut in one store or channel affects the others (cannibalization, flows), so the schedule is shared.

What it consists of

When a cut pays off

  • A price cut by share m at gross margin g = 1 − c/p pays off if volume rises at least (1 + u) times where u = m/(g − m). At a margin of exactly 70% a 30% cut needs a 75% volume rise, 40% already 133% and 50% 250%; at the exact example margin of 69.7% these are 76, 135 and 254%.
  • The expected volume rise at elasticity 2 is (1 − m)^−e − 1: for 30% it is 104%. A deep cut pays in margin only when it clears stock that would otherwise be liquidated cheaper; otherwise it buys volume with margin.

Leftovers, time and sell-through

  • Season model (illustrative): stock 1,000 units, cost RUB 1,000, price RUB 3,300 excluding VAT, full-price demand decays 16% a week, the leftover after 20 weeks sells at a liquidation price of 25% of full. Without a cut 71% is sold with 286 units left.
  • Timing matters more than depth: a 30% cut from week 7 gives RUB 1,732,682 of gross profit while the same cut from week 9 gives RUB 1,689,339 and from week 5 RUB 1,658,162; in the model the optimum is week 7 because earlier margin is lost and later demand is already too weak.

The reduction ladder

  • The −20/−40/−60% ladder (weeks 7, 11, 15) sells 100% of stock and yields RUB 1,636,929 against RUB 1,689,339 for one 30% cut that leaves 103 units; steps win on sell-through but lose on margin if the first step is set late.
  • Steps are held at least 2-3 weeks, otherwise a step effect cannot be told from noise; the next step depth grows slower than demand falls, otherwise the cut stops working.

Cutoff and liquidation

  • Brand: cutoff by remaining stock and channel: leftovers above the threshold go to an outlet, wholesale partners or a trade-in program instead of a deep cut in the main channel; this must be agreed with partners to avoid devaluing their stock.
  • Retailer: cutoff by weeks of cover and age; leftovers after the last step are consolidated in one place or passed to liquidation to free space and warehouse. The decision compares the net liquidation price with the margin lost by further cutting.

Brand and retailer: the markdown schedule in practice

  • Brand: a brand plans markdown for its own channels and price support for partners; the partner schedule is set by the contract (sale windows, compensation), not only by brand mathematics.
  • Retailer: a retailer plans markdown by category, store and item role; the schedule is shared across the chain but weeks-of-cover triggers are computed by store and channel.
  • Brand: a brand sends the leftover after the last step to an outlet, closed sales for staff and partners or carries it to the next season, and the decision depends on brand policy and channel rules.
  • Retailer: a retailer consolidates the leftover after the last step into stock stores and an online sale or passes it to wholesale buyers of leftovers; the liquidation price is set by the leftover market.
  • Brand: returns of unsold goods from partners (stock rotation, contractual return) enter the brand markdown calculation as deferred leftover; without them sell-through on shipments is overstated.
  • Retailer: customer returns to stores and online put goods back on sale with a delay and sometimes with lost quality; the markdown calculation includes the return share and return period.
  • Brand: a brand tracks the price index in a channel and the price gap between channels; a deep markdown in one channel breaks price discipline and must be agreed.
  • Retailer: a retailer tracks competitor prices on the same items: the markdown schedule adapts to the market but does not go below the threshold where a cut stops paying in margin.
  • Brand: a brand shows finance the effect of markdown on revenue, margin, the returns provision and the partner season; the assessment uses fast and slow demand decay scenarios.
  • Retailer: a retailer shows finance the margin bridge (full price, steps, leftover), the provision for inventory write-down and the effect on turnover and free space.

Where the data comes from

Data sourceFields
Weekly sales, prices and stockSold, stock, price by SKU life week; launch date; full-price share.
Item economicsCost, price excluding VAT, channel commission, holding and capital cost, liquidation price.
Elasticity and life curveElasticity estimate for the group (see the elasticity document) and the demand profile by life week.

How to build it step by step

  1. Step 1. Set the season target: target sell-through by the cutoff date and the minimum gross margin for the category.
  2. Brand: Step 2. Determine in which channels and when a cut is allowed and agree windows with partners; fix price support.
  3. Retailer: Step 3. Split the range by role and age; set weeks-of-cover triggers for the first step.
  4. Step 4. Build the demand curve by week at full price (decay) and estimate the group elasticity.
  5. Step 5. Compute the payoff threshold u = m/(g − m) and compare it with the expected rise (1 − m)^−e − 1 for candidate depths.
  6. Step 6. Scan the start week of the first step and choose the maximum of gross profit within the allowed price range.
  7. Step 7. Build the ladder: depths, weeks and transition conditions; run scenarios (fast decay, slow decay).
  8. Step 8. Define the cutoff: the leftover and date after which stock goes to liquidation or another channel; compute the net liquidation price after costs.
  9. Step 9. Show finance the margin by step and the total as a bridge (full price, steps, leftover) and the effect on the provision for inventory write-down.
  10. Step 10. Compare actual with the step plan weekly; when sell-through deviates beyond the threshold revisit the start and depth.

Formulas

MeasureFormulasNote
Margin after a cutg(m) = 1 − c / (p·(1 − m))A cut of m lowers margin more than it seems: at g = 60% a 30% cut leaves 43%.
Volume uplift thresholdu = m·p / (p·(1 − m) − c) = m / (g − m)The neutral gross profit condition against selling at full price; ignores capital and space.
Expected volume upliftu_exp = (1 − m)^(−e) − 1e is absolute elasticity; compare with the threshold.
Sell-throughST, % = sold in period / (sold + closing stock) × 100A brand computes it on shipments and on sell-out separately.
Season gross profitGP = Σ_w sold_w · p·(1 − m_w) + leftover · p · k_liq − c · Q₀Cost of the whole purchase Q₀, not of what was sold: the leftover is already paid for.
Demand on a stepd_w(m) = d_w · (1 − m)^(−e), sold = min(stock, d_w(m))Decay d_w = d₁·exp(−λ·(w − 1)).
Cutoff criterionliquidate if p·k_liq − costs ≥ p·(1 − m_next)·sell_next − c_holdingA comparison of net receipts on the two paths per unit.

Formulas in notation

Expected uplift
u_exp = (1 − m)^{−e} − 1
Constant elasticity e.
Demand decay
d_w = d₁ · exp[ −λ·(w − 1) ]
λ is the full-price decay rate.
Weekly sales
s_w = min[ S_{w−1}, d_w·(1 − m_w)^{−e} ]
S_w is stock at the end of week w.

Worked example

Five markdown schedules on a stock of 1,000 units over 20 weeks

ScheduleSell-through, %Leftover, unitsRevenue, RUB thousandGross profit, RUB thousand
No markdown71.42862,5911,591
One cut of 30% from week 510002,6581,658
One cut of 30% from week 989.71032,6891,689
One cut of 50% from week 1395502,6561,656
Ladder 20 / 40 / 60% (weeks 7, 11, 15)10002,6371,637

All schedules are computed in code on one demand-decay model and elasticity 2. The best of the set is the single 30% cut from week 9 (gross profit RUB 1,689,339) but it leaves 103 units; the ladder sells everything but loses on profit. The start-week scan peaks at week 7. The model is illustrative: parameters are fitted on the data of your own category.

Figures

Stock by week under different schedules
Line chart: No markdown: 890 → 286; One cut of 30% from week 9: 890 → 103; One cut of 30% from week 5: 890 → 0; Ladder 20 / 40 / 60% (weeks 7, 11, 15): 890 → 002004006008001,00013579111315171920No markdown, 1: 890No markdown, 20: 286One cut of 30% from week 9, 1: 890One cut of 30% from week 9, 20: 103One cut of 30% from week 5, 1: 890One cut of 30% from week 5, 20: 0Ladder 20 / 40 / 60% (weeks 7, 11, 15), 1: 890Ladder 20 / 40 / 60% (weeks 7, 11, 15), 20: 0units
  • No markdown
  • One cut of 30% from week 9
  • One cut of 30% from week 5
  • Ladder 20 / 40 / 60% (weeks 7, 11, 15)
Data as a table
MeasureNo markdownOne cut of 30% from week 9One cut of 30% from week 5Ladder 20 / 40 / 60% (weeks 7, 11, 15)
1890890890890
2796796796796
3716716716716
4648648648648
5590590530590
6541541429541
7499499343475
8463463270419
9432400207371
10406347154330
11384302109269
1236526370216
1334923037171
143352029133
15324178060
1631415800
1730514100
1829812600
1929211300
2028610300

Stock by week under different schedules. The ladder and early steps empty the stock while no cut leaves a significant leftover.

Gross profit by schedule
Bar chart: No markdown 1,591, One cut of 30% from week 5 1,658, One cut of 30% from week 9 1,689, One cut of 50% from week 13 1,656, Ladder 20 / 40 / 60% (weeks 7, 11, 15) 1,63705001,0001,5002,000No markdownNo markdownGross profit, RUB thousand: 1,5911,591One cut of 30% from week 5One cut of 30% fromweek 5Gross profit, RUB thousand: 1,6581,658One cut of 30% from week 9One cut of 30% fromweek 9Gross profit, RUB thousand: 1,6891,689One cut of 50% from week 13One cut of 50% fromweek 13Gross profit, RUB thousand: 1,6561,656Ladder 20 / 40 / 60% (weeks 7, 11, 15)Ladder 20 / 40 / 60%(weeks 7, 11, 15)Gross profit, RUB thousand: 1,6371,637RUB thousand
Data as a table
MeasureGross profit, RUB thousand
No markdown1,591
One cut of 30% from week 51,658
One cut of 30% from week 91,689
One cut of 50% from week 131,656
Ladder 20 / 40 / 60% (weeks 7, 11, 15)1,637

Gross profit by schedule. The best schedule combines timing and depth; a deep single cut is the most expensive.

Profit and the start week of a 30% cut
Line chart: Gross profit, RUB thousand: 1,310 → 1,5981,3001,4001,5001,6001,7001,800135791113151719optimumGross profit, RUB thousand, 1: 1,310Gross profit, RUB thousand, 19: 1,598RUB thousand
Data as a table
MeasureGross profit, RUB thousand
11,310
21,419
31,512
41,591
51,658
61,716
71,733
81,709
91,689
101,672
111,658
121,646
131,635
141,626
151,618
161,612
171,606
181,602
191,598

Profit and the start week of a 30% cut. The maximum is at week 7: earlier margin is lost, later the leftover cannot sell.

Required and expected volume rise under a cut
Line chart: Required rise, %: 17 → 254; Expected at e = 2, %: 23 → 300010020030010%20%30%40%50%Required rise, %, 10%: 17Required rise, %, 20%: 40Required rise, %, 30%: 76Required rise, %, 40%: 135Required rise, %, 50%: 254Expected at e = 2, %, 10%: 23Expected at e = 2, %, 20%: 56Expected at e = 2, %, 30%: 104Expected at e = 2, %, 40%: 178Expected at e = 2, %, 50%: 300%
  • Required rise, %
  • Expected at e = 2, %
Data as a table
MeasureRequired rise, %Expected at e = 2, %
10%1723
20%4056
30%76104
40%135178
50%254300

Required and expected volume rise under a cut. At e = 2 the expected rise exceeds the threshold at every depth but the gap is narrow; at e = 1.5 a 30% cut no longer pays (rise 71% against a threshold of 76%).

Gross profit bridge of the ladder by step
Waterfall: Full price +1,056, Cut 20% +345, Cut 40% +193, Cut 60% +43, Residual at liquidation price 0, Total 1,63705001,0001,5002,000Full priceFull priceFull price: +1,056+1,056Cut 20%Cut 20%Cut 20%: +345+345Cut 40%Cut 40%Cut 40%: +193+193Cut 60%Cut 60%Cut 60%: +43+43Residual at liquidation priceResidual atliquidation priceResidual at liquidation price: 00TotalTotalTotal: 1,6371,637RUB thousand
  • positive
  • negative
  • total
Data as a table
StepChangeLevel
Full price+1,0561,056
Cut 20%+3451,401
Cut 40%+1931,594
Cut 60%+431,637
Residual at liquidation price01,637
Total—1,637

Gross profit bridge of the ladder by step. The main contribution comes from full price; steps add, and the leftover at liquidation price takes part away.

Specifics by category

Apparel
Demand fades fast toward season end, especially for outerwear: the first step comes earlier than for basics and outerwear leftovers are carried over only when holding cost is low.
Footwear
Sells worse due to broken size runs: a cut will not help with edge sizes, so first-step actions add consolidation of leftovers across stores; a high unit cost increases the loss from a deep step.
Accessories
Shallow and late cuts: basic accessories can be held until the next season while gift sets lose value after the holiday, so they are liquidated right after the peak.
Kids
Size groups go obsolete: leftover of a needed size is lost, so steps are shorter and deeper; school items are cut right after September.
Men
Basics allow carrying leftovers and shallow cuts; a step ladder works on seasonal collections while constant cuts train customers to wait for discounts.
Women
More fashion items and a high sale share: a schedule by role (image items hold price longer), the cannibalization probability within the category is high, so cuts are planned by group, not by model.

Common mistakes

  • Cuts are set looking only at the discount without computing the threshold u = m/(g − m) → volume does not repay margin → compare expected rise with the threshold.
  • The first step is postponed → leftovers grow while demand fades → trigger it by weeks of cover.
  • A deep single cut → margin lost and customers trained to wait → a ladder with transition conditions.
  • Profit is computed only on what was sold → the leftover and its liquidation stay hidden → count the cost of the whole purchase and add the leftover.
  • Windows are not agreed with partners → partner stock is devalued and returns grow → a shared calendar and price support.

Russian specifics

  • Inventory in reporting is measured at the lower of cost and net selling price (FSBU 5/2019): the markdown schedule affects the provision for inventory write-down, so finance receives it in advance. Rule: Order 180n of 15.11.2019: Ministry of Finance Order of 15.11.2019 No. 180n (FSBU 5/2019 Inventories) (rule check not completed).
  • The price on the tag and in a promotion is stated including 22% VAT; the cut is calculated on the price excluding VAT so margin is not distorted by the rate. Rule: Tax Code Art. 164: Tax Code, Art. 164 (VAT rates).
  • Sale terms and dates must match advertising and the retail sales rules including returns; return periods affect the markdown schedule online. Rule: Decree 657 of 30.05.2026: Government Decree of 30.05.2026 No. 657 (rules of retail sale of goods) (rule check not completed).

Open interactively →