Jewelry Brand Inventory Planning: The Cash Trap Guide

Most jewelry founders plan inventory by gut. They reorder bestsellers when stock "feels low" and over-buy a new drop because the vendor minimums were attractive. The result is the category's signature failure mode: cash frozen in a vault. This guide gives you the planning system the public comps imply. A monthly open-to-buy budget, days-of-inventory targets by SKU class, a clean split between core replenishment and seasonal drops, a safety-stock formula tuned for high-AOV pieces, and the compliance rules that quietly shape every buy.

Inventory is the cash trap nobody benchmarks

Inventory is the largest line on a jewelry balance sheet, and it is the one almost nobody benchmarks against the category. The three public comps make the stakes concrete. Movado carries about 186 days of inventory (1.96x annual turns), though it is a watch-led wholesale comp that carries real accounts receivable, unlike the two cash-up-front DTC brands. Signet carries about 172 days (2.13x). Made-to-order Brilliant Earth carries just about 75 days (4.86x). Those numbers come from SEC 10-K XBRL filings, FY2026 and FY2025, computed as days of inventory = 365 / (annual COGS / ending inventory). The takeaway is the part operators miss: the production model, not brand size, sets how long your cash sits in the vault. The stock-heavy model ties up roughly 2.5x more cash per unit of COGS to do the same job.

It is not just turns. Made-to-order collapses the whole cash-conversion cycle. We compute Signet's cash conversion cycle at about 105 days (DIO 172, minus days payable outstanding around 68, plus days sales outstanding around 1), because it finances some inventory through vendor payables. Brilliant Earth's is about 44 days (DIO 75, DPO around 31, DSO near 0), and it collects cash up front on made-to-order so it carries almost no accounts receivable. The production model is a working-capital decision before it is an operations one.

Scale this to a private brand. We use the stock-heavy end (~2x turns) rather than made-to-order's 4.86x because most private DTC brands hold stock rather than building to order, so they resemble Signet more than Brilliant Earth. At roughly 2x turns you hold about half a year of COGS in product at any moment. On a brand doing $5M in COGS, that is about $2.5M of cash sitting as unsold inventory. When we talk to founders running a brand this size, the thing they keep saying is that they only discovered the number when they went looking for a line of credit, and by then inventory days had quietly crept past 180. The signal that this is fixable: Signet grew revenue while holding inventory dead flat. Ending inventory moved just +0.04% year over year ($1,936.6M to $1,937.3M) while revenue rose +1.6% ($6,703.8M to $6,813.6M). Flat inventory on rising sales is the definition of inventory days managed as a hard constraint. For the full vertical picture, see our jewelry financial benchmark report.

Build an open-to-buy budget (the cash-control backbone of a buy)

Open-to-buy (OTB) is the budget that decides how much you are allowed to spend on new inventory in a given period, after accounting for what you will sell, what you already hold, and what you have already committed to vendors. It is the single best discipline for a jewelry brand because it converts a vibe ("the minimums looked good") into a number you either have room for or you do not.

The formula, worked in retail dollars: OTB (retail) = planned sales + planned markdowns + desired ending inventory, minus beginning inventory, minus on-order inventory. Then convert to a cost budget by multiplying by your cost-to-retail ratio. A 2.5x markup is a 40% cost-to-retail ratio. Here is one holiday drop worked end to end.

Two rules make OTB work in practice. First, keep 10 to 25% of the budget uncommitted at the start of the season so you can chase a winner instead of being locked into a buy you placed in July. Second, if the formula comes back negative, you are over-bought for the period: there is no open-to-buy budget, and the job is to work down existing stock, not place an order. When we've struggled with this, what worked was treating a negative OTB as a hard stop, not a number to argue with.

Set days-of-inventory and weeks-of-supply targets by SKU class

A single blended turnover target misleads a high-AOV brand, because a $4,000 signature piece and a $120 core stud should not carry the same number of days. Plan by SKU class instead. The 2026 jewelry operating benchmarks below are industry-reported operating ranges, not audited standards, so treat them as a planning ruler rather than a law. Core replenishable styles target 3 to 4 turns and 90 to 120 days of inventory. High-ticket signature pieces run slower, 2 to 3 turns and 120 to 180 days. New launches and limited drops are planned on forward-weeks-of-supply rather than a days target, because they have no replenishment tail.

The reason to separate weeks of supply from forward-weeks-of-supply is that they answer different questions. Weeks of supply looks backward at recent average sales. Forward-weeks-of-supply looks ahead at the forecast for the upcoming window. For a steady core style they are close enough. For a holiday drop where demand is front-loaded, only the forward number is honest, because last week's sales do not predict next week's. The pattern we see again and again is operators running every SKU off a trailing weeks-of-supply number and then being shocked when the seasonal piece sells out in week two.

Separate core replenishment from seasonal drops

Core and drop inventory fail in opposite directions, so you plan them with different tools. Core SKUs sell every month, so they get a reorder point plus safety stock: you replenish back to a days-of-inventory target whenever stock crosses the trigger. Seasonal drops have a defined window and no long tail, so they get planned by forward-weeks-of-supply and sell-through. You buy to the window, watch sell-through in the first week, and either reorder fast or let it run down. Mixing the two is exactly how a brand over-buys a drop on attractive minimums and stocks out a bestseller in the same month.

Variants are where jewelry over-buying multiplies. Every style fans out into metal, size, and stone combinations, and stocking each one to the same depth manufactures deadstock. Forecast at the style level first, then allocate down to variants using historical mix, not an even spread. If yellow gold is 60% of a style's sales and ring sizes 6 to 7 are 70% of demand, the buy should mirror that, with safety stock concentrated on the fast variants and the slow combinations held thin or made to order. When we talk to founders at this stage, the ones who got unstuck did it by cutting variant depth on the long tail, not by forecasting harder.

Safety stock for high-AOV, low-unit-count SKUs

Safety stock is the buffer that protects you from stocking out when demand or lead time runs hot. The standard method is Safety Stock = Z x sigma_D x sqrt(L), where Z is the service-level factor (1.65 for a 95% service level), sigma_D is the standard deviation of demand, and L is lead time. The output is a unit count, and the practical bands land around 1 to 2x lead-time demand for steady core sellers and 2 to 3x for long-lead or high-stockout-penalty SKUs.

The jewelry-specific twist is that high AOV makes every buffer unit expensive, so over-buffering core SKUs is a direct cash leak you can see on the balance sheet. A 3-unit buffer on a $30 stud is rounding error; a 3-unit buffer on a $3,000 signature ring is $9,000 of frozen cash per SKU. Set the buffer to the service level the SKU actually needs, not to a comfort number. One caveat worth flagging: for high-value SKUs with intermittent, lumpy demand, the naive formula understates the buffer you need, and simulation does a better job than a closed-form calculation. If a long-lead signature piece sells in unpredictable bursts, model it rather than trusting the square-root formula alone.

Deadstock, write-downs, and the compliance constraints on your buy

Deadstock is the back end of every over-buy, and the discipline is markdown cadence: review aged inventory on a fixed schedule, mark it down before it becomes a write-down, and feed the lesson back into next season's OTB. We did not find an audited jewelry deadstock benchmark to anchor a percentage to, so we won't invent one. Treat it qualitatively: if a style has not turned in two planning cycles, it is a markdown candidate, full stop.

Several compliance rules shape the buy directly, and they are planning inputs, not afterthoughts. (This is general information, not legal advice.) GAAP carries inventory at the lower of cost or net realizable value (ASC 330), so a fall in gold prices can force write-downs on high-metal-content finished goods and work-in-process: metal-price exposure is a real input to how much aged high-karat stock you should hold. The FTC Jewelry Guides (16 CFR Part 23) require separate SKUs for natural versus lab-grown versus treated stones and for metal fineness, which multiplies your SKU count whether you like it or not. Jewelers block insurance caps how much high-value inventory you can safely hold per location and safe, which caps on-hand stock at the top end. And conflict-minerals rules (Dodd-Frank Section 1502, covering 3TG including gold) plus the Kimberley Process for rough diamonds can extend supplier-qualification and reorder lead times, which feeds straight back into your safety-stock math. Each one either lengthens lead time or caps on-hand stock, so each one is a direct input to OTB and safety stock.

For the broader financial framework behind these moves, see our guide to how inventory financing works, and the interim CFO services overview if you want a second set of eyes on your buy.

Treat inventory days as a hard operating constraint, not a buying preference. The public comps prove the lever works: the disciplined brand grew sales while holding inventory dead flat, because it bought to an open-to-buy budget and a days-of-inventory target instead of to attractive vendor minimums. The cash you free is the cash that was going to fund your line of credit.

Sources and methodology

Jewelry financial benchmark (internal grounding). This guide reuses vertical-specific data points from our jewelry financial benchmark report: days of inventory 75 to 186 across Brilliant Earth, Signet, and Movado; inventory turns 1.96x to 4.86x; made-to-order turning roughly 2.5x faster than stock-heavy; and the roughly $2.5M of cash in product on a $5M-COGS brand at ~2x turns. All trace back to SEC 10-K XBRL primary data.

SEC EDGAR (primary). Annual 10-K XBRL financials were pulled directly via SEC EDGAR, accessed June 14, 2026. Signet Jewelers (CIK 0000832988), FY ended 2026-01-31: revenue $6,813.6M, COGS $4,119.0M, ending inventory $1,937.3M (prior year $1,936.6M), accounts payable $767.0M. Brilliant Earth (CIK 0001866757), FY ended 2025-12-31: revenue $437.5M, COGS $186.0M, ending inventory $38.3M (FY2024 base), accounts payable $15.7M, near-zero accounts receivable. Movado (CIK 0000072573): revenue $671.3M, COGS $307.7M, ending inventory $156.7M, carried from the benchmark report.

Ratios (computed). Days of inventory = 365 / (COGS / ending inventory). Inventory turns = COGS / ending inventory. Days payable outstanding = accounts payable / COGS x 365. Days sales outstanding approximated at 0 to 1 (both brands collect at or near point of sale). Cash conversion cycle = DIO + DSO - DPO. Signet: DIO 171.7, DPO 67.9, DSO ~1, so CCC about 105 days. Brilliant Earth: DIO 75.1, DPO 30.9, DSO ~0, so CCC about 44 days. Brilliant Earth's DIO uses the FY2024 inventory base and its DPO uses FY2024 accounts payable against FY2025 COGS, so read the ~44-day cash conversion cycle as an approximation.

Operating benchmarks (web). The 2026 DTC jewelry inventory ranges (turns 2 to 4x, days-of-inventory 90 to 180, weeks-of-supply 13 to 26, forward-weeks-of-supply 8 to 20, OTB keep 10 to 25% uncommitted, safety stock 1 to 2x core and 2 to 3x long-lead) are synthesized from 2026 jewelry inventory-KPI guidance (Valigara, Useryze, JCK). These are industry-reported operating ranges, not audited standards, and are kept distinct from the SEC primary numbers throughout.

Open-to-buy and safety-stock methods. The OTB formula (planned sales + planned markdowns + desired ending - beginning - on-order = OTB at retail, then x cost-to-retail ratio) follows Faire and Shopify retail-planning guidance. The safety-stock formula (Safety Stock = Z x sigma_D x sqrt(L)) follows ISM guidance, with the simulation caveat for high-value, intermittent-demand SKUs drawn from INFORMS simulation work.

Compliance. Inventory-planning constraints (GAAP ASC 330 lower-of-cost-or-NRV, FTC Jewelry Guides 16 CFR Part 23, jewelers block insurance limits, Dodd-Frank Section 1502 conflict minerals, and the Kimberley Process) were sourced from US regulatory research, June 2026. This is general information, not legal advice.

Limitations. A jewelry-specific Storeleads vertical cut and the founder-call corpus were not available this run; per-store inventory and SKU counts are not fabricated, and operator-voice lines are drawn from documented, anonymized founder pains. No audited jewelry deadstock or markdown-rate benchmark was found, so the deadstock section is kept qualitative.

Frequently asked questions