Graves 1999 combined variability
Combines demand variability and lead-time variability into one safety stock formula instead of approximating them separately.
Replenishment Engine
Reorder points you can defend. Min/max levels backed by research.
Stocklyst calculates safety stock, reorder points, and min/max levels from demand and lead-time behavior instead of static guesswork. The goal is not more settings. The goal is better replenishment decisions at every branch.
Numbers you can defend
6 formulas
Graves, Syntetos-Boylan, Cornish-Fisher, and Nahmias components are there because the number needs credibility.
Different demand gets different treatment
4 classifications
Smooth, erratic, intermittent, and lumpy items do not all behave the same, so the engine does not treat them the same.
Calculated where stock decisions happen
Per branch
Every location gets its own calculation output instead of inheriting one generic company-wide threshold.
Protects perishables
Shelf-life aware
Max levels are capped where inventory ages out, so overstock does not quietly turn into expiry waste.
01 — Calculation engine
Why the numbers deserve trust
The formulas are named because the citations are the proof. Each component solves a specific inventory failure mode that simpler threshold systems usually ignore.
Syntetos-Boylan 2005 classification
Sorts items into Smooth, Erratic, Intermittent, or Lumpy patterns so calculation logic matches how demand actually behaves.
Cornish-Fisher skewness correction
Adjusts the z-score when demand is not normally distributed, which is common in real inventory histories.
Nahmias 1982 perishable cap
Prevents max levels from climbing above what can be sold before expiry on shelf-life-constrained items.
Nahmias 1994 anti-death-spiral
Prevents the engine from falsely learning lower demand from periods where the item was actually stocked out.
Zero-inclusive standard deviation
Keeps zero-demand periods in the variability math so intermittent items are not treated like clean linear sellers.
02 — Engine output
From formula to replenishment decision
The engine runs continuously for every item at every branch, producing safety stock, reorder points, and min/max levels without forcing the team to maintain static thresholds by hand.
What the output gives operators
- Know what to reorder at each branch instead of relying on one blanket company rule
- Use safety stock buffers that account for both demand variability and lead-time uncertainty
- See why intermittent and bursty items get different treatment from smoother sellers
- Cap max levels where shelf life turns overstock into waste
03 — Where it helps most
Where manual min/max rules usually break down
The strongest fit is any operation where one static threshold keeps failing across branches, intermittent demand, or perishability.
Retail chains
Each store gets branch-specific reorder points based on its own demand pattern instead of inheriting one company-wide threshold.
Perishable goods
Shelf-life-aware caps reduce waste for bakeries, grocers, and food distributors where overstock turns into expiry.
Seasonal and intermittent demand
Burst-driven items like spare parts and holiday stock get safer treatment than a generic moving-average rule can provide.
Growing businesses
As new items and branches are added, the engine scales automatically instead of forcing manual min/max setup everywhere.
04 — FAQ
Frequently asked questions
How does Stocklyst calculate safety stock?
Stocklyst uses the Graves 1999 combined variability formula, which accounts for both demand variability and lead time variability in the same calculation rather than treating them independently.
What demand patterns does the engine classify?
The engine uses Syntetos-Boylan 2005 classification to categorize items into Smooth, Erratic, Intermittent, or Lumpy demand patterns. Each pattern receives tailored calculation parameters.
How does the engine handle intermittent demand items?
Intermittent and Lumpy items receive adjusted calculation parameters, including stronger safety factors and demand treatment better suited to irregular ordering behavior.
What is Cornish-Fisher skewness correction?
Standard safety stock math assumes normal demand, but real inventory demand is often skewed. Cornish-Fisher adjusts the z-score so the engine does not under- or over-stock based on a bad assumption.
How does perishable inventory capping work?
For shelf-life-constrained items, max stock is capped so the oldest units can realistically sell before expiry, reducing waste from overstocking perishables.
Does the engine protect against death-spiral stockouts?
Yes. The Nahmias anti-death-spiral guard prevents stockouts from collapsing demand history and causing the engine to keep lowering reorder points incorrectly.
Ready to stop
6 peer-reviewed formulas · ADI 1.08 · Confidence 92%