How the Economy Moves Through the Supply Chain: The Bullwhip Effect

In 2020 and 2021, the world economy delivered a lesson in supply chain dynamics. Consumer spending shifted from services to goods, and the shift was modest compared with what followed upstream: retailers over-ordered, distributors over-ordered on top of them, factories ran out of components, shipping rates multiplied, and semiconductors went on allocation. By 2022 and 2023, many of the same companies were writing down excess inventory as demand normalised.

None of this required consumer demand to move very much. It is the result of a mechanism every supply chain contains, and which turns small changes in the economy into large swings in production, inventory and cash.

The bullwhip effect

The term became widely known through a 1997 article by Hau Lee, V. Padmanabhan and Seungjin Whang, who described what Procter & Gamble had found in its Pampers supply chain: consumption of nappies by babies was steady, but the orders that distributors placed with P&G swung widely, and P&G's own orders to its suppliers swung even more. Each tier was acting rationally. Together they amplified the signal.

In the simulation, consumer demand rises 15% for three months and then returns to normal. Each tier forecasts from the orders it receives and tries to rebuild its stock to cover the lead time. The retailer's orders swing twice as much as demand; the distributor's almost four times; the factory's almost six times, with a peak 80% above normal followed by weeks with no orders at all. The factory sees a boom and a bust where the consumer saw a modest, temporary rise.

Four causes, all ordinary

• Forecasting from orders, not demand. Each tier reads the orders of the tier below as if they were demand, and adds its own reaction to the noise.
• Lead times and safety stock. Longer lead times mean more stock in the pipeline, and every change in the forecast is multiplied by the lead time when the target is rebuilt.
• Batching. Full trucks, minimum orders and monthly purchasing cycles turn smooth demand into lumpy orders.
• Shortage gaming and price swings. When supply is rationed, customers order more than they need; when prices are expected to rise, they buy ahead. Both create phantom demand that disappears later.

Why it matters to finance

The further a company sits from the final consumer, the more cyclical its revenue, its capacity utilisation and its working capital. Industrial suppliers, component makers and raw material producers experience every slowdown as a recession and every recovery as a shortage. For FP&A, this means that the economic indicators in the forecast should be read through the company's position in the chain: a one-point slowdown in consumption can mean a ten-point fall in orders for a business three tiers upstream.

It also means that the order book is a poor guide at turning points. Orders surge just before the destocking begins, and collapse while final demand is already recovering. Companies that budget on order intake alone tend to add capacity at the top of the cycle and cut it at the bottom.

Dampening the whip

1. Share point-of-sale and inventory data with customers and suppliers, so that each tier forecasts from real consumption.
2. Shorten lead times, the most powerful lever, because the amplification grows with them.
3. Smooth the replenishment rules, reacting to only part of the gap between target and actual stock each period, as planners call dampening.
4. Allocate on past sales in shortages, not on current orders, which removes the incentive to over-order.
5. Watch the leading indicators of destocking: the ratio of customer inventories to their sales tells when the next fall in orders is coming.