Please select your location to see content specific to your country

x

Bullwhip effect explained

Covid-19: “In Trouble, We Double” Bullwhip Effect Explained

Covid-19: “In Trouble, We Double” Bullwhip Effect Explained

Due to COVID-19, border restrictions and the ‘circuit breaker’ measures in March and early April, Singapore consumers were panic buying. Egg prices temporarily rose and shelves were emptied, raising questions on whether Singapore’s food supply would be affected. Distributors felt pressured to import more when supermarkets were asking for more. Mr Ma Chin Chew, Chief Executive of Local Egg Farm N & N Agriculture, said it imported more eggs during the circuit breaker period when demand spiked.1

However, 3 months later, the situation has reversed. Suppliers are now facing an oversupply of eggs, causing egg prices to plummet by at least two cents each as reported by several local newspapers. Kim Hock egg distributor1, as a result, had to throw away 250,000 eggs imported from Thailand when customers and supermarkets complained that the eggs had gone bad. Farmers and distributors have also been cutting supply.

This is a classic example of the bullwhip effect. For non-supply chain people: the bullwhip effect is a phenomenon where relatively small changes in demand at the end-consumer level lead to (unreasonably) large orders and inventory levels at the manufacturer.

There are a few explanations as to why this is happening; the main contributors to the bullwhip effect are due to a lack of information sharing in the supply chain, long lead times from manufacturer to retailer and the need for each node in a supply chain to keep its inventory and buffer stock. The human component of overreacting to issues (“in trouble, we double”) is also one of the biggest reasons the bullwhip effect exists. Despite it being a well-researched phenomenon, in practice, the bullwhip effect is still very uncommon.

Strategies to mitigate the bullwhip impact

Recently, Slimstock played a simulation game with participants from the supply chain industry to access the impact of the bullwhip effect and tested a few effective strategies to mitigate the impact. In this game, we simulated a supply chain of hand sanitizer facing a sudden bump in demand and let the participants take on a role in the supply chain (one person played as a retailer, wholesaler, distributor or manufacturer). We played several rounds with different configurations to test the effectiveness of different strategies.

The game started with a traditional supply chain. When we say “traditional”, we meant a supply chain with no communication between the different roles, other than the order they place with each other. On top of that, the supply chain had long lead-times, making it difficult to respond effectively and quickly to a sudden spike in demand. The goal of the simulation game was to achieve the lowest total cost across the supply chain (with backorders incurring costs and, on the other hand having inventory incurring costs. Hence participants had to balance the cost of lost sales/late shipments vs. the cost of holding inventory).

The first game

Demand first game
Figure 1: Demand throughout the first game – participants only knew demand in the week it occurred, and the true demand was only known to the retailer who had no way of sharing information with the other teams.

During the first game, we learned a few things. We mainly found out that in general manufacturers were running the risk to end up with high inventory levels at the end of the crisis. Whereas wholesalers run a very large risk of backorders. Retailers run the lowest risk of incurring heavy costs.

If we look at figure 2 below, we notice the order size fluctuation throughout the 20 rounds. We see the manufacturer preparing early for the crisis, but the distributor and wholesaler do not start pulling any product in anticipation (participants did know a crisis was coming, they did not, however, know exactly when and how severe). The retailer responded to the crisis as it occurs. As they know what the actual demand was, they could respond the latest in terms of ordering. They only need to react as it is happening.

Order fluctuation by role
Figure 2: Order fluctuation by role throughout the first game

If we look at the inventory fluctuation throughout the game for each of the roles, we see that the anticipation of the manufacturer backfired. As they did not have any information on true demand, they kept producing too long and kept building up inventory up until week 9, long after the spike in demand was over.

Inventory fluctuation by role
Figure 3: Inventory fluctuation by role throughout the first game

Overall, the wholesaler had built up a significant number of backorders which incurred them a very high cost. In real life, this is likely to occur after this COVID-19 crisis as well because many retailers have a major power advantage over their wholesalers (and distributors). They are close to demand, very large, and keep very minimum inventory compared to the rest of the supply chain. As a community, this gives us something to think about, as we’ll likely see many wholesalers and distributors incurring large costs, and we’ll see many manufacturers produce until long after this crisis will be over. this might lead to large periods of no production at all and could cause severe effects on employment and stability at the manufacturing level.

The second game

For the second game, we made a lot of changes to the setup of the supply chain. Of course, we modelled the crisis slightly different so participants had no idea what and when the crisis would come. On top of that, we let participants openly collaborate and communicate and we halved the lead-times between the links in the supply chain allowing them to respond faster. This greatly improved the total cost incurred by the supply chains, even at the demand that was on average 25% higher than in the previous game, the total costs were cut by as much as 75%.

 

Order fluctuation by role throughout the second game
Figure 4: Order fluctuation by role throughout the second game

In figure 4, we observed the order fluctuation by the role in the second game. We noticed the manufacturer, again, anticipated a crisis. This time, however, the production soon stabilised right after the crisis. This is because of sharing information through the whole supply chain. Similarly, the manufacturer was running the highest risk of being left with a large chunk of the inventory at the end of the crisis, but the risk was significantly lower, as you can observe in the inventory progression in figure 5 below.

 

Inventory fluctuation by role throughout the second game
Figure 5: Inventory fluctuation by role throughout the second game

At the peak of its inventory the manufacturer only had roughly two weeks of products in stock. A lot less than the 10 weeks peak of the inventory we witnessed in the first game.

What did we learn

Supply chain management is not easy. It involves many parties and uncertainties. However, we’ve shown that collaborating, sharing information, and or focusing on getting shorter lead-times to offer a good way to reduce total supply chain costs. Not only does it allow supply chains as a whole to react faster to a crisis. It also lowers the cost for everybody involved.

Speak to an expert about your inventory challenges

Ryan Square

Ryan Ching

Call_icon mail_icon LinkedIn_icon