Whether it is a storm, fire, strike, management failure or trade conflict, the risks facing companies are numerous. There is also danger if things go badly for suppliers and customers. If a supplier that is difficult to replace fails or an important customer runs out of money, then there is a crisis. The stronger the division of labor and specialization in production, the more that companies are closely interwoven and interdependent in a network of supply relationships. An interruption in the production chain can be all the more disastrous. Contagion in the value chain can affect entire industries and exacerbate economic downturns. In order to strengthen crisis resilience, companies can take precautions, for example, with warehousing, diversification of suppliers and sufficient capital reserves.
Companies are closely intertwined through a large number of supply relationships. As a result, a supplier’s financial distress can quickly turn into a business risk for their customers and even for an entire industry. The shocks and their effects can be so severe that they affect entire industrial sectors and thus exacerbate the economic downturn. So far, economic research has mainly dealt with the interdependence between special industries and their influence on the overall economy. Little is known yet how the complex value chains depend on the interdependencies between individual companies.
Shocks and supplies
Every company is part of a more or less complex network of suppliers, sub-suppliers and customers. Companies supply each other and sell their products on a common market. If one of these companies has to interrupt their production or there are delivery bottlenecks for other reasons, their customers have to react. This response is determined by two factors.
On the one hand, companies can compensate for a large part of such shocks through their own precautions and adjustments in their production. The distribution of purchases among different suppliers and warehousing are examples of this. Even in the event of a major interruption and delivery bottlenecks and price increases, many companies are flexible enough to adapt their production or to switch to other suppliers.
On the other hand, with long-term supply contracts, the exclusivity of a supplier or patents can make such adjustments difficult. In this case, shocks in a specific company spread faster and more strongly in the production network and can build up each other. If a company gets into trouble, its customers and suppliers will feel it. The larger the company concerned, the more rigid the production, or the more branches the network has to other companies, the more the shock spreads and the greater the impact on the economy as a whole.
Connections that existed at the beginning and at the end of the period have a dark background. All other relationships were either broken off before 2000 or only started after 1995. The distance between two points shows the geographical distance between the companies, the thickness shows the strength of the relationship.
According to data, the complexity of a supplier and customer network reflects the consequences of a shock. Not all companies have the same importance in the value chain.
A failure of production at one of the nodes, i.e., a heavily networked control facility, affects a large number of companies in the network. It’s also known that connections to geographically close companies are preferred. A production shock therefore typically has a particularly severe impact on customers and suppliers in the vicinity.
It is not always easy for economists to establish these chains of effects and identify them empirically. A multitude of influencing factors often makes it almost impossible to isolate the direct effect on a single company. Events that affect a company alone are rare and difficult to pinpoint. You can use a special trick to do this. For the United States between 1978 and 2013, they look at how local natural disasters such as fires, storms, flooding or heavy snowfall affected companies and their upstream and downstream partners. Then they compare the course of business with similar companies that did not suffer directly from disaster, but are intertwined with the affected companies through their network. The stronger the shock and the more significant the role of a company in the network, the stronger the effects of such a shock should be felt by other companies.
In the event of a local shock such as a natural disaster, the company's sales growth falls by an average of 5 percentage points, and for customers by 2 to 3 percentage points.
Estimating by disaster
First, I estimate that the revenue growth of a company hit by a natural disaster will decrease by an average of 5 percentage points. This should have a direct impact on customer sales. I also estimate that such a shock reduces customer sales by 2 to 3 percentage points. This corresponds to a decrease of 25% compared to the average growth rate over three quarters.
The vertical line indicates the time of the natural disaster and the vertical axis the loss in the share value in percent. In the first month after the shock, the market value of the supplier concerned is reduced by an average of 3%, while that of the customer is reduced by up to 1%, i.e., by a third of that.
After a production stop, the stock market value of an affected company falls by an average of 3%. In a company network, this leads to a loss of value for customers of up to 1% of the market value.
In addition, a shock also spreads horizontally to other companies that are not directly affected themselves, but supply customers who also buy from the suppliers directly affected. Such that other, not directly affected, companies are also infected by having to accept up to 4 percentage points lower sales growth via their customers. A company’s crisis does not only spread vertically along a value chain, but also horizontally between suppliers at the same processing level. An entire network or an entire industry can become infected.
My estimates are limited to the effects of natural disasters. With my approach, however, I can identify shocks that are initially limited to individual companies and can thus filter out the propagation effects separately. As I emphasized, it is very likely that I will generalize my results to other interruptions in the value chain, e.g., as a result of strikes, management changes or trade disputes. It does not depend on the cause of the production downtime, but on the importance of the goods and services that a customer receives from its suppliers. The more difficult it is for a customer to replace a failing supplier, the more they will be affected by an interruption in the previous processing stage. This applies in particular to industries that are subject to strict regulation and in which patents are held by a few providers. This can significantly increase the contagion effects of company-specific shocks.