Short vs. long-term procurement contracts when supplier can invest in cost reduction

https://doi.org/10.1016/j.ijpe.2020.107652Get rights and content

Abstract

In dynamic markets where cost of components changes fast, buyers typically auction off regular short-term contracts to fully leverage supplier competition in each period to continuously source from the lowest-cost supplier. However, too much competition through short-term contracts does not incentivize the incumbent supplier to make relation-specific investments in reducing costs, as future business is not assured. We investigate this trade-off, between leveraging supplier competition in each period versus incentivizing incumbent’s investment, with a stylized two-period model in which the buyer decides whether to auction off short-term contracts in each period or auction off a single long-term contract spanning both periods. In both cases, we characterize the optimal incumbent supplier’s investment, the suppliers’ equilibrium bidding strategy and the buyer’s expected cost. Our analysis shows that the supplier always invests more in a long-term contract. However, the buyer’s cost depends on supply base size: it prefers short-term contracts for large supply base size, otherwise it prefers long-term contract. Moreover, we find that system cost is typically lower with short-term contracts and that the suppliers are always better off with short-term contracts. Finally, adding non-discriminatory or discriminatory reserve prices to our model does not fundamentally modify the trade-off, but we find that auctions with discriminatory reserve price are better at balancing this trade-off compared to long or short-term contracts.

Introduction

Procurement managers frequently use electronic reverse auctions to source standardized and well specified items like memory circuits, printed circuit boards, power chords and cable connectors. In dynamic markets, like consumer electronics, that see frequent evolution of technology, suppliers’ cost of producing these items can change from period to period. For instance technological investments in either new production technology or new production material can change manufacturer’s cost from one period to the next, often in uncertain ways. As an example, a semiconductor manufacturer might invest in new material or a new production process (technology) that it uses for fabricating semiconductor chips. Indeed, the new material (or production process) would allow the chips to be more efficient or faster which is necessary for the manufacturer to maintain its market position. However the manufacturer can be impacted in uncertain ways in working with new technology or material, e.g., unknown impurities in the manufacturing environment could decrease production yield (hence increasing cost) or on the other hand the new material might be more robust to environmental impurities thus increasing the yield (and decreasing the cost). Supply–demand dynamics could further change a supplier’s cost for a buyer from one period to another in uncertain ways, specially when supplier makes such technological investments. For instance another buyer might value production with supplier’s new technology more or less, which would change the opportunity cost (supplier’s) of dedicating its capacity (assuming fixed capacity) to a particular buyer. In order to discover the current best market price, among all the potential suppliers whose cost might change from one period to the next, procurement managers often organize reverse auctions periodically (see Carbone, 2004).

On the other hand, by assuring long-term business to a supplier, a buyer can gain cost savings from supplier’s relation-specific investments (in form of time, effort and resources) that are idiosyncratic to the buyer (see Cisternas and Figueroa, 2015). These investments are dedicated to a particular buyer. Moreover, these investments are made by the supplier only after it gets business from the buyer, i.e., these investments cannot be replicated by other potential suppliers (who do not have buyer’s business) because they either require knowledge about buyer specific needs or require access to buyer’s specialized programs both of which can only be gained by working with the buyer. We provide three different examples of investments falling into this category. (1) Investments made by supplier in developing buyer-specific software (see [9]). For instance, a supplier might need to develop additional buyer-specific ERP modules that can be integrated into suppliers’ broader ERP system. Such an additional module would result in better production planning at supplier’s end and consequent reduction in production and inventory costs of the supplier, despite the fact that it would initially be expensive and time-consuming for the supplier to invest in such an alignment. A better integration can further be achieved through investments in human capital (e.g. manufacturing engineers developing knowledge about the buyer), or in warehouse investments dedicated to a specific buyer, notably in lowering future inventory and transportation expenses (Dyer and Ouchi, 2002, Williamson, 1983). (2) Buyer-specific production cost reduction could also occur due to supplier’s production learning — which requires not only sufficient production volume but also engineering trials that use expensive production capacity at supplier’s end (along with employee time and effort) for controlled experiments in optimizing production processes specific to the buyer’s order fulfillment (see Terwiesch and Bohn, 2001). (3) Some large buyers (e.g. Ikea or Walmart) have their own energy efficiency assessment programs which determine potential of improvement in energy efficiency of their suppliers, which, if realized, could allow supplier to reduce its production cost. Typically, the buyer would offer its suppliers access to its energy efficiency program via a third-party auditor (for which the supplier pays) who identifies the potential for improvement and the supplier then invests in realizing gains (Nguyen et al., 2018). Note that the common feature of all these examples is that supplier’s relation-specific investment occur only after gaining experience (and business) of working with the buyer; the investments almost always result in production efficiency specific to the buyer and finally, the buyer’s contribution (in terms of cost and effort) towards these relation-specific investments is negligible in comparison to supplier’s contribution, i.e., the relation-specific investments that we consider in this paper are primarily being driven by the supplier.1

Indeed, these relation-specific investments are different from technological investments. The former are made only by the incumbent whereas the latter is made by all the suppliers (incumbent and non-incumbents), i.e., irrespective of whether they get buyer’s business or not. In this paper, we focus on these relation-specific investments made by the supplier towards lowering its buyer-specific production cost after it gets business from the buyer. The level of these relation-specific investments depends on the continuity of business that the supplier anticipates from the buyer: intuitively, higher levels of relation-specific investment made by a supplier are riskier in short-term contracts (compared to long-term contract) as it has to compete again for buyer’s business in a market where cost of suppliers can change from one period to the next. Thus, a buyer can better incentivize its supplier towards making higher relation-specific investments by offering a longer-term contract to the supplier which in turn can benefit the buyer from the cost advantages that the supplier can offer the buyer.

Thus, there exists a trade-off in buyer’s sourcing strategy: it can either gain higher cost savings derived from relation-specific made by the supplier by assuring long-term contracts to suppliers or it can fully leverage supplier competition in each period by only offering short-term contracts.

To capture this trade-off we present a stylized 2-period model in which symmetric suppliers independently draw fresh costs in both periods (to reflect the changes in suppliers’ cost from periodic investments in new technology). The buyer has two sourcing options: (1) it can organize an open descending (or a sealed second-price) auction at the beginning of each period and give a short-term contract, spanning a single period, to the lowest bidder in each period, or (2) it can organize a single open descending (or a sealed second-price) at the beginning of the first period and give a long-term contract, spanning both the periods, to the lowest bidder. To capture relation-specific investments made by the supplier that has won the first-period auction (from now on the incumbent supplier, in contrast with a first-period auction loser that we define as a non-incumbent supplier), the model assumes that the incumbent supplier (in either the single or the two-auction setting) can make relation-specific investment which stochastically reduce the supplier’s second-period cost. The level of investment made by supplier in either auction setting is a decision variable.

Intuitively, risk of losing buyer’s business soon (in a short-term contract) would disincentivize supplier from making relation-specific investments. For similar reason, greater competition would increase the risk of investment in short-term contract. In fact we find that the difference in investments between the long and the short-term contract is increasing with the supply base size. However, with greater supply base size the buyer increases its chances of drawing a lower cost in both periods through short-term contracts as compared to long-term contracts. Thus, the buyer’s decision on long or short-term contracts critically depends on its supply base size.

To quantitatively compare the long-versus-short term sourcing strategies, we next investigate buyer’s cost in both auction settings. For this, we characterize the equilibrium bids that suppliers would submit in both auction settings. As the incumbent supplier can make relation-specific investments, hence, the equilibrium bids must take into account the cost of investment and the resulting cost improvements that the suppliers would gain. We find that typically the buyer would prefer short-term contracts for large supply base size and would prefer long-term contract for smaller supply base size. However, the buyer’s preference for long-term or short-term contract has a more nuanced dependence on the supply base size. This is because in a two-period setting, the investment in short-term contracts drops to zero beyond a certain supply base size, at which point the long-term contract can become more preferable for the buyer. Moreover we show that suppliers are always better off, in expectation, by participating in auctions that give away short-term contracts. Finally, we find that the system cost (i.e., the sum of production cost and investment) is lower with short-term contracts than with longer-term contracts.

We also numerically investigate how our findings are affected when the buyer can optimally set non-discriminatory, and discriminatory, reserve prices to lower its procurement cost (for instance when the buyer has access to an inexpensive outside option). Neither non-discriminatory reserve prices, nor discriminatory reserve prices change the fundamental trade-off between leveraging period-to-period supplier competition versus incentivizing incumbent supplier’s relation-specific investment. However, we find that discriminatory reserve prices are a better tool for balancing this trade-off in comparison to short or long-term contract. More specifically, with discriminatory reserve prices the buyer can organize an auction in each period and thus leverage period-to-period supplier competition, as in the short-term contracts case, but at the same time it can discriminate in favor of the incumbent supplier in the second auction to incentivize a high investment from the incumbent, as in the long-term contract case. We find that a contract with optimally set discriminatory reserve prices always performs better than both the long-term and the short-term contract cases. We further find that buyer’s expected cost in the contract with discriminatory reserve prices is closer to this in a long-term contract when supply base size is smaller, and is closer to this in short-term contracts when supply base size is bigger, consistent with our previous findings. However, with discriminatory reserve prices the buyer has sufficient control over its cost as a result of which its expected cost decreases monotonically in the supply base size.

The rest of this paper is organized as follows: we first review the procurement auctions literature in Section 2. Then we introduce the model in Section 3 and determine supplier optimal investment in Section 4. Next, in Section 5 we characterize and compare the buyer’s expected cost, suppliers’ surplus and system cost in both settings. Finally, we discuss the impact of reserve prices on our model in Section 6, and we present the conclusion in Section 7. All the proofs are presented in Appendix.

Section snippets

Literature review

Our work relates to the literature in procurement auctions that investigates supplier investment. Some papers completely focus on the supplier investment decision, observing how it is affected by specific factors, like the auction format (Arozamena and Cantillon, 2004) or the commitment to a mechanism (Dasgupta, 1990, Piccione and Tan, 1996). Unlike these papers, our objective is not to observe what affects supplier investment levels, but rather to determine the influence that the supplier

Model

We consider a 2-period model in which a buyer needs to procure one unit of a homogeneous good in each period. Without any loss of generality, we normalize the buyer’s demand to one unit per period. The buyer organizes dynamic open descending price reverse auction(s) to procure its demand for the two periods (we will investigate later whether the buyer should organize a single auction to procure both periods demand or should organize a separate auction for reach period). The open descending

Incumbent supplier’s relation-specific investment

In this section, we investigate the implications of the incumbent supplier making relation-specific investments which reduce its per-unit cost for a specific buyer. As mentioned in Section 1, these are for example supplier investments in improving its energy efficiency, in integrating its operations to the buyer’s supply chain, or in production learning (e.g. time, effort and resources required) for engineering experiments conducted to optimize production processes specific to buyer’s order

Buyer’s expected cost and suppliers’ surplus

In this section, we characterize the buyer’s expected cost in both the single and the two-auction settings. For this, we first characterize the equilibrium bidding strategies of the suppliers in both auction formats and then characterize the difference in the buyer’s expected cost between the single and the two-auction settings. We also compare the single-auction and two-auction setting for differences in supplier’s surplus and system cost.

Reserve prices

This section investigates how the results would change if the buyer uses optimally set reserve prices in both auction settings. In a classical single shot reverse auction, reserve prices are used to better manage buyer’s procurement cost when the buyer has access to a viable outside option (which could be a non-participating supplier or buyer’s in-house production). In our single-auction setting, the buyer’s rationale for setting optimal reserve price remains similar to the classical single

Conclusion

In dynamic markets where supplier technology evolves fast, a buyer regularly auctions off new contracts to stay abreast of the best price that it can receive from its supply base. However, short-term contracts might not incentivize the supplier to make relation-specific investments that can reduce its production cost; because such an investment would be risky for the supplier as it is not sure of winning consequent auctions. On the contrary a longer-term contract can better incentivize the

References (25)

  • CisternasG. et al.

    Sequential procurement auctions and their effect on investment decisions

    Rand J. Econ.

    (2015)
  • DasguptaS.

    Competition for procurement contracts and underinvestment

    Internat. Econom. Rev.

    (1990)
  • Cited by (0)

    View full text