Effects of demand uncertainty reduction on the selection of financing approach in a capital-constrained supply chain

Highlights

• We investigate the impact of DUR on the selection of financing approach.

• Our investigation covers both exogenous and endogenous wholesale pricing cases.

• The increase of DUR does not always benefit the supply chain players.

Abstract

This study investigates how demand uncertainty reduction (DUR) affects the decisional dynamics within a supply chain, which comprises a supplier and a capital-constrained retailer, who chooses between bank credit and trade credit financing. A comprehensive scenario analysis suggests the retailer should accept trade credit when DUR is high, trade credit risk premium is moderate, and wholesale price is exogenous and low. However, the retailer should adopt trade credit only when both DUR and production cost are not high, and wholesale price is set endogenously. We further relax the assumption on the bank’s risk attitude and find most results still hold.

Introduction

Advanced data analytics techniques incorporating data mining and machine learning are effective in reducing uncertainty in customer demand (Analytics Insight, 2020); and they have been increasingly applied in firms across industries. Amazon, for instance, uses sophisticated machine learning-based data analytics to generate accurate demand forecasts for use by firms worldwide in the form of service of Amazon Forecast. Such forecasts are found to be much more accurate compared with the traditional prediction methods (Business Wire, 2019). As reported by Amazon Forecast, the leading performance marketing company Heroleads succeeds in increasing its demand forecast accuracy to 99%; the global refrigerator manufacturer Arneg S.p.A. achieves a 91% accuracy in predicting the energy needs of its refrigerators; and Omotor can provide demand forecasting services with an accuracy improvement of more than 50%. This increase of prediction accuracy has been termed demand uncertainty reduction (or DUR hereinafter) by Li and Petruzzi (2017).

The effect of DUR on supply chain management has received tremendous attention by researchers including Chong et al., 2017, Arunachalam et al., 2018, Govindan et al., 2018. Apparently, DUR will help firms to reduce the mismatch between supply and demand, thus improving their operational and inventory decisions (Xu et al., 2010, Fleischhacker and Fok, 2015, Zhai et al., 2020). Unfortunately, from the financial perspective, the importance of DUR has not received sufficient attention that it deserves. Pezza (2011) obtains the views of 145 corporate executives and senior financial managers on the effects of demand uncertainty on firms’ financial performance. He concludes that cash position is the most affected, as demand uncertainty exerts both internal and external pressures on firms’ operations. Internally, greater inventory investment is required to buffer against stock-outs and externally, uncertain sale revenue makes it more difficult for firms to maintain their financial viability. Therefore, effective DUR is desired especially by capital-constrained firms to maintain robust cash flow and hence reducing the risk of default, a critical factor employed by the corresponding lending agents or institutions in determining the interest rate (Lai et al., 2009). Wang et al. (2019b) have shown that increased demand uncertainty may motivate the lenders to charge a higher interest rate and that may further influence the decisions and performance of participants in supply chain and lending market.

To clarify the effects of DUR from the financial perspective, this study considers the typical financial issue that a capital-constrained supply chain may have access to more than one financing sources through, for example, bank credit financing (BCF) and trade credit financing (TCF). These are the two most common financing approaches provided by banks and upstream suppliers within the supply chain, respectively (Kouvelis and Zhao, 2017). Several studies have demonstrated that DUR will affect the effectiveness of BCF and TCF. Through numerical experiments, Jing et al. (2012) show that reduced demand uncertainty is more effective in improving supply chain efficiency under TCF than BCF. Also, Chen (2015) finds that the ratio of the capital-constrained retailer’s (or the supplier’s) profit under BCF to the profit under TCF varies with the degree of demand uncertainty. The latter study suggests that DUR can influence not only the retailer’s cost-benefit trade-off between BCF and TCF, but also the supplier’s willingness to offer trade credit. Therefore, under the influence of DUR, the equilibrium conditions of the decisional dynamics within a supply chain in the coexistence of BCF and TCF will be affected. In spite of the influence of DUR on the operations-finance interface, most of the existing studies concern only single financing approaches, and none of them has compared the would-be effects of DUR under multiple financing approaches (e.g., BCF and TCF). To address this research gap, this study introduces DUR into a capital-constrained supply chain model and analytically compares the effects of DUR on the decisions and performance of all supply chain members under BCF and TCF.

Based on a basic supply chain consisting of two self-interested members (a supplier and a retailer), this study investigates how they will make operational and financial decisions in response to different degrees of DUR. The analysis is performed within a non-cooperative strategic game setting and involves two alternative financing approaches. Specifically, the retailer (he) is capital-constrained and has to seek financing, which can be either BCF or TCF. On the other hand, the supplier (she) is well funded. What is more, she is in a position to provide trade credit to the retailer but may decide whether she should make such an offer. However, bank credit is always available - a convention adopted in the supply chain finance literature (Kouvelis and Zhao, 2012, Kouvelis and Zhao, 2017, Jing et al., 2012, Cai et al., 2014). Therefore, this non-cooperative game involves only the strategic interactions between the retailer and the supplier. If the supplier is willing to offer trade credit and TCF can bring higher profit to the retailer than BCF, TCF will be adopted. In contrast, the adoption of BCF should satisfy one of following conditions: (i) the supplier does not offer trade credit and thus the retailer can only rely on BCF; (ii) the supplier is willing to offer trade credit but the retailer prefers BCF.

Moreover, this analysis further considers two scenarios, which are due to the consideration of exogenous and endogenous wholesale pricing. In the former case, the wholesale price is given or fixed due to the high cost for changing the terms of wholesale price contract (Khanjari et al., 2013; Jin and Luo, 2017). In other words, the supplier is a price taker and cannot determine the wholesale price to her own advantage. In the latter case, however, the supplier plays the role of a price maker and can optimise the wholesale price so that she may extract maximum benefit from other supply chain members (Jing et al., 2012, Kouvelis and Zhao, 2012). In both cases, the strategic positions taken by the retailer and the supplier will be significantly different. In the endogenous wholesale pricing setting, the supplier always acts as the leader of a Stackelberg game as she can influence the retailer’s choice between BCF and TCF through wholesale price setting. As shown by Jing et al. (2012), if the retailer chooses BCF and that choice will undermine the supplier’s benefit, the latter can simply set the wholesale price sufficiently high in order to force the retailer to adopt TCF instead. However, in the case of exogenous wholesale pricing, the supplier can only decide whether she should offer trade credit or not. Therefore, it is important to consider these two cases of wholesale pricing in order to better understand the related channel dynamics.

Therefore, this study will explore the scenarios formed by combinations of the two cases of wholesale pricing and the two alternative financing approaches, with the consideration of DUR that indicates how demand uncertainty can be successfully reduced. The objective is, for each scenario, to obtain the equilibrium solution that describes the optimal decisions and profits of the supplier and retailer under a specified set of conditions, including the degree of DUR, the product cost, and the credit risk premium. The retailer’s decisions involve the optimal order quantity and the financing approach; and the supplier will decide on the optimal wholesale pricing strategy (in the case of endogenous pricing) and on whether she should offer the retailer trade credit. Also, the bank will decide on the fairly priced interest rate corresponding to the retailer’s loan amount. By deriving the equilibrium solutions for these scenarios, the effects of DUR on the selection of financing approach and the operational decisions and performance of all supply chain members under BCF and TCF can be investigated.

To model the decision processes of the supplier and the retailer and the interactions between them, we consider the case where a newsvendor-like retailer faces uncertain demand and orders products from the supplier. Both parties are assumed to be risk-neutral profit maximisers (Kouvelis and Zhao, 2012, Jing et al., 2012, Cai et al., 2014). The degree of DUR is parameterised by using the method of mean-preserving spread (Gerchak and Mossman, 1992, Rothschild and Stiglitz, 1970) and is known by all participants with the bank included. On the basis of this model setting, we first consider the basic situation where the bank is assumed to be risk neutral. The effects of DUR on the optimal decisions and profits of both channel members under BCF and TCF as well as the selection of financing approach are investigated. Moreover, we extend the model to consider the role played by the bank, which is averse to loss due to the possibility of the retailer’s default. An investigation is performed on how the bank’s different risk attitudes will affect the decisional dynamics, especially in the selection of financing approach by the retailer.

This research contributes to the existing literature and practices in the following ways. First, our analysis complements the existing studies (e.g., Lau and Lau, 2002, Miyaoka and Hausman, 2008, Li and Petruzzi, 2017) that focus on the effects of DUR by exploring the interactions between a supplier and a financially constrained retailer. It reveals that both channel members’ operational characteristics and profitability are subject to the degree of DUR. In particular, we find that a greater DUR will motivate the supplier to charge a higher wholesale price and make the retailer order less under TCF but more under BCF. Also, the achievement of a higher DUR does not necessarily lead to a greater benefit for either the supplier or the retailer. This finding is counterintuitive but concurs with the well-documented phenomenon of “accuracy trap” (see Laucka, 2005). Our analysis does not only exemplify the existence of “accuracy trap”, but also points out that, due to the strategic interactions of members in the supply chain with financing needs, endlessly striving for higher demand forecast accuracy is not necessarily beneficial. Specifically, due to the increase in DUR, the supplier can mostly benefit in the case of endogenous wholesale pricing but will suffer a loss in the case of exogenous wholesale pricing. In contrast, the retailer can always benefit from a higher DUR only when the wholesale price is exogenous but will suffer a loss when the wholesale price is endogenous.

Second, our study explores the relationship between the degree of DUR and the equilibrium solutions to the selection of financing approach in two wholesale pricing cases. Previous research (e.g., Lau and Lau, 2002, Li and Petruzzi, 2017) has emphasised the importance of the supplier’s power to set the wholesale price and proved that such power will change her strategic positions within the supply chain. Our study extends these works by investigating the effects of DUR on the selection of financing approach when the supplier assumes different strategic positions. We find that, in general, TCF is more suitable in the case of endogenous than exogenous wholesale pricing, especially when the degree of DUR is not high. Specifically, in the case of exogenous wholesale pricing, TCF should be adopted under high DUR, moderate trade credit risk premium and low exogenous wholesale price. In the case of endogenous wholesale pricing, the adoption of TCF only requires that both DUR and the production cost are not high. Otherwise, BCF should be adopted. This result is generally consistent with the findings of empirical studies such as Elliehausen and Wolken, 1993, Brennan et al., 1988. Based on the balance sheet data of the U.S. small nonfinancial firms, they find that firms with a higher default probability (indicated by lower demand uncertainty) or a higher profit margin (indicated by lower production cost) tend to depend more on TCF in practice.

Thirdly, our work has been extended to examine the optimal financial strategy for both channel parties when the external lending institutions (such as banks) are risk averse. We find that, when the bank is loss averse, the retailer’s selection between BCF and TCF generally does not depart from the case when the bank is risk neutral. Only in the case of exogenous wholesale pricing with high demand uncertainty, the retailer may turn to TCF due to the bank’s loss aversion. Also, we find that the loss-averse bank will charge a higher interest rate as the degree of the bank’s loss aversion increases.

The remainder of this paper is organised as follows. Section 2 reviews the related literature on the effects of DUR on operations management and financial issues. The formulation of the supply chain model that incorporates DUR is given in Section 3. 4 Exogenous wholesale price, 5 Endogenous wholesale price study the conditions for equilibrium in the case of exogenous and endogenous wholesale pricing, respectively. In Section 6, numerical studies are conducted to present and compare the effects of DUR on the equilibrium solutions in both wholesale pricing cases. Section 7 describes the effects of the bank’s loss-averse attitude on the selection of financing approach and also its decision on the interest rate. Insights derived from this study and suggestions for future extensions are summarised in Section 8. All major proofs are given in the Online Appendix.