I begin this research by observing a crucial shift in the global automotive industry. A new round of technological revolution is pushing automobile manufacturers away from conventional internal combustion engines and toward electrified mobility. In this context, the new energy vehicle has become one of the most important solutions to reduce reliance on fossil fuels and to mitigate carbon emissions. The traction battery pack, as the core energy storage unit, largely determines the driving range, safety, and long-term durability of a new energy vehicle. Consequently, investment decisions in traction battery pack technology have direct implications for market demand, consumer confidence, and the sustainable growth of the entire electric vehicle ecosystem.
My study focuses on two interrelated strategic problems in the new energy vehicle supply chain. The first problem is the R&D investment decision for traction battery packs. I examine three possible arrangements: the vehicle manufacturer develops the battery alone; the battery supplier develops the battery alone; or both parties collaborate in the development process. The second problem is the remanufacturing strategy for retired traction battery packs. I investigate how the choice of recycling entity affects the supplier’s incentive to invest in remanufacturing process innovation, and how that innovation influences production and pricing decisions.
In the remainder of this article, I first introduce the analytical framework of consumer anxiety and its relation to traction battery pack innovation. Then I present three game-theoretic models of R&D investment. After that, I develop two closed-loop supply chain models for remanufacturing, where either the supplier or the manufacturer is responsible for collecting retired battery packs. Throughout the article, I use mathematical derivation, comparative analysis, and numerical examples to provide clear managerial insights.
1. Research Background and Motivation
The rapid expansion of the new energy vehicle market has brought the traction battery pack to the center of public attention. From the perspective of consumers, range anxiety remains one of the most frequently cited barriers to adoption. Even when a vehicle is fully charged, drivers worry about the availability of charging stations, the degradation of battery capacity in cold weather, and the time required for recharging. In addition, resale anxiety discourages potential buyers because the residual value of an electric vehicle depends strongly on the remaining health of its traction battery pack. If the battery degrades quickly, the resale price of the whole vehicle falls sharply. Therefore, improving battery performance is an urgent task that requires sustained investment in research and development.
From the perspective of the environment, a large wave of retired traction battery packs is approaching. Early electric vehicles are now reaching the end of their useful battery life. If these batteries are disposed of improperly, hazardous metals and electrolytes may pollute soil and water. In contrast, an efficient recycling and remanufacturing system can recover valuable materials and reduce the lifecycle cost of the traction battery pack. Many suppliers and manufacturers have therefore begun to design closed-loop supply chains for end-of-life batteries. In such a system, the remanufacturing process converts used batteries into serviceable products, and process innovation plays a key role in lowering the cost of that conversion.
Motivated by these observations, I address the following questions in this research. How should firms choose between independent and collaborative R&D modes for traction battery packs? How does consumer anxiety affect R&D investment, price, and profit? In the remanufacturing context, should the battery supplier or the vehicle manufacturer collect retired batteries? How does the choice of the collection entity influence the supplier’s incentive to invest in remanufacturing innovation? I summarize the main variables and notation in Table 1.
| Symbol | Definition |
|---|---|
| \( \alpha \) | Consumer anxiety intensity |
| \( \theta \) | Traction battery pack R&D level |
| \( I \) | R&D investment cost coefficient |
| \( w \) | Wholesale price of the battery |
| \( p \) | Market price of the new energy vehicle |
| \( q \) | Production quantity and demand |
| \( S \) | Fixed upfront investment needed for self-development |
| \( \lambda \) | Manufacturer’s cost-sharing ratio in supplier-led R&D |
Throughout the R&D analysis, I assume a linear demand function. The demand for the new energy vehicle increases with the R&D level \( \theta \) and decreases with consumer anxiety \( \alpha \) and retail price \( p \). I normalize market size and price sensitivity to one, which gives
$$ q = 1 – p + (1-\alpha)\theta. $$
In particular, when \( \alpha=1 \), consumer anxiety completely offsets the demand-enhancing effect of innovation, and demand is only \( 1-p \). When \( \alpha=0 \), there is no anxiety, and R&D contributes fully to increased demand. This formulation allows me to capture the intuition that R&D on the traction battery pack can reduce anxiety by improving durability and vehicle range.
For the cost structure, I assume that a firm investing in R&D level \( \theta \) incurs a convex investment cost \( I\theta^2 \). The quadratic form reflects diminishing returns to R&D effort. If both firms participate in R&D, each firm pays for its own research activities. I also normalize the production cost to zero because the production cost does not change the qualitative comparison among the three R&D modes.
2. Three Modes of Traction Battery Pack R&D
2.1 Manufacturer-Led R&D
In the manufacturer-led R&D mode, which I denote as Mode D, the vehicle manufacturer develops the traction battery pack independently. This mode is observed when an automaker builds an in-house battery division and makes a large upfront investment \( S \) to create factories, laboratories, and technical teams. The manufacturer chooses the R&D level \( \theta_d \) and the vehicle price \( p_d \) to maximize its profit:
$$ \max_{p_d,\theta_d} \; \pi_{md}=p_d\left(1-p_d+(1-\alpha)\theta_d\right)-I\theta_d^2-S. $$
Using backward induction, the manufacturer chooses the optimal R&D level and price. The first-order conditions lead to the following equilibrium results under the condition \( I>(1-\alpha)^2/4 \):
$$ \theta_d^*=\frac{1-\alpha}{4I-(1-\alpha)^2}, \quad p_d^*=q_d^*=\frac{2I}{4I-(1-\alpha)^2}. $$
Because the manufacturer directly faces consumer demand, it can respond quickly to shifts in consumer anxiety. I denote the equilibrium profit as
$$ \pi_{md}^*=\frac{4I^2}{\left[4I-(1-\alpha)^2\right]^2}-I\left(\frac{1-\alpha}{4I-(1-\alpha)^2}\right)^2-S. $$
This profit is positive only when the upfront investment \( S \) is below a threshold \( S_0 \). If the initial cost of building an in-house battery supply chain is too high, the manufacturer may prefer not to pursue independent R&D.
2.2 Supplier-Led R&D
In the supplier-led R&D mode, which I denote as Mode U, the traction battery supplier is responsible for most of the research effort. The manufacturer, however, can share a proportion \( \lambda \) of the supplier’s R&D cost. This cost-sharing contract is signed before the supplier chooses its innovation level; otherwise, cost sharing cannot provide an incentive. The supplier chooses \( \theta_u \), and the manufacturer chooses the price \( p_u \) and the sharing ratio \( \lambda \). The profit functions are
$$ \pi_{su}=\left(w\right)\left(1-p_u+(1-\alpha)\theta_u\right)-(1-\lambda)I\theta_u^2, $$
$$ \pi_{mu}=\left(p_u-w\right)\left(1-p_u+(1-\alpha)\theta_u\right)-\lambda I\theta_u^2. $$
Solving this Stackelberg game, I obtain the equilibrium in Mode U. In this mode, the supplier has no direct contact with consumers and therefore cannot easily observe changes in consumer anxiety. Consequently, its optimal R&D level is lower than in the other two modes. I compare this result with Mode D and Mode C in Section 2.4.
2.3 Collaborative R&D
In the collaborative R&D mode, which I denote as Mode C, both the supplier and the manufacturer contribute to different parts of traction battery pack development. The supplier works on electrochemistry, cell design, and safety features, while the manufacturer contributes know-how about vehicle integration, thermal management, and consumer preferences. Let the supplier’s R&D effort be \( \theta_s \) and the manufacturer’s effort be \( \theta_m \). The final R&D level is
$$ \theta_c=\theta_s+\theta_m. $$
The two parties simultaneously choose their respective research efforts before the manufacturer sets the retail price. The profit functions are
$$ \pi_{sc}=w\left(1-p_c+(1-\alpha)\theta_c\right)-I\theta_s^2, $$
$$ \pi_{mc}=\left(p_c-w\right)\left(1-p_c+(1-\alpha)\theta_c\right)-I\theta_m^2. $$
Collaboration gives the supplier access to downstream market information and gives the manufacturer access to core battery technology. I derive the equilibrium R&D level, price, and demand by solving the first-order conditions of both firms.
3. Comparative Analysis of the Three R&D Modes
After solving the three models, I compare the equilibrium outcomes in terms of R&D investment, production, pricing, and profit. The main structural comparison is presented in Table 2.
| Decision Variable | Mode D | Mode U | Mode C |
|---|---|---|---|
| Traction battery pack R&D level \( \theta \) | High | Lowest | Equal to Mode D |
| Vehicle price \( p \) | Lowest | Medium | Highest |
| Vehicle demand \( q \) | Highest | Lowest | Medium |
| Supplier profit \( \pi_s \) | Not applicable | Lower | Higher |
3.1 R&D Investment Decisions
My first finding concerns the level of R&D investment. I write the result as a proposition.
Proposition 1. In the three R&D modes studied, the supplier-led R&D mode yields the lowest traction battery pack R&D level. The manufacturer-led mode and the collaborative mode yield the same R&D level:
$$ \theta_u < \theta_d = \theta_c. $$
The reason is that the manufacturer sits closer to the final market and can perceive consumer anxiety more accurately. When the manufacturer participates in R&D, either alone or with the supplier, the R&D direction is more aligned with actual market needs. As a result, R&D is more effective at generating demand. My result also indicates that the supplier-only mode is inferior from a technological improvement perspective.
I further examine the effect of consumer anxiety \( \alpha \) on R&D investment.
Proposition 2. In every R&D mode, the optimal R&D level is decreasing in consumer anxiety:
$$ \frac{\partial \theta_d}{\partial \alpha}<0, \qquad \frac{\partial \theta_u}{\partial \alpha}<0, \qquad \frac{\partial \theta_c}{\partial \alpha}<0. $$
This result has an intuitive explanation. If consumers worry about limited range and uncertain resale value, the additional demand generated by a marginal improvement in the traction battery pack is relatively small. Firms therefore earn a lower return on R&D and optimally reduce investment. Strong consumer anxiety not only suppresses current sales but also damages long-term innovation incentives.
Moreover, I observe that the gap between different modes shrinks as \( \alpha \) increases. When anxiety is high, even the most market-oriented firm loses its motivation to invest, so the difference between manufacturer-involved R&D and supplier-only R&D becomes smaller.
3.2 Production and Pricing Decisions
Next I compare the prices and quantities in equilibrium.
Proposition 3. The equilibrium vehicle prices and demands satisfy
$$ p_d < p_u < p_c, \qquad q_u < q_c < q_d. $$
Mode C has the highest retail price because both parties contribute specialized R&D resources, and the decentralized division of research tasks may create additional coordination costs. These costs are partly passed on to final consumers through a higher price. Mode D has the lowest retail price because the manufacturer coordinates the whole battery production and vehicle assembly process internally, avoiding double marginalization in the battery supply chain.
For demand, the manufacturer-led mode generates the highest quantity. The direct link between R&D and final demand means that the manufacturer can adjust production plans quickly when consumer anxiety changes. Mode U leads to the lowest quantity because the supplier’s technological effort is less responsive to market conditions and because the wholesale price remains an external contract term.
3.3 Profit and Strategy Choice
I now compare supplier profits and manufacturer profits under different modes.
Proposition 4. For the traction battery supplier, collaborative R&D always yields a higher profit than supplier-led R&D:
$$ \pi_{su} < \pi_{sc}. $$
Collaboration allows the supplier to share R&D costs and benefit from the manufacturer’s downstream information. Consequently, the supplier has a strict preference for Mode C over Mode U.
The manufacturer’s profit comparison is more nuanced. The result depends on the upfront investment \( S \), the wholesale battery price \( w \), and the R&D cost coefficient \( I \).
Proposition 5. (i) When the upfront investment \( S \) is small, Mode D gives the manufacturer the highest profit. When \( S \) exceeds a threshold, Mode U becomes better for the manufacturer than Mode D. (ii) When the R&D cost coefficient \( I \) is small, Mode C is better for the manufacturer than Mode U. When \( I \) is large and the wholesale price \( w \) is also high, the manufacturer prefers to leave R&D entirely to the supplier.
These results show that neither pure independence nor pure outsourcing dominates in all circumstances. I summarize the profit comparison in Table 3.
| Condition | Preferred Mode | Main Intuition |
|---|---|---|
| Small upfront cost \( S \), moderate \( I \) | Mode D | Manufacturer internalizes demand and avoids double marginalization |
| Large \( S \), small \( I \) | Mode C | R&D cost is shared; high innovation increases demand |
| Large \( S \), large \( I \), high \( w \) | Mode U | R&D is expensive; manufacturer avoids R&D expenditure |
From a supply chain perspective, Mode D and Mode C are more effective at improving traction battery pack quality than Mode U. Because both modes involve the manufacturer, the research agenda is aligned with consumer expectations, leading to better range, durability, and resale value. For policymakers and managers, my results suggest that the vehicle manufacturer should actively participate in battery R&D rather than passively relying on upstream suppliers alone.
3.4 Numerical Illustration of the R&D Model
I now illustrate these analytical results with numerical examples. I set \( \alpha=0.4 \), \( w=0.4 \), and \( I=0.6 \). The simulated patterns confirm my theoretical conclusions.
| Parameter | Mode D | Mode U | Mode C |
|---|---|---|---|
| R&D level when \( \alpha \) increases | Decreases | Decreases | Decreases |
| Vehicle price when \( I \) increases | Decreases | Decreases | Decreases |
| Demand when \( I \) increases | Decreases | Decreases | Decreases |
| R&D level ranking | High | Lowest | High |
| Demand ranking | Highest | Lowest | Medium |
The numerical results show that when consumer anxiety is high, firms are less willing to invest in traction battery pack innovation. Even a moderate reduction in anxiety can substantially increase optimal R&D effort. Therefore, an important policy recommendation is to reduce consumer anxiety through measures such as battery health warranties, reliable charging networks, and transparent secondhand battery certification.
4. Remanufacturing Strategies for the Traction Battery Pack
In the second half of this research, I shift attention from the forward supply chain to the reverse supply chain. After several years of use, traction battery packs reach a point where their capacity fades and they are no longer suitable for vehicle traction. At that point, they must either be recycled for materials or remanufactured for second-life applications. Because of the complexity of battery chemistry, disassembly, and testing, remanufacturing can be expensive. Investment in process innovation is therefore important to reduce the variable cost of remanufacturing.
I consider a closed-loop supply chain with one battery supplier, one vehicle manufacturer, and consumers. The supplier performs the remanufacturing operation and sells the remanufactured battery to the manufacturer at a wholesale price \( w \). The manufacturer assembles the battery into a vehicle and sells it to consumers. I assume that all collected traction battery packs can be remanufactured and used in new vehicles. The production quantity is equal to the collection quantity.
The demand is characterized by the inverse demand function
$$ p = 1 – q. $$
The variable cost of remanufacturing is
$$ C(x)=c_r(1-\eta x), $$
where \( c_r \) is the unit remanufacturing cost without any process innovation, \( x \) is the remanufacturing R&D level, and \( \eta \in[0,1] \) is the conversion efficiency of R&D into cost reduction. A higher \( \eta \) means that a given R&D effort reduces remanufacturing costs more strongly. The R&D investment cost is \( Kx^2 \), where \( K \) captures the degree of difficulty of such innovation.
I define two recycling models:
- Model S: the battery supplier collects retired traction battery packs directly from consumers.
- Model M: the vehicle manufacturer collects retired packs through its dealer network and sells them to the supplier at a transfer price \( g \).
The two structures are compared in Table 5. The collection cost is assumed to be \( hq^2 \), where \( h \) is the collection difficulty coefficient.
| Feature | Model S (Supplier Collects) | Model M (Manufacturer Collects) |
|---|---|---|
| Recycling entity | Battery supplier | Vehicle manufacturer |
| Collection cost | Paid by supplier | Paid by manufacturer |
| Used battery transfer | Not needed | Manufacturer sells to supplier at price \( g \) |
| R&D responsibility | Supplier | Supplier |
| Main advantage | Direct material flow back to supplier | Convenient access to end users through dealers |
4.1 Model S: Supplier Recycling
In Model S, the supplier’s profit consists of the revenue from selling the remanufactured traction battery pack minus the collection cost and the R&D cost:
$$ \pi_{ss}=(w-c_r(1-\eta x_s))q_s – h q_s^2 – Kx_s^2. $$
The manufacturer’s profit is
$$ \pi_{ms}=(p_s-w)q_s = (1-q_s-w)q_s. $$
The decision sequence is as follows. First, the supplier chooses the R&D level \( x_s \) and the wholesale price \( w_s \). Second, the manufacturer chooses the production quantity \( q_s \). Solving this Stackelberg game gives the equilibrium summarized in Table 6.
| Case | R&D Level \( x_s \) | Wholesale Price \( w_s \) |
|---|---|---|
| \( K_1 < K \le K_2 \) | \( x_s=1/\eta \) | Boundary solution with maximum R&D |
| \( K>K_2 \) | Interior solution satisfying \( \frac{\partial \pi_{ss}}{\partial x}=0 \) | Interior solution satisfying \( \frac{\partial \pi_{ss}}{\partial w}=0 \) |
I state the key comparative statics in Proposition 6.
Proposition 6. In Model S, when R&D is relatively easy, a higher conversion efficiency \( \eta \) lowers the optimal R&D level. When R&D is relatively difficult, a higher \( \eta \) raises the optimal R&D level. Formally, for \( K>K_2 \),
$$ \frac{\partial x_s}{\partial \eta}>0, \qquad \frac{\partial x_s}{\partial K}<0, \qquad \frac{\partial x_s}{\partial c_r}>0, \qquad \frac{\partial x_s}{\partial h}<0. $$
The intuition is clear. If the R&D challenge is easy, the supplier can already achieve the maximum feasible R&D level at low cost, so a further increase in conversion efficiency only reduces the required effort. If the R&D challenge is difficult, a higher conversion efficiency means that each unit of R&D investment yields a larger cost reduction. The supplier therefore increases its research effort. I also find that an increase in the initial remanufacturing cost \( c_r \) motivates the supplier to invest more in innovation, because innovation preserves the profit margin.
4.2 Model M: Manufacturer Recycling
In Model M, the manufacturer collects retired traction battery packs and transfers them to the supplier at a price \( g \). The supplier pays the transfer price and also invests in remanufacturing R&D. The profit functions are
$$ \pi_{sm}=(w-c_r(1-\eta x_m))q_m – gq_m – Kx_m^2, $$
$$ \pi_{mm}(p_m-w-c_0)q_m + g q_m – h q_m^2. $$
I summarize the equilibrium conditions in Table 7.
| Case | R&D Level \( x_m \) | Wholesale Price \( w_m \) |
|---|---|---|
| \( K_3 < K \le K_4 \) | \( x_m=1/\eta \) | Boundary solution with maximum R&D |
| \( K>K_4 \) | Interior solution satisfying \( \frac{\partial \pi_{sm}}{\partial x}=0 \) | Interior solution satisfying \( \frac{\partial \pi_{sm}}{\partial w}=0 \) |
I also derive a useful monotonic relation between the transfer price \( g \) and the wholesale price \( w_m \).
Proposition 7. In Model M, the equilibrium wholesale price \( w_m \) is increasing in the transfer price \( g \):
$$ \frac{\partial w_m}{\partial g}>0. $$
This finding has an important implication. When the manufacturer charges the supplier a higher price for collected batteries, the supplier treats that price as an additional cost of production and raises the wholesale price. This leads to a higher retail price and lower consumer demand in the forward market.
4.3 Comparison of Model S and Model M
I next compare the two recycling models in terms of R&D level, price, quantity, and profit.
Proposition 8. The supplier’s remanufacturing R&D level in Model S is always weakly higher than in Model M:
$$ x_s \ge x_m. $$
Moreover, the retail price is lower in Model S, whereas the demand is higher in Model S:
$$ p_s < p_m, \qquad q_s > q_m. $$
The intuition is that in Model S the supplier avoids paying the transfer price \( g \). The supplier’s marginal cost is lower, so it retains more profit margin and has a stronger incentive to invest in remanufacturing innovation. A higher R&D level reduces the variable remanufacturing cost, which allows the supplier to set a lower wholesale price. The lower wholesale price is partly passed on to consumers as a lower retail price, which stimulates demand. From the perspective of the whole supply chain, supplier-led recycling is more effective in promoting the circulation of remanufactured traction battery packs.
I compare supplier profits in the two models.
Proposition 9. The supplier always earns more profit in Model S than in Model M:
$$ \pi_{ss} > \pi_{sm}. $$
Because Model S gives the supplier lower acquisition cost and higher demand, the supplier clearly prefers to collect retired battery packs by itself. This is an important message for battery manufacturers. If the supplier can build its own recovery network, perhaps with the help of local recyclers or dealership partners, it can achieve both higher profit and better environmental performance.
The manufacturer’s profit comparison is more complex. I find that the manufacturer often prefers to be the recycling entity, but not always. The complete result is described in Proposition 10.
Proposition 10. The manufacturer earns higher profit in Model M than in Model S in most parameter regions. However, when the collection difficulty \( h \) is low and the R&D difficulty \( K \) lies in an intermediate range, the manufacturer earns more profit in Model S than in Model M. In that case, both supply chain members prefer Model S, and the chain is coordinated in terms of recycling responsibility.
I summarize the profit comparison in Table 8, where \( h_1 \), \( h_2 \), \( K_7 \), and \( K_8 \) are positive thresholds derived from the equilibrium expressions.
| Collection Difficulty \( h \) | R&D Difficulty \( K \) | Profit Comparison |
|---|---|---|
| Low: \( 0<h<h_2 \) | Intermediate: \( K_7<k<k_8 \) | \( \pi_{ms}>\pi_{mm} \) |
| Low: \( 0<h<h_2 \) | Other \( K \) | \( \pi_{ms}<\pi_{mm} \) |
| High: \( h>h_2 \) | Any \( K \) | \( \pi_{ms}<\pi_{mm} \) |
This result has a practical implication. When collection is not too difficult and R&D for remanufacturing is neither extremely easy nor extremely difficult, the supplier and the manufacturer have aligned incentives. The supplier collects the used traction battery packs, invests in innovation, and stimulates demand; the manufacturer benefits from higher sales volume and lower battery cost. In that region, no conflict arises over the choice of recycling entity.
5. Numerical Observations for the Remanufacturing Model
To validate the analytical results, I use the following numerical values: \( \eta=0.2 \), \( c_r=0.4 \), and \( h=0.9 \). I vary the R&D cost coefficient \( K \) over a sufficiently wide range. Table 9 reports the qualitative patterns.
| Variable | Model S | Model M |
|---|---|---|
| R&D level \( x \) as \( K \) increases | Decreases | Decreases |
| R&D level ranking | Higher | Lower |
| Vehicle price ranking | Lower | Higher |
| Quantity ranking | Higher | Lower |
| Supplier profit ranking | Higher | Lower |
The numerical examples confirm that Model S dominates Model M in terms of both remanufacturing innovation and total market expansion. These findings can be explained by the fact that supplier-led recycling shortens the reverse logistics chain and avoids the additional transfer price. In contrast, manufacturer-led recycling introduces an extra layer of transaction cost that reduces the supplier’s profit margin and weakens the incentive to invest in innovation.
The simulations also illustrate the interaction between R&D efficiency and collection difficulty. When collection difficulty is moderate and R&D is difficult, the manufacturer may still prefer to control the recycling process because doing so gives it a secondary revenue stream from selling used batteries to the supplier. However, when the transfer price is too high, the supplier responds by increasing the wholesale price, which hurts final demand. Therefore, the manufacturer should not set an excessively high transfer price if its objective is to maximize the total profit of the supply chain.
6. Managerial Insights and Policy Implications
My research produces several important implications for the management of traction battery pack development and end-of-life processing.
First, vehicle manufacturers should not completely delegate traction battery pack R&D to upstream suppliers. In the supplier-led R&D mode, the innovation level is the lowest, and vehicle demand is also the lowest. A prominent reason is that consumers’ range anxiety and resale anxiety are best understood by downstream firms that interact directly with buyers. If the vehicle manufacturer participates in battery research, whether independently or cooperatively, the R&D effort becomes more effective and more responsive to market changes.
Second, collaborative R&D between the supplier and the manufacturer benefits the supplier in most cases. It can reduce R&D costs, share technical risk, and improve the total added value of the traction battery pack. For the manufacturer, collaboration is desirable when the upfront cost of establishing in-house development is high and when the cost of R&D is not excessively high. If the wholesale price of batteries is already high, the manufacturer may be reluctant to co-fund upstream R&D because the additional sales generated by innovation may not cover the extra cost.
Third, in the recycling phase, I recommend that end-of-life traction battery packs should preferably be collected by the supplier that performs the remanufacturing operation. Supplier-led recycling creates a shorter and more efficient closed loop. This design avoids the markup of the manufacturer-operated collection channel and preserves funds for process innovation. The numerical results support this recommendation: supplier-led recycling yields a higher remanufacturing R&D level, lower retail price, higher demand, and higher supplier profit.
Fourth, if the manufacturer takes on the collection responsibility, it should not overprice the used traction battery pack when selling it to the supplier. Although the manufacturer earns transfer revenue from the sale, the corresponding increase in supplier cost is passed down through the wholesale price. Higher wholesale prices reduce vehicle demand and may eventually lower the manufacturer’s profit as well. A moderate transfer price can balance the incentives of both parties.
Fifth, the ability to convert R&D effort into actual cost savings is a key driver of remanufacturing innovation. When the remanufacturing process is inherently difficult, high conversion efficiency of R&D is crucial. Suppliers should therefore invest in digital traceability, automated disassembly, and diagnostic tools to increase the efficiency \( \eta \). These capabilities make R&D spending more productive and incentivize the supplier to invest in remanufacturing innovation.
Finally, I link the two parts of this research through the concept of the traction battery pack life cycle. In the design stage, R&D can improve battery durability and reduce future recycling costs. In the end-of-life stage, remanufacturing innovation can lower the cost of reusing the battery pack. A full life-cycle strategy requires the manufacturer and supplier to coordinate both forward and reverse decisions. The traction battery pack is not merely a component of the vehicle; it is also a resource that can be refurbished, resold, and re-used. A successful strategy should therefore integrate product innovation, supply chain collaboration, and closed-loop recycling.
7. Conclusions
In conclusion, this research provides a systematic game-theoretic analysis of traction battery pack R&D and remanufacturing strategies in the new energy vehicle industry. I obtain several important findings.
For the R&D investment strategy, I compare manufacturer-led R&D, supplier-led R&D, and collaborative R&D. The results show that consumer anxiety and R&D cost both suppress the optimal R&D level. The supplier-led R&D mode performs worst in terms of both innovation and market demand. The manufacturer-led mode achieves the highest demand and the lowest retail price. Collaborative R&D benefits the supplier but does not always maximize the manufacturer’s profit.
For the remanufacturing strategy, I find that the identity of the recycling entity has a substantial impact on process innovation. When the supplier is responsible for collection, the R&D level is always weakly higher than when the manufacturer collects the used batteries. Supplier-led recycling also leads to lower retail prices and higher quantities. The supplier always prefers this mode, while the manufacturer usually prefers to handle recycling itself, except in cases where collection difficulty is low and R&D difficulty is intermediate.
These findings have direct relevance for managers and regulators. Firms should account for consumer psychology when deciding how much to invest in the improvement of the traction battery pack. They should also avoid supply chain structures that add unnecessary transaction costs in the reverse channel. Government policies that encourage collaboration between battery suppliers and vehicle manufacturers can accelerate innovation in battery technology. At the same time, policies that support supplier-based recycling networks can improve the economic viability of battery remanufacturing and reduce the environmental burden of retired batteries.
Finally, I note some limitations of this study. I consider a linear demand function and a single-period setting. In future research, a two-period model can be developed to capture the competition between new and remanufactured traction battery packs. Future work can also incorporate the manufacturer’s sales effort, asymmetric information, or the strategic choice between vertical integration and alliance formation. These extensions would offer deeper insight into the evolving market for traction battery packs.