As the current global environmental situation is gradually becoming grim and the energy problem is gradually becoming prominent, all countries in the world are paying attention to the research of electric vehicles. The large-scale promotion of the use of electric vehicles has the functions of ensuring the energy security, improving the local environment, and promoting economic development (Ahman, 2006). Marmiroli et al. (2020) compared the environmental assessment of electric and conventional light-duty vehicles. The results showed that electric vehicles have advantages in urban environments due to the large amount of stopping and regenerative breaking that occurs during urban transportation. The trend of replacing conventional fuel power with new energy power in the future automobile industry is gradually clear (Specchia et al., 2005). However, the market share of electric vehicles for urban logistics is much lower than the potential in some countries (İmre et al., 2021). To estimate the potential and possible directions of electric vehicle adoption, several scholars used the Bass diffusion model (Park et al., 2011; Lee et al., 2019; Brdulak et al., 2021). The Bass model captures the growth trend of a new product’s sales forecast. Yu et al. (2018) also used the diffusion model to assist the evaluation of policies’ effectiveness in the adoption of electric vehicles in China. Li et al. (2020) examined electric vehicles in urban logistics distribution systems under a sharing economy. They found that in the first simulation, sharing distribution networks can reduce trip times and costs. The second simulation estimated different timepoints of electricity price, which indicates that low electricity prices reduce traffic congestion and cost. Furthermore, they suggested that an increase in carbon tax reduces the CO₂ emissions. Gao et al. (2022) evaluated inceptives for the adaptation electric vehicles in urban logistic, and they analyzed different factors including price, driving rate, and decision related to time window. This suggests that price and driving range of the electric vehicle motivate the consumer to adopt EV. Furthermore, wider time window also motives the consumer to adopt EV. Juvvala and Sarmah (2021) analyzed the policy electric vehicles in city logistics, and they used a real-life case study and use the MACO mathematical model test. The hypothesis found that the purchase subsidy and the zone entry fee play an important role in city logistics. Roumboutsos et al. (2014) found that electric vehicles help urban logistics significantly and improve urban terrific congestion, and they also suggested that electric vehicles reduce the pollution emissions. Based on a real-world application, Duarte et al. (2016) assessed the adequacy of electric vehicles for urban logistics in Lisbon. They compared the energy impacts of switching to electric mobility and found that the battery electric vehicles lead to low vehicle usage energy consumption. Quak et al. (2016) addressed recent advancements, challenges, and prospects for the large-scale use of electric freight vehicles in daily city logistics. Iwan et al. (2019) illustrated that electric vehicles are suitable for urban logistics by investigating the potentials of electric mobility in commercial transport in terms of travel patterns and daily mileage. Ehrler et al. (2021) analyzed a real-life case of electric vehicles for grocery logistics in Germany and argued the prerequisites and perspectives for a successful introduction of electric freight vehicles for urban logistics. Some scholars even look into the sharing economy to improve the routing model for urban logistics distribution using electric vehicles (Verena et al., 2019; Zuo et al., 2019).

Firms should take the liability to reduce environmental externalities via green technology investments. While complying with government regulations, Yalabik and Fairchild (2011) argued that government subsidies are more effective than penalties in encouraging firms’ technology adoptions to become environment-friendly. Their findings suggest that the incentive policy is a stronger driving force for firms’ environmental innovation than the penalty policy. By using a lifecycle approach, Ding et al. (2014) drew the impact of government policies on how to motivate firms to reduce their pollution externalities by producing environment-friendly products. Ding et al. (2016) explored a decision-making mechanism for an environmentally sustainable supply chain that is jointly constrained by environmental carrying capacities and carbon caps, and also takes government policy incentives into account. They addressed the impact of government policy incentives on value transition and profit allotment in different settings of the collaborative supply chain system.

The industrialization of electric vehicles in urban logistics needs the government to issue policies on supporting the construction of infrastructure and encouraging the research and development of key technologies. Governments in different countries have taken differential policies to support measures in the development of electric vehicles, including research and development, demonstration projects, and market support (Ahman, 2006; Lebeau et al., 2016). Mirhedayatian and Yan (2018) evaluated the impacts of supporting policies for electric vehicles on company and environment. Juvvala and Sarmah (2021) carried out a real-life case study showing that car purchase subsidies and zone entry fees are important considerations when promoting electric vehicles in urban logistics. Luo et al. (2014) analyzed how the government’s subsidies for the purchase prices would affect the supply chain of new energy vehicles and proved that it is the most effective way to stimulate the sales of new energy vehicles.

Carbon trading, as a market-based instrument, has been studied and used to optimize resource allocation and reduce carbon emissions. Most of the related research studies have been conducted at the macro level of country or industry, including carbon price, allocation approach of carbon emission rights, comparison of trading schemes in different countries, and the impact of emissions trading on economy and industries (Comodi et al., 2016; Kanamura, 2016; Bakhtyar et al., 2017; Zhang et al., 2017). Wang et al. (2019) proposed an ideal vehicle combination including electric vehicles and fuel vehicles to achieve low carbon emissions and distribution costs in the urban logistics network. There are only a few quantitative assessments of the impact of carbon trading and carbon emission constraint on energy-intensive firms, e.g., profits, cost transfers, innovation, and production planning (Gong and Zhou, 2013). To meet the regulatory requirements, Benjaafar et al. (2013) used relatively simple and extensive models to illustrate how to consider carbon emission in operational decisions with procurement, production, and inventory management. Toptal et al. (2014) studied the joint decision of manufacturing firms’ production planning and green technology investment under the constraint of carbon cap-and-trade policy and compared the impact of different policies on manufacturing firms’ optimal order quantity and green technology investment decision.

Existing studies of electric vehicles by and large have focused on the passenger vehicles with associated technologies, infrastructures, and government policies. The research on the marketization potential in urban electric logistics industry remains rare, combined with government policies and the carbon trading context. To fill this research gap, our study intends to explore the economic feasibility of electric vehicles in urban logistics by taking the view of both manufacturers and business users. The pricing model with an applicable government policy scheme is deduced to be compatible with the market development mechanism.

It is the government’s responsibility to internalize environmental externalities by motivating business organizations to reduce emissions through regulation and incentive policies. In the logistics service industry, it is shown as the gradual replacement of conventional fuel vehicles with electric vehicles promoted by government incentive policies, leading to the marketization of electric vehicles in urban logistics. China has become world’s largest producer and seller of new energy vehicles, and world’s largest market for electric vehicles developed with policy support. With the consideration of fast development of electric vehicles in the Chinese market, promoted by government policies, we will focus our study on the reality in China. In terms of government policies in China, the purchase tax of new energy vehicles was exempted in 2014, and various related policies were subsequently issued with a clear indication to support the application of electric vehicles in urban logistics. According to the statistics of China Association of Automobile Manufacturers (CAAM), electric vehicles in urban logistics began to sprout in 2014 and achieved a blowout from 2015 to 2017, encouraged by government policies. However, the government issued a policy for electric vehicles with subsidy scheme change from 2018. Compared with previous years, the subsidy dropped sharply. Financial subsidy was reduced by 40% and over 50% in 2018 and 2019, respectively, and announced to be withdrawn in 2023, which may significantly impact the marketization of electric vehicles in urban logistics. Withdrawing subsidies is expected and will come sooner or later. The relevant issue is how to arrange the process of reducing the government subsidy so that it will be compatible with the requirement for the marketization of electric vehicles in urban logistics, leading the industry into a rational development stage in which fair market competition is compatible with low carbon emission and the internalization of environmental externalities.

Consider a supplier–buyer business relationship that involves business transactions of urban logistics vehicles between manufacturers and business users in the competitive market. For convenience, we consider all the manufacturers in the market of urban logistics vehicles as a whole, and all the business users as a whole. Manufacturing companies manufacture logistics vehicles on a made-to-order basis, and sell them directly to consumers. Similarly, logistics service providers purchase logistics vehicles for business operations. Due to the increase in the demand for logistics industry, electric vehicles are gradually replacing conventional fuel vehicles. The manufacturers are beginning to produce logistics electric vehicles to gradually replace the conventional fuel vehicles.

We assume that the production of electric logistics vehicles requires additional initial investment, resulting in a cost disadvantage compared with conventional fuel vehicles in the competitive market. To encourage the manufacturers and logistics service providers to improve their environmental performance by reducing environmental pollution, the government enforces regulation by imposing a penalty on manufacturing and using conventional fuel vehicles, and grants policy incentives to subsidize manufacturing and buying electric logistics vehicles. The government policy incentives will help reduce the cost disadvantage of the electric logistics vehicles in the marketplace.

We assume that in the competitive market, the purchase price of electric logistics vehicles for the business user is relatively similar to the market price, which may be lower than the manufacturer’s production cost of the electric logistics vehicles. The manufacturer will be compensated by the government subsidy only when it sells an electric logistics vehicle and shares the subsidy with the business users by coordinating the transaction price through negotiation. The purchase price of conventional fuel vehicles is determined by the market price.

We further assume that the electric logistics vehicles gradually replace the conventional ones in the market place, which means that the logistics service providers use both electric vehicles and conventional fuel vehicles simultaneously with an increasing usage of the former and decreasing usage of the latter. The usage costs of electric vehicles in urban logistics mainly include vehicle purchase cost and vehicle operation cost (including electricity cost, maintenance cost, insurance cost, and other expenses). Purchase cost includes purchase tax expense. We assume that purchase tax is free for electric vehicles in urban logistics, but is charged at a portion of the purchase price for conventional fuel vehicles.

Carbon trading is used to set the cap for carbon emission and then create a carbon market where emitters can buy and sell emission credits. The cap can be either placed on the source side (manufacturers) or placed on the end side (business users). However, the cap is not allowed to set for both the source side and the end side at the same time for the same industry to avoid double counting. The carbon cap is set at a much lower level than the average level of carbon emissions in producing and using logistics vehicles. Hence, emitters need to buy carbon credits to comply with the carbon cap whenever their carbon emissions exceed the cap. When discussing the scenario under carbon trading context, for convenience, we assume that the carbon cap-and-trade policy is implemented at the same time when the electric logistics vehicles are produced and used.

Regarding government policies, there are three alternatives to be considered: 1) the current implemented scheme in which the government subsidy is proportional to the production cost of electric vehicles, which is based on the fact that the manufacturer’s battery procurement cost is proportional to the production cost of electric vehicles; 2) the rationalized subsidy scheme in which the government subsidy is granted on the increment cost basis, which means the government subsidy only compensates for the additional costs incurred by the investment of electric vehicles compared with those of conventional fuel vehicles; and 3) the rationalized subsidy scheme together with the carbon cap enforced either on the manufacturers or business users. In the following section, these three schemes will be discussed.

As mentioned previously, the break-even analysis of the investment of the electric vehicles in urban logistics involves two important parties, the manufacturers as suppliers and the urban logistics service providers as business users, and both are profit driven. Thus, the payback for their investments must be fully considered. The excessively long payback period of either party will cause them to give up the investments in electric vehicles in urban logistics. Only if they can achieve their break-even relatively simultaneously can there be enough driving force to make the marketization of electric vehicles in urban logistics sustainable. Hence, when studying the market development of electric vehicles in urban logistics from the viewpoints of manufacturers and business users, we need to comprehensively consider the operating conditions and payback periods of both the parties.

The investment in electric vehicles in urban logistics reduces emissions and internalizes environmental costs. Meanwhile, the internalization of environmental costs will have certain impacts on business cashflows. To make the marketization of electric vehicles in urban logistics economically feasible in the competitive market, the government imposes penalties on conventional fuel vehicles to have them bear environmental costs and increase their marginal costs. The incentive policies encourage the investment of electric vehicles in urban logistics by making up for the extra operating costs that manufacturers must pay to produce electric vehicles, thus reducing their marginal costs. It can be seen that in order to make electric vehicles in urban logistics competitive in the market and to raise the awareness of environmental protection, it is essential in the initial stage to implement government policies so that both the production and the usage of electric vehicles in urban logistics are well motivated toward their economic feasibility.

In the next section, our model formulation and analysis will be presented. For convenience, notations in Table 1 are used for defining variables and parameters.

Similarly, from the perspective of business users, the investment valuation of electric vehicles in urban logistics is also estimated in terms of the incremental difference in revenue and cost, compared with conventional fuel vehicles. For the convenience of comparison and analysis, we assume that all electric vehicles in urban logistics produced by the manufacturers are sold in the market. Regarding transportation cost measurement, we assume that it is charged in terms of the maximum allowable load carried by vehicles. Influenced by battery power consumption and charging time, the mileage of electric vehicles in urban logistics is shorter than that of conventional fuel vehicles on average. In addition, business users also bear the fine (in T _o ) brought by conventional fuel vehicles that have not yet been replaced.

When electric vehicles in urban logistics are put into operation and begin to partially replace the conventional fuel vehicles in the market, for year t, the business users will hold electric logistics vehicles (in ∑ t = 1 n Q e t ) accumulatively purchased up to t years, the business users’ annual operating cash flow can be expressed as follows: S G e D e ∑ t = 1 n Q e t + S G o D o ∑ t = 1 n Q t − ∑ t = 1 n Q e t − P e t Q e t + P o Q t − Q e t − A e ∑ t = 1 n Q e t + A o ∑ t = 1 n Q t − ∑ t = 1 n Q e t − T o ∑ t = 1 n Q t − ∑ t = 1 n Q e t , where ( ∑ t = 1 n Q t − ∑ t = 1 n Q e t ) is the amount of conventional fuel vehicles that have not been replaced. By deducting the opportunity cost of only holding conventional fuel vehicles S G o D o ∑ t = 1 n Q t − P o Q t − A o ∑ t = 1 n Q t − T o ∑ t = 1 n Q t , the business users’ annual cash flow can be simplified as S G e D e − G o D o − A e − A o ∑ t = 1 n Q e t − P e t − P o Q e t + T o ∑ t = 1 n Q e t . Then, from the view of logistics service providers as the business users, the net present value of investing in electric vehicles in urban logistics, denoted by NPV _u , can be expressed as follows: N P V u = ∑ t = 1 n S G e D e − G o D o − A e − A o ∑ t = 1 n Q e t − P e t − P o Q e t + T o ∑ t = 1 n Q e t 1 + r 2 t . (6)

We refer to the sales price of manufacturers selling electric vehicles in urban logistics and the purchase price by business users collectively as the transaction price. As the suppliers and buyers of electric vehicles in urban logistics, both the manufacturers and the business users are facing the issue of jointly determining the optimal transaction price so that both of them can reach their break-even at relatively the same time, which is crucial to the sustainable development of electric vehicles in urban logistics market. Therefore, as the overall benefit of the integrated system including manufacturers and business users is optimum, both parties will come to collaboration only if the following holds: N P V m = − I + ∑ t = 1 n P e t − P o Q e t − C e t − C o Q e t + T e t + T o Q e t 1 + r 1 t ≥ 0 , (7) N P V u = ∑ t = 1 n S G e t D e t − G o D o − A e − A o ∑ t = 1 n Q e t − P e t − P o Q e t + T o ∑ t = 1 n Q e t 1 + r 2 t ≥ 0 . (8)

These are the necessary prerequisites for a lasting collaborative relationship between the manufacturers and business users. The goal of the collaboration is to realize the unification of both individual interest and collective interest, and to finally achieve Pareto optimum. To satisfy Eqs 7, 8, the manufacturers and business users must negotiate a transaction price P _et that satisfies both parties. Since the determination of P _et goes through a negotiation process between the manufacturers and business users, we use the Rubinstein game approach (Rubinstein, 1982) to describe the bargaining process of negotiating P _et . First, we identify a feasible region that satisfies the interests of both parties. From Eqs 7, 8, we can obtain the lower and upper bounds of P _et denoted by P _e-min (from △NPV _m ≥ 0) and P _e-max (from △NPV _u ≥ 0), respectively, as follows: P e − min = I − ∑ t = 1 n T e t + T o Q e t − C e t − C o Q e t − P o Q e t 1 + r 1 t ∑ t = 1 n Q e t 1 + r 1 t , (9) P e − max = ∑ t = 1 n S G e D e − G o D o − A e t − A o t + T o ∑ t = 1 n Q e t + P o Q e t 1 + r 2 t ∑ t = 1 n Q e t 1 + r 2 t , (10) where [P _e-min , P _e-max ] is the feasible interval to find the optimal value of P _et within which the manufacturers and business users make a reasonable allocation of mutual benefits (profits). According to the Rubinstein’s game theory, for both parties to coordinate, the individual portion of shared profits under coordination must be not less than that of a non-coordination scenario. Assuming an indefinite bargaining game between the manufacturers and business users in our case, let x* (0 ≤ x* ≤ 1) be the profit share of the manufacturers and 1-x* be the profit share of the business users. The unique sub-game perfect Nash equilibrium, namely, the optimal profit sharing ratio of the manufacturer, is expressed as x* = (1-δ ₂)/(1-δ ₁ δ ₂) (if δ ₁ = δ ₂, then x* = 1/(1+δ)), where δ ₁ and δ ₂ are the discount factors of the manufacturers and business users, respectively (here, discount factor is the patience degree of participants for 0 ≤ δ ≤ 1, which can be seen as the cost of bargain). Considering that the determination of profit sharing depends on the transaction price that satisfies both the manufacturers and business users, the ratio of the profit allotment between the two parties can then be expressed in terms of the transaction price as follows: x * 1 − x * = 1 − δ 2 δ 1 − δ 1 δ 2 = P e − max − P e * P e * − P e − min , (11) where P _e * denotes the optimal value of the transaction price and P _e-max and P _e-min are shown in Eqs 9, 10. Rewriting Eq. 11 with the coordination of participants, the optimal transaction price is obtained as follows: P e * = δ 2 − δ 1 δ 2 P e − max + 1 − δ 2 P e − min 1 − δ 1 δ 2 = 1 − x * P e − max + x * P e − min . (12)

As seen previously, the lower and upper bounds of P _et depend on the sales volume, initial investments of the manufacturers, production costs, government policies, incremental revenue, and operating cost of electric vehicles in urban logistics and also the price of fuel logistics vehicles. They have joint impacts on the determination of the optimal transaction price through the negotiation process. Given the production and sales volume and other parameters, the number of years for the manufacturers and business users to reach their break-evens can be obtained, which present their payback periods. The optimal coordinated transaction price is obtained when the game process comes to an equilibrium at which both parties reach their break-evens at relatively the same time.

In this section, in addition to the government subsidy offered to compensate for the cost disadvantages of electric vehicles, we further study the impact of introducing carbon trading on the economic valuation of the electric vehicles in urban logistics.

With the consideration of carbon emission and carbon cap, when introducing the electric logistics vehicles to gradually replace the conventional fuel vehicles, carbon credits for additional carbon emissions are involved. The carbon emissions at time period t (t = 1,2, … ,n) can be expressed as follows: e e Q e t + e o Q t − Q e t = q t + Z   f o r   q t ≥ 0 , (13) where (Q _t − Q _et ) is the sales quantity of conventional fuel vehicles. Equation 13 states that the total carbon emissions in year t is the sum of the emissions associated with producing both the electric logistics vehicles and the conventional fuel vehicles, which equals to the annual carbon cap plus the purchased carbon credits from an external source in year t. Carbon emission costs incurred from purchasing carbon credits (i.e., carbon emission rights) at time period t can then be expressed as follows: H t = h t q t = h t e e Q e t + e o Q t − Q e t − Z . (14)

Equation 14 is derived from Eq. 13. As we assumed, the carbon cap can be either placed on the manufacturers or on the business users. Thus, we will discuss these two scenarios separately. If it is the manufacturer who participates in carbon trading, with the carbon credit purchase cost of Eq. 14 taken into consideration, we can obtain its NPV from Eq. 7 as follows: N P V ′ m = − I + ∑ t = 1 n P e − P o Q e t − C e t − C o Q e t + T e t + T o Q e t − H t 1 + r 1 t ≥ 0 . (15)

If it is the business user who participates in carbon trading, with the carbon credit purchase cost of Eq. 14 taken into consideration, we can obtain its NPV from Eq. 8 as follows: N P V ′ u = ∑ t = 1 n S G e D e − G o D o − A e t − A o t ∑ t = 1 n Q e t − P e t − P o Q e t + T o ∑ t = 1 n Q e t − H t 1 + r 2 t ≥ 0 , (16) where the impact of the cost of purchasing carbon credits is similar to the penalty cost imposed on carbon emissions. This is true in reality because if the carbon caps were not set low enough, then firms would not actively reduce carbon emissions.

In a way similar to Section 4.4, the optimal transaction price between the manufacturer and business user can be determined through negotiation using the Rubinstein game approach. For the case of the manufacturers to participate in carbon trading, we use Eq. 15 to develop the lower bound of the transaction price as follows: P e − min = I − ∑ t = 1 n T e t + T o Q e t − C e t − C o Q e t − H t − P o Q e t 1 + r 1 t ∑ t = 1 n Q e t 1 + r 1 t . (17)

Together with the upper bound of the transaction price as shown in Eq. 10, we can determine the optimal transaction price with the carbon cap set on manufacturers. On the other hand, for the case of business users to participate in carbon trading, we use Eq. 16 to develop the upper bound of the transaction price as follows: P e − max = ∑ t = 1 n S G e D e − G o D o − A e t − A o t + T o ∑ t = 1 n Q e t − H t + P o Q e t 1 + r 2 t ∑ t = 1 n Q e t 1 + r 2 t . (18)

Together with the lower bound of the transaction price as shown in Eq. 9, we can determine the optimal transaction price with the carbon cap set on business users.

Electric vehicles in urban logistics include minivan and light truck, mainly used in the fields of intra-city express and fresh food logistics. The data are estimated on average based on real market situation shown in Table 3. With the development and market penetration of the electric logistics vehicles, its variable production cost gradually reduces as the battery cost decreases with technological improvement.

Table 4 shows that under the current subsidy scheme and pricing, the NPVs of the electric vehicles in urban logistics for the manufacturers and business users will become positive in 2024 and 2020, respectively, i.e., the former will achieve its break-even in 10 years and the latter in 6 years. This incurs unbalanced payback periods between the two parties and will result in unwillingness from the manufacturers, implying that the current subsidy scheme may be inappropriate in a certain sense.

Under the current implemented government subsidy mode with T’ _et = βC _et (0 < β ≤ 1), i.e., the government subsidy is proportional to the production cost of electric vehicles. Using the unit variable cost and current actual government subsidy for the electric logistics vehicles, the value of the coefficient β can be calculated to show its trend in Figure 2. The current government subsidy seems illogical since it incurs unbalance in the market due to its being directly proportional to the production cost of the electric vehicle, which cannot be compatible with the reality and does not properly consider the market competition with fuel vehicles. The initial growth trend of β indicated that the battery costs fell faster than subsidies and also meant that the subsidies were excess than needed. The delayed reaction of the government resulted in the sharp drop from 2017 to 2019. Due to the market competition, it would be unlikely for the manufacturers to make adjustment of the sales price for the subsidy decline. The loss caused by the subsidy decline basically should be covered by production cost reduction or sales increase, which needs longer time periods. Hence, this may not be acceptable by the manufacturers who have to take much longer to achieve their break-even. Noticing that it is the manufacturers that make the initial investment of producing the electric logistics vehicles by undertaking heavy upfront cost relative to the business users, this implies that the manufacturers originally take longer time to reach its break-even than the business users, which is evidenced in reality.

From the aforementioned analysis, we can see that the current implemented government subsidy scheme needs to be rationalized to follow the market mechanism. Therefore, Eq. 2 is used to rationalize the government subsidy on the incremental cost basis, and Eq. 12 is used to coordinate the transaction price for both parties so as to ensure that both parties can reach their break-evens at relatively the same time and achieve sustainable development when the subsidy ceases. Based on the incremental cost, the value of the subsidy coefficient α is set to be less than 1 with three levels of 15%, 23%, and 30% for sensitivity analysis. The rationalized government subsidy, coordinated transaction price, and payback time periods for both parties are calculated with the collected data.

The comparison in Table 5 shows that for the rationalized scheme with α levels of 23%, the accumulated subsidies during the time periods toward the break-even are 336,000 RMB, which is rather less than that under the current implemented subsidy scheme of 418,000 RMB. Their payback periods are relatively the same, which means the manufacturer is overcompensated under the current subsidy scheme.

Under the rationalized government subsidy, Table 4 shows that both the manufacturers and business users reach their break-even at the same time in 9, 8, and 7 years at the subsidy coefficient level of 15%, 23%, improve routing for urban logistics distribution using and 30%, respectively. The investment break-even time periods are shortened with the increase of the government subsidy, which leaves the room for negotiation between business organizations and governments regarding how long is acceptable for the manufacturers and business users to break-even while facing the risks they undertake. Under the rationalized subsidy scheme with α levels of 23%, the coordinated transaction prices of the electric vehicles are 126,700 RMB, higher than 122,500 RMB under the current subsidy scheme. This implies that the reduced government subsidies compensate manufacturers by the increased transaction prices. In contrast, the higher purchase price paid by the business users prolongs their payback periods. It explains that under the rationalized scenario, the incomes associated with electric logistics vehicles are rebalanced between the manufacturers and business users.

The impact of environmental standards on decision performance can be analyzed by observing the number of years to reach break-even changes as the carbon cap is introduced. With the carbon cap enforced on either the manufacturers or on the business users, it is expected that their payback time periods will be prolonged due to the imposed cost impact of paying for carbon credits, as shown by the case example in Table 4 that their payback time periods are extended for one more year . The total carbon emissions were 11.3 million tons in 2015 and will be 5.5 million tons in 2025, which is cut in half in 10 years for the given estimated data. We could see the positive effect brought by the marketization of electric logistic vehicles under the carbon trading context.

The government subsidy is not affected by the enforcement of the carbon cap as assumed in Eq. 2. Compared to the previous scenario not considering carbon trading, as shown in Table 6, the coordinated transaction price is higher when the carbon cap is enforced on the manufacturers due to the carbon emission costs undertaken by the manufacturers. In contrast, the coordinated transaction price is lower when the carbon cap is enforced on the business users. These can be explained by Eqs 17, 18 where the impacts of the carbon costs paid due to excess emission over the carbon cap for the manufacturers and business users are opposite to the transaction price.