Carbon emission permit matchmaking transaction method based on Gal-Shapley stable matching algorithm

Abstract

The invention provides a carbon emission permit matchmaking transaction method based on a Gal-Shapley stable matching algorithm. The method comprises the following steps: step 1, listing by both parties: carrying out buying or selling listing according to the demanded quantity or quota margin of the parties; step 2, social welfare calculation: the system integrates the lists of the two parties, considers that the lists are sold as producers and the lists are purchased as consumers, and combines any two of the buying lists Hi and the selling lists Lj to calculate the sum of the two lists to obtain the social welfare of the market; step 3, preference sorting: regarding each listing as a participant, setting that a matching set of the other party has own preference, and taking a social welfare value generated by combining any listing of the two parties as a preference value of the two listing parties; step 4, performing bilateral matching based on a Gal-Shapley algorithm: forming a transaction process according to listing information of each subject and preference sorting, and performing bilateral matching; and step 5, transaction settlement.

Description

Technical Field

[0001] The present invention discloses a method for improving the efficiency of carbon emission rights trading, in particular to a carbon emission rights matching trading method based on the Gale-Shapley stable matching algorithm, and belongs to the technical field of carbon trading. Background Art

[0002] The carbon trading mechanism is one of the important ways to achieve low-carbon development. In February 2021, my country officially implemented the "Carbon Emission Trading Management Measures (Trial)", and the Shanghai trading platform was launched on July 16 of the same year, marking the entry of domestic carbon emission trading into the market construction and development stage. By the end of 2023, the annual carbon dioxide emissions covered by the national carbon market will be about 5.1 billion tons, involving 2,257 companies participating in the performance and liquidation, making it the world's largest carbon market covering greenhouse gas emissions at this stage. However, judging from the performance of the national carbon trading market in 2023, in the first half of the year, the market activity was low and carbon prices were relatively stable, but in the second half of the year, there was a surge in trading volume and drastic fluctuations in carbon prices.

[0003] At present, my country's carbon trading market has been established for a short time, and the current market trading model is immature. The main trading methods are listing agreement transactions and bulk agreement transactions. When there is a surge in the number of listing applications, a large number of transaction orders reported by corporate entities may accumulate, resulting in a decrease in market transaction efficiency. Participants in the new market lack experience and may engage in irrational buying and selling operations such as emotional transactions, which in turn aggravates carbon price fluctuations, forms a vicious circle, and affects the overall market development. Based on the digital platform of the carbon trading market, the present invention proposes a carbon emission rights matching transaction method based on the Gale-Shapley stable matching algorithm, aiming to improve market transaction efficiency, stabilize carbon price fluctuations, and cope with the adverse effects of the surge in trading volume. By establishing a multi-agent model of the matching transaction model for simulation experiments, the applicability of the model is evaluated from the perspective of market transaction efficiency and carbon price fluctuations. Summary of the invention

[0004] Technical problems solved by the present invention:

[0005] In order to improve the transaction efficiency of carbon trading and stabilize carbon price fluctuations, the present invention proposes a market matching mechanism based on the Gale-Shapley stable matching algorithm. The market matching mechanism uses the matching principle to match the quotations of buyers and sellers on the basis of the market digital platform to facilitate transactions. It can quickly and accurately match buyers and sellers, reduce transaction costs, and enhance market liquidity.

[0006] The solution of the present invention:

[0007] The first step is to list the two parties. In the two-way listing transaction of carbon trading, multiple companies selling carbon quotas and multiple companies buying carbon quotas enter the market and buy or sell according to their own demand or quota surplus. Assume that there are m buyers and sellers {H1, H2, ..., H i ……,H m}, n sold listings {L1,L2,……,L j ……,L n}, the company H that sells carbon quotas i The listing price is The listing volume is Company L that buys carbon quotas j The listing price is The listing volume is First, before the carbon market opens, the trading platform announces the market reference price P x , and bilateral enterprises will list according to the reference price,

[0008]

[0009] Market players conduct the following automatic matching and transaction based on the information listed by both parties. The specific key steps include:

[0010] The second step is to calculate social welfare. According to the definition of market social welfare in economics, the social welfare value is the sum of consumer surplus and producer surplus, and the market configuration that maximizes total welfare is called efficient. Therefore, before starting to match, the system will integrate the listings of both parties, considering that the listing sale is the behavior of the producer, and the listing purchase is the behavior of the consumer, and will buy the listing H i (i=1,2,……,m) and the selling order L j (j=1,2,……,n) The total surplus can be obtained by calculating the sum of any two of them, which is the social welfare of the market, as shown in formula (1):

[0011]

[0012] Among them, U ij For the social welfare of the market; ij is the transaction price; Q ij It's the trading volume.

[0013] The third step is preference sorting. Each listing is regarded as a participant, and each participant is assumed to have his own preference for the matching set of the other party. The social welfare value generated by the combination of any listings of both parties is taken as the preference value of the two parties.

[0014] The preference value U generated by pairing with different listings ij , sort them from largest to smallest as the preference ranking for listing. Get the preference ranking list for buying listings As shown in formula (2) and the preference ranking list of selling listings As shown in formula (3).

[0015]

[0016] The fourth step is bilateral matching based on the Gale-Shapley algorithm. According to the listing information and preference ranking of each subject, a transaction process is formed and bilateral matching is performed. Figure 1 As shown, specifically:

[0017] The first company to buy carbon quotas will be listed on H i in accordance with List the companies that sell carbon quotas with a preference ranking of 1 Make a matching request;

[0018] ·Then the company that sold the carbon quota will list L j in accordance with Matching selection is performed according to the following rules:

[0019] √ If the company that sells carbon quotas lists L j If the number of accepted requests is 0, the next round of matching will begin;

[0020] √ If the quantity is 1, the match is temporarily accepted; if the quantity is 2 or more, the match is sorted by preference, and the unmatched carbon quota buying enterprise H′ i Enter the next round of matching;

[0021] ·In the next round of matching, companies that buy carbon quotas will be listed in order of preference The carbon quota selling enterprise with a ranking of +1 will be listed for matching request, and so on;

[0022] Finally, when all the carbon quota sellers or buyers are successfully matched, the matching ends.

[0023] Step 5: Transaction settlement. After the matching is completed, a stable matching result is formed. The transactions are completed in sequence according to the matching results. The smaller quota of one party is used as the transaction volume as shown in formula (4), and the weighted average of the listed prices of both parties is used as the transaction price as shown in formula (5). The weighted average of all completed orders is used as the market reference price for the next round of matching as shown in formula (6).

[0024] Q ij =Min(Q i ,Q j ) (4)

[0025]

[0026] Among them, Qij is the volume, Q i is the buyer listing quantity, Q j is the seller listing quantity, P ij is the transaction price, P i The buyer's bid, P j is the seller's bid, P x ′ is the market reference price.

[0027] The advantages and beneficial effects of the present invention are:

[0028] 1. Reduce transaction costs, improve order efficiency, and stabilize carbon price fluctuations. The matching transaction model matches the orders of buyers and sellers through an automated system, which is particularly effective in markets with large trading volumes. Traditional listing transactions are inefficient and may lead to high transaction costs. The matching transaction model relies on matching algorithms to process a large number of orders in a short period of time, reducing transaction delays and human errors, thereby significantly reducing transaction costs, improving market liquidity, reducing carbon price fluctuations, and increasing traders' trust and participation in the market, making them more willing to participate in market transactions.

[0029] 2. Combination of stability and maximization of social welfare. The Gale-Shapley algorithm itself is committed to finding stable matches, that is, no buyer or seller will prefer to change the existing matching status. When the matching basis adopted by the present invention is the maximum social welfare of each order to be traded, the algorithm not only ensures stability, but also pays attention to the social welfare of the market as a whole. This means that the resource allocation of the trading market is more efficient and can maximize the economic benefits of the market as a whole.

[0030] 3. Transaction anonymity and market fairness. In the traditional trading model, traders may take advantage of information asymmetry to gain unfair advantages. In the matching trading model, the identity of the trader is kept confidential, and only the trading instructions and matching results are public. This anonymity can effectively prevent market manipulation and insider trading, ensure that every trader can participate in the transaction under the same conditions, and improve the fairness and transparency of the market. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 It is a flowchart of the two-way listing transaction model.

[0032] Figure 2 It is a schematic diagram of multi-agent relationship.

[0033] Figure 3 Transaction efficiency fluctuation chart of simulation results. DETAILED DESCRIPTION

[0034] The simulation model data is based on the national carbon trading market transaction data published by the Shanghai Environment and Energy Exchange. The main simulation tools in this invention include: 1. Programming language: JavaSE-1.8; 2. Simulation framework: Mason 21; 3. Development environment: Eclipse IDE 2020-12.

[0035] ① Carbon emission rights matching transaction model process

[0036] The process of integrating the two-way listing trading model with the market matching trading mechanism is as follows Figure 1 Specifically, enterprises participate in matching transactions and apply for listings based on the information published by the market. The market receives listing information from both buyers and sellers, generates a preference ranking of the number of counterparties based on the listing volume and listing price of each listing, and then performs Gale-Shapley matching based on the preference ranking list to form a stable matching result and conduct transaction settlement. Finally, the market reference price of carbon quotas is updated based on the settlement records to enter the next round of transactions.

[0037] ②Multi-agent relationship modeling

[0038] In the entire trading model, the participating entities can be abstracted into three categories: 1. Government; 2. Key emission units; 3. Carbon trading platform. The key attributes that link these three are carbon prices and carbon quotas. Carbon prices are formed after the key emission units participating in the transaction conduct market transactions based on their own business conditions and decision-making criteria. It is the core of the trading model. Effective price information can provide key emission units with a reference for calculating emission reduction costs and improve their enthusiasm for emission reduction. Carbon quotas are a means for the government to control the total amount of carbon emissions, and are also an important basis for key emission units to make trading decisions. Draw an interactive relationship diagram of the relationship between each subject, such as Figure 2 shown.

[0039] ③ Main body model construction

[0040] Government subject modeling: From the perspective of the initial allocation of carbon emission rights, the behavior of government agents in the carbon trading part is mainly to allocate free carbon emission quotas to key emission units in the primary market according to the intensity-based baseline method. Referring to the data released by the Ministry of Ecology and Environment, there are 2,257 key emission units participating in the national carbon trading market in 2023, with an annual emission coverage of more than 5 billion tons. According to the quota allocation and settlement rules, it is believed that the total amount of pre-allocated quotas for the second compliance period (2021-2022) is approximately 7 billion tons, as shown in formula (7). On the basis of setting the total amount of quotas, assuming that each key emission unit obtains quotas independently, without considering the power generation and heating capacity of specific enterprises, according to the normal distribution with a mean of 3,101,462 tons and a standard deviation of 700,000 tons as shown in formula (8), the quota amount allocated to the enterprise is greater than 26,000 tons as shown in formula (9), and the carbon quota quantity of each enterprise is randomly allocated to achieve a total amount of carbon quotas of approximately 7 billion tons. It can be concluded that the constraints for simulating the government's allocation of quotas for key emission units are:

[0041] Total quota limit:

[0042]

[0043] Distribution of quota allocations:

[0044] CarbonCredit~N(3101462,700000 2 )(8)

[0045] Limitation of quota for key emission units:

[0046] CarbonCredit i ≥26000 (9)

[0047] CarbonCredit is the amount of carbon quota allocated to the i-th key emission unit.

[0048] Enterprise entity modeling:

[0049] Here, key emission units are divided into two categories according to the positive or negative difference between emissions and quotas. The first category is enterprises that buy carbon quotas if their actual emissions are greater than their carbon quotas, and the second category is enterprises that sell carbon quotas if their actual emissions are less than their carbon quotas. The number of enterprises that buy carbon quotas is as shown in formula (10), and the actual emission limit is as shown in formula (12). The number of enterprises that sell carbon quotas is as shown in formula (11), and the actual emission limit is as shown in formula (13). According to the trading statistics of the national carbon trading market in 2023 announced by the Shanghai Environment and Energy Exchange, the total trading volume is 211,943,768 tons. Based on the survey of the trading model of the national carbon trading market, it is assumed that the purpose of corporate trading is only to fulfill the contract. Thus, further parameter settings are obtained. The gap of key emission units is about 212 million tons, that is, the total amount of carbon quotas that enterprises that buy carbon quotas need to purchase is about 212 million tons, and the total amount of surplus quotas of enterprises that sell carbon quotas is also about 212 million tons. The specific expression is as follows: Formula (14). In this way, the performance objectives of each enterprise can be achieved after all transactions are completed. However, in fact, there will still be situations where enterprises that sell carbon quotas after performance settlement have excess quotas or enterprises that buy carbon quotas fail to purchase sufficient quotas. Therefore, when realizing the method of generating the actual carbon emissions of each enterprise, the present invention adopts a uniformly distributed random number and sets a boundary to achieve the purpose of limiting the total amount and retaining randomness. It can be concluded that the constraint conditions for generating the actual emission data of key emission units are:

[0050] Distribution of the number of enterprises of different categories among key emission units:

[0051] Number High ~U(0,2257);(10)

[0052] Number Low =2577-Number High (11)

[0053] The actual carbon emission limits of enterprises that buy carbon quotas and enterprises that sell carbon quotas are:

[0054] CarbonEmission High ≥CarbonCredit High ; (12)

[0055] CarbonEmission Low ≤CarbonCredit Low (13)

[0056] Total transaction limit:

[0057] ∑TradeVolume High≈212000000≈∑TradeVolume Low (14)

[0058] Under the two-way listing and trading model combined with the market matching mechanism, enterprises that buy carbon quotas and enterprises that sell carbon quotas will apply for buying and selling listings respectively. It is assumed that the purpose of enterprises participating in carbon trading is only to fulfill their obligations and they are active in trading in the market. After the fulfillment purpose is achieved, they will no longer participate in carbon trading. Based on this setting, each enterprise will apply for listing during the market matching stage. The listing information includes: 1. Enterprise number; 2. Listing quantity; 3. Listing price; 4. Listing direction, as shown in Table 1.

[0059] Table 1 Listing attribute information of enterprise entities

[0060]

[0061] According to the transaction model, the constraints for enterprise listing are as follows:

[0062] Listing quantity constraints:

[0063] 0≤TradeVolume≤100000≤|CarbonEmission-CarbonCredit| (15)

[0064] If(|CarbonEmission-CarbonCredit|)≤1000

[0065] Then TradeVolume=|CarbonEmission-CarbonCredit| (16)

[0066] Listing price constraints:

[0067] 0.9*P x ≤Tradeprice≤1.1*P x (17)

[0068] Character explanation: TradeVolume is the transaction volume, CarbonEmission is the actual carbon emission of the enterprise, CarbonCredit is the carbon quota of the enterprise, Tradeprice is the listing price, P x This is the market reference price.

[0069] Market entity modeling:

[0070] In the simulation model, market entities act as a bridge between corporate entities. They first provide the market reference price of carbon quotas for traders participating in the market, then accept the listing and trading information declared by the enterprises, match transactions in accordance with the trading rules of the two-way listing and trading model combined with the market matching mechanism, settle transactions based on the matching results, record settlement information and update the market reference price of carbon quotas.

[0071] ④Transaction efficiency analysis

[0072] Under the market matching transaction mechanism of the simulation model, enterprises list by referring to the market price, and match through matching transactions, and pair the buy and sell listings that can achieve the maximum market efficiency to reach a transaction. Here, the ratio of the number of completed orders to the number of listed orders in each round is used as the transaction efficiency, as shown in formula (18), and the fluctuation of transaction efficiency is analyzed.

[0073] Rate=Num deal / Num list* 100% (18)

[0074] Among them, Rate is the transaction efficiency, Num deal is the number of completed orders, Num list The number of pending orders.

[0075] First, the simulation results are divided into buy-listing (High) transaction efficiency and sell-listing (Low) transaction efficiency. The descriptive statistics are shown in Table 2. The fluctuation of transaction efficiency is as follows: Figure 3 shown.

[0076] Table 2 Descriptive statistics of buy and sell listing transaction efficiency

[0077]

[0078] The whole process of the simulation trading experiment is divided into the early stage, the middle stage and the late stage. In the early stage of the transaction, under the pressure of fulfilling the contract and clearing the payment, the enterprise entities with the intention of buying and selling concentratedly submitted transaction declarations to the market entities. Under the setting of the matching transaction mode, the transaction efficiency between buyers and sellers can be stabilized at about 50%. However, as the number of enterprises and the listing volume decrease, the transaction efficiency begins to fluctuate greatly, reaching a maximum of 100% and a minimum of close to 5%. The results show that the matching transaction mode is suitable for enterprises with high enthusiasm for participating in carbon trading and a large number of transaction orders. The transaction efficiency is high and stable, while when the number of participating enterprises is small, the transaction efficiency will fluctuate greatly.

[0079] ⑤Analysis of carbon price fluctuations

[0080] From the descriptive statistical results of the average transaction price, we can see that the standard deviation of the simulation results is 1.35, and the maximum carbon price of 59.60 and the minimum carbon price of 52.65 are both within 10% of the mean, with a small degree of variation. From the perspective of skewness, the simulation results are right-skewed, that is, as transactions continue to occur, the number of companies participating in the transaction gradually decreases, and the stabilizing effect of the matching transaction model on the carbon price is also gradually decreasing. From the perspective of kurtosis, the simulation results are flat compared to the normal distribution, indicating that the overall transaction price does not change much, and the matching transaction model has played a role in stabilizing the carbon price.

Claims

1. A carbon emission rights matching transaction method based on the Gale-Shapley stable matching algorithm, characterized by: The steps include: Step 1: Bilateral listing: In the bidirectional listing transaction of carbon trading, multiple companies selling carbon quotas and multiple companies buying carbon quotas enter the market and buy or sell according to their own demand or quota surplus; Step 2: Calculate social welfare: The social welfare value is the sum of consumer surplus and producer surplus, and the market configuration that maximizes total welfare is called efficient; therefore, before starting to match, the system will integrate the listings of both parties, considering that the listing sale is the producer behavior and the listing purchase is the consumer behavior, and will buy the listing H i With sell listing L j The total surplus calculated by combining any two of them is the social welfare of the market; Step 3, preference sorting: treat each listing as a participant, assume that each participant has his own preference for the matching set of the other party, and take the social welfare value generated by the combination of any listings of both parties as the preference value of the two parties; Step 4: bilateral matching based on Gale-Shapley algorithm: form a transaction process based on the listing information and preference ranking of each entity, and perform bilateral matching; Step 5, transaction settlement: After the matching is completed, a stable matching result is formed, and the transactions are completed in sequence according to the matching results. The transaction volume is the smaller quota, and the weighted average of the listed prices of both parties is the transaction price; and the weighted average of all completed orders is used as the market reference price for the next round of matching.

2. According to claim 1, a carbon emission rights matching transaction method based on the Gale-Shapley stable matching algorithm is characterized by: In step 1, there are m purchase listings {H1, H2, ..., H i ……,H m }, n sold listings {L1,L2,……,L j ……,L n }, the company H that sells carbon quotas i The listing price is The listing volume is Company L that buys carbon quotas j The listing price is The listing volume is Before the carbon market opens, the trading platform announces the market reference price P x , and bilateral enterprises will list according to the reference price, 3. According to claim 1, a carbon emission rights matching transaction method based on the Gale-Shapley stable matching algorithm is characterized by: In step 2, social benefits are specifically: Among them, U ij For the social welfare of the market; ij is the transaction price; Q ij It's the trading volume.

4. A carbon emission rights matching transaction method based on the Gale-Shapley stable matching algorithm according to claim 1 or 3, characterized in that: In step 3, the preference values ​​U generated by pairing with different listings are ij , and arranged them from largest to smallest as a preference ranking for listing; Get a preference-ordered list of buy listings As shown in formula (2); the preferred ranking list of sell listings As shown in formula (3); 5. According to claim 1, a carbon emission rights matching transaction method based on the Gale-Shapley stable matching algorithm is characterized by: In step 4, specifically: 4.1 The first company to buy carbon quotas is listed on H i in accordance with List the companies that sell carbon quotas with a preference ranking of 1 Make a matching request; 4.2 Then the company that sold the carbon quota is listed on the L j in accordance with Matching selection is performed according to the following rules: 4.3 The companies that buy carbon quotas in the next round of matching will be listed in the order of preference The carbon quota selling enterprise ranked plus 1 in the middle will be listed for matching request, and so on; 4.4 Finally, when all the matches between the enterprise selling carbon quotas or the enterprise buying carbon quotas are successfully completed, the matching ends.

6. According to claim 5, a carbon emission rights matching transaction method based on the Gale-Shapley stable matching algorithm is characterized by: In step 4.2, if the enterprise selling carbon quotas lists L j If the number of accepted requests is 0, the next round of matching will be entered; if the number is 1, the matching will be temporarily accepted; if the number is 2 or more, the matching will be sorted according to the preference, and the carbon quota buying enterprise H′ that has not been successfully matched i Enter the next round of matching.

7. According to claim 1, a carbon emission rights matching transaction method based on the Gale-Shapley stable matching algorithm is characterized by: In step 5, the smaller quota is used as the transaction volume: Q ij =Min(Q i ,Q j ) (4) Among them, Q ij is the volume, Q i is the buyer listing quantity, Q j The seller listing volume.

8. According to claim 7, a carbon emission rights matching transaction method based on the Gale-Shapley stable matching algorithm is characterized by: In step 5, the weighted average of the two parties' listing prices is the transaction price: Among them, P ij is the transaction price, P i The buyer's bid, P j It is the seller's bid.

9. The carbon emission rights matching transaction method based on the Gale-Shapley stable matching algorithm according to claim 8 is characterized by: In step 5, the weighted average of all executed orders is used as the market reference price for the next round of matching: Among them, P′ x It is the market reference price.

Images