Crypto asset evaluation system

Abstract

Provided is a system capable of improving the evaluation accuracy of crypto assets through agent-based simulation by dynamically considering characteristic elements of the crypto assets that differ for each platform (supply volume, supply method, service value, market size, productivity, etc.) as a complex system. Model parameters including parameters that define the trading characteristics of crypto assets on a platform is input through an input unit 11. A model calculation unit 20 determines information regarding the future value of the crypto assets in accordance with an evaluation model in which the trading trends of the crypto assets on the platform is defined using the model parameters input to the input unit 11. The information regarding the future value of the crypto assets determined by the model calculation unit 20 is output through an output unit 12.

Description

Cryptocurrency valuation system

[0001] This invention relates to a system for evaluating crypto assets such as tokens.

[0002] Various methods have been proposed for valuing crypto assets such as tokens (see Non-Patent Document 1).

[0003] “Tokenomics: Dynamic adoption and valuation.” Cong, L. W. , Li, Y. &Wang, N. (2021), The Review of Financial Studies, 34(3), 1105-1155.

[0004] However, since the valuation of a particular cryptocurrency does not take into account asset characteristics such as different token issuance schedules, growth rates, staking with other cryptocurrencies, and the trading characteristics of various agents, there is room for improvement in evaluating that cryptocurrency.

[0005] Therefore, the present invention aims to provide a system that can improve the accuracy of cryptocurrency valuation by dynamically considering elements of cryptocurrency characteristics (supply quantity, supply method, service value, market size, productivity, etc.) that differ for each platform as a complex system through agent-based simulation.

[0006] The crypto asset valuation system of the present invention comprises: an input unit into which model parameters including parameters that define the trading characteristics of a crypto asset on a platform are input; a model calculation unit that determines information regarding the future value of a crypto asset according to a valuation model that defines the trading trends of the crypto asset on the platform based on the model parameters input to the input unit; and an output unit that outputs information regarding the future value of the crypto asset determined by the model calculation unit.

[0007] Diagram illustrating the configuration of a crypto asset valuation system as one embodiment of the present invention. Flowchart showing the algorithm according to the present invention. Flowchart showing the model initialization algorithm. Flowchart showing the first token supply algorithm. Flowchart showing the second token supply algorithm. Flowchart showing the token staking algorithm. Flowchart showing the price determination algorithm. Flowchart showing the token trading algorithm. Flowchart showing the token unlock algorithm. Diagram illustrating the output format of the token price trend, etc. Diagram illustrating the output format of the number of token holders trend. Diagram illustrating the output format of the token trading volume trend. Diagram illustrating the output format of the market size trend. Diagram illustrating the output format of the token supply volume trend. Diagram illustrating the input format of the token supply schedule, etc. Diagram illustrating the input format of the market size growth schedule, etc. Diagram illustrating the input format of the token trading characteristics parameters, etc.

[0008] (Configuration) The crypto asset valuation system shown in Figure 1, as one embodiment of the present invention, comprises an input unit 11, an output unit 12, and a model calculation unit 20.

[0009] The input unit 11 receives parameters related to the "market environment," "agent characteristics," "simulation settings," and "output parameters," which form the basis for evaluating crypto assets (e.g., tokens). Table 1 summarizes the classification of the input parameters ("market environment," "agent characteristics," "simulation settings," and "output parameters") and how to set the values ​​of each input parameter.

[0010]

[0011] ​The model calculation unit 20 is composed of a calculation processing unit and a storage device, etc. Data related to parameters input through the input unit 11 is stored in the storage device. The calculation processing unit reads the necessary data and crypto asset valuation algorithm (software) from the storage device and performs calculation processing for valuing crypto assets such as tokens according to the algorithm. This crypto asset valuation algorithm focuses on utility value and is an extension of the probabilistic model of Non-Patent Document 1 by a unique agent-based algorithm.

[0012] The output unit 12 outputs the calculation results from the model calculation unit 20 in a form that can be recognized by the user through sight, hearing, touch, etc. The output unit 12 is composed of, for example, an image output device and / or an audio output device. The output data includes the total utility value of the tokens, the utility value per unit token, the total governance value of the tokens, and the governance value per token at the end of the simulation.

[0013] (Function) When simulating in a virtual world, the model is first initialized (Figure 2 / STEP 10).

[0014] Specifically, agents are created and defined in the virtual world (Figure 3 / STEP 102). Specifically, a number of agents corresponding to the specified initial number of users (N0) are imagined, and the parameters (a and u) and duration w of the utility function held by each agent are defined. Furthermore, the initial amount k of tokens held by each agent i in the virtual world is defined. i,0 This is defined (Figure 3 / STEP 104).

[0015] Market growth potential and platform (PF) productivity in the virtual world A t The next PF productivity A in the virtual world is updated (Figure 2 / STEP 11). Specifically, it is updated according to geometric Brownian motion with the initial value being the initial PF productivity A0. t+1 Currently, PF productivity A t productivity drift coefficient μ p , the diffusion coefficient of productivity σ p, based on the length of the period \(T\) (e.g., the number of days in a year (\(= 365\))) and the standard normal distribution \(Z\sim N(0, 1)\), it is defined according to the relational expression (01) (see Non-Patent Document 1).

[0016] A t+1 = A t exp{(\(\mu\) p −(\(\sigma\) p )) 2 / 2) / T+\(\sigma\) p ×Z\(\sim\)N(0, 1)×T 1 / 2}‥(01).

[0017] In the virtual world, the market size expands according to a predetermined schedule. A predetermined number of agents are created in the environment.

[0018] Subsequently, tokens are supplied to agent \(i\) in the virtual world (Figure 2 / STEP 12). Specifically, in the virtual world, tokens are supplied to the corresponding agent \(i\) according to the following two forms.

[0019] In the first token supply method, tokens are supplied to token holders, including participants or stakers, in the virtual world. Specifically, a token pool is first created in the virtual world (Figure 4 / STEP 212). The pool size is calculated by multiplying the predetermined new supply amount for each period by the proportion allocated to the first token supply method. Next, agents who meet the reward receiving conditions are selected in the virtual world (Figure 4 / STEP 214). These conditions include having a token holding amount greater than 0 and a staked token holding amount greater than 0. Furthermore, tokens are awarded to the selected agents from the token pool in the virtual world (Figure 4 / STEP 216). Rewards are randomly (uniformly distributed) selected from agents who meet the above conditions, and one token is awarded to the selected agent. Subsequently, it is determined whether or not there are tokens remaining in the token pool in the virtual world (Figure 4 / STEP 218). If the result of the determination is positive (Figure 4 / STEP 218...YES), the process from agent selection (Figure 4 / STEP 214) onwards is repeated in the virtual world. If the result of the determination is negative (Figure 4 / STEP 218...NO), the token supply process in the virtual world is stopped.

[0020] In the second token supply method, tokens are supplied by the exchange. Specifically, it is first determined whether or not an IEO (Initial Exchange Offering) has taken place in the virtual world (Figure 5 / STEP 222). If the result of this determination is negative (Figure 5 / STEP 222...NO), the token supply process is not executed in the virtual world. On the other hand, if the result of this determination is positive (Figure 5 / STEP 222...YES), a token pool is created in the virtual world (Figure 5 / STEP 224). Next, tokens are supplied to the exchange from the token pool in the virtual world (Figure 5 / STEP 226). Then, the exchange sells the tokens to the agent in the virtual world (Figure 5 / STEP 228).

[0021] Next, agent i stakes the token in the virtual world (Figure 2 / STEP 13). Specifically, agent i's token holdings k in the virtual world i,t The optimal token holding amount is k i,t * The following is determined (Figure 6 / STEP 32): Optimal token holdings k at time t. t * The token price P at time t is the token price P. t , Number of users N at time t t Platform productivity A at time t t , risk-free rate r, marginal utility diminishing parameter α of the utility function (0 < α < 1), drift coefficient μ of platform productivity p , and the magnitude of utility relative to the value provided by the platform e ｕi Based on this, it is defined according to relational equation (02) (see equation 28 in Non-Patent Document 1).

[0022] P t k t * = N t A t e ｕi ((1-α) / (r-μ p )) 1 / α ... (02).

[0023] (10 / 24) It seems that the changes regarding the r - μ^p part have not been communicated, so I will add them here. Please change the r - μ^p part to r - (G_t / M_t*P_t) - μ^p. This is an extension of equation 28 in Non-Patent Document 1 by optimizing equation (22) in this document, so please add an explanation as necessary. However, G_t is the governance value of the platform at time t, M_t is the token supply at time t, and P_t is the token price at time t.

[0024] If the result of the assessment is positive (Figure 6 / STEP 32...YES), the token will be staked in the virtual world (Figure 6 / STEP 34). On the other hand, if the result of the assessment is negative (Figure 6 / STEP 32...NO), the token will not be staked in the virtual world (Figure 6 / STEP 36).

[0025] Next, limit orders are placed by agents in the virtual world (Figure 2 / STEP 14). Agent i includes a "rational agent".

[0026] In the virtual world, limit orders by rational agents (parameters defining the trading characteristics of crypto assets) are defined as shown in Table 2.

[0027]

[0028] The token holding evaluation function f is defined according to relational equations (21) and (22) (see equations 2 and 15 in Non-Patent Literature 1). The function f defined by equation (21) takes an arbitrary token quantity k, an evaluator agent i, and an evaluation time t as inputs, and outputs the evaluation of the input token quantity by agent i at time t.

[0029] According to relation (21), in the virtual world, the utility dν at time t obtained from the services provided by the platform for agent i is i,t , the token price P at time t t , the amount of tokens held at time t k t , Number of users N at time t t Platform productivity A at time t t , risk-free rate r, marginal utility diminishing parameter α of the utility function (0 < α < 1), drift coefficient μ of platform productivity p , and the magnitude of utility relative to the value provided by the platform e ｕi (u i The correlation between the agent type and the platform participation cost (opportunity cost) φ is defined.

[0030] f(k,i,t)=dν i,t= (P t k t ) 1-α (N t A t e ｕi ) α dt-φdt-P t k t rdt+P * k * u^pdt+k * (G / M)dt (21).

[0031] According to relation (22), in the virtual world, with respect to agent i, the total payoff and the utility dν obtained from the services provided by the platform are... i,t , and profit P from token price fluctuations t k t (E[dP t ] / P t The correlation between these two is defined.

[0032] total payoff=dν i,t +P t k t (E[dP t ] / P t ) ... (22).

[0033] Furthermore, the token's price is calculated in the virtual world before the token is traded (Figure 2 / STEP 15).

[0034] When calculating the price of a token in the virtual world, it is determined whether or not there are buy orders for the token (Figure 7 / STEP 510). Specifically, this is determined by whether or not the record of limit buy orders for each agent for the current period is empty. If it is determined that there are no buy orders for the token in the virtual world (Figure 7 / STEP 510...NO), it is determined whether or not there are sell orders for the token in the virtual world (Figure 7 / STEP 511). Specifically, this is determined by whether or not the record of limit sell orders for each agent for the current period is empty. If it is determined that there are no sell orders for the token in the virtual world (Figure 7 / STEP 511...NO), the final trading price of the token at the end of the previous period is determined as the price in the virtual world (Figure 7 / STEP 516). On the other hand, if it is determined that there are sell orders for the token in the virtual world (Figure 7 / STEP 511...YES), the reserved price of the token is determined as the price in the virtual world (Figure 7 / STEP 517).

[0035] If it is determined that there are buy orders for tokens in the virtual world (Figure 7 / STEP 510...YES), it is determined whether there are sell orders for tokens in the virtual world (Figure 7 / STEP 512). If it is determined that there are no sell orders for tokens in the virtual world (Figure 7 / STEP 512...NO), the holding price of the tokens in the virtual world is determined as the price (Figure 7 / STEP 518).

[0036] On the other hand, if it is determined that there are sell orders for tokens in the virtual world (Figure 7 / STEP 512...YES), it is determined whether there is an intersection between the demand and supply curves of the tokens in the virtual world (Figure 7 / STEP 513). If it is determined that there is no intersection (Figure 7 / STEP 513...NO), it is determined whether there are buy orders at a price higher than the sell orders (Figure 7 / STEP 514). If it is determined that there are no buy orders in the virtual world (Figure 7 / STEP 514...NO), the final trading price of the tokens at the previous period in the virtual world is determined as the price (Figure 7 / STEP 516). If it is determined that there are buy orders in the virtual world (Figure 7 / STEP 514...YES), the lowest buy order price at which a transaction was completed in the virtual world is determined as the price (Figure 7 / STEP 517). If it is determined that there is an intersection (Figure 7 / STEP 513...YES), the price at that intersection is determined as the price (Figure 7 / STEP 519).

[0037] When trading tokens in the virtual world, it is determined whether there are any matching or suitable buy and sell orders (Figure 8 / STEP 522). Specifically, it is determined whether there are any buy orders at a price higher than a given sell order. If it is determined that there are no matching orders in the virtual world (Figure 8 / STEP 522...NO), the tokens are not traded in the virtual world. If it is determined that there are matching orders in the virtual world (Figure 8 / STEP 522...YES), agents who placed sell orders below the price in the virtual world release their tokens (Figure 8 / STEP 524). Specifically, the supply of unheld tokens increases according to the number of sell orders placed by the agent, and the supply of tokens held by the agent decreases. Also, the amount of tokens held by the agent decreases according to the number of sell orders. Furthermore, agents who placed buy orders at or above the price in the virtual world acquire tokens (Figure 8 / STEP 526). Specifically, the supply of unheld tokens decreases according to the number of buy orders placed by the agent, and the supply of tokens held by the agent increases. Furthermore, the token holdings of the agent will increase in proportion to the number of buy orders.

[0038] Next, the token is staked by agent i in the virtual world (Figure 2 / STEP 16).

[0039] Furthermore, it is determined whether or not another trading round will be conducted in the virtual world (Figure 2 / STEP 17). If the result of this determination is positive (Figure 2 / STEP 17...YES), the process from Agent i's limit order (Figure 2 / STEP 14) onwards is repeated in the virtual world. On the other hand, if the result of this determination is negative (Figure 2 / STEP 17...NO), the token is unlocked in the virtual world (Figure 2 / STEP 18). Unlocking makes it possible to trade on the market.

[0040] When unlocking tokens in the virtual world, it is determined whether the token has been held for a certain period of time or longer (Figure 9 / STEP 82). Specifically, each token held by the agent has information about its holding period, which is compared with the token lock period (T1) to determine if it is unlocked. Unlocking is performed only for tokens whose holding period is equal to or longer than the token lock period (T1). Each token also has information about whether it is locked or not, and locking and unlocking are performed by updating this information. If the result of this determination is positive (Figure 9 / STEP 82...YES), the token will be changed to a tradable state in the virtual world from the next period (Figure 9 / STEP 84). If the result of this determination is negative (Figure 9 / STEP 82...NO), the token will remain tradable in the virtual world in the next period as well (Figure 9 / STEP 86).

[0041] Next, the token is staked by agent i in the virtual world (Figure 2 / STEP 19).

[0042] Then, it is determined whether or not the simulation will be run again (Figure 2 / STEP 20). If the result of this determination is positive (Figure 2 / STEP 20...YES), the process from updating market growth potential and platform productivity in the virtual world (Figure 2 / STEP 11) onwards is repeated. On the other hand, if the result of this determination is negative (Figure 2 / STEP 20...NO), the series of processes or simulations ends.

[0043] The model calculation unit 20 calculates the output results at the end of the simulation. The growth token value is calculated at its present value (at the start of the simulation), and the model terminates after n periods.

[0044] The total utility value TV at the end or current time is equal to the token supply M at the end. end , token price P at the end end , discount rate r d It is defined according to relation (41), based on the number of reference points (e.g., 0:00 each day) in a specified period (t) and the number of reference points per period T (t ≤ T (e.g., T = 365 if one period is one year)). The token price P at the end end This is the final transaction price obtained by the processing shown in Figure 7 when entering the END block in Figure 2. The token supply at the end is M. end This represents the amount of tokens supplied up to the point when the END block in Figure 2 is reached, as shown by the processing in Figures 4 and 5. CF and the discount rate r are not simulation results, but are based on the assumptions of the model user / defender. Since the final transaction price is determined probabilistically, it is assumed that the model will be run multiple times with similar parameters.

[0045] TV=P end ×M end / (1+r) d ) t / T ... (41).

[0046] The utility value per token tv at the end of the term or at the present time is calculated as the total utility value tv and the token supply M at the end of the term. end Based on this, it is defined according to relation (42).

[0047] tv = TV / M end ‥(42).

[0048] Total governance value GV at the end end * is defined according to relational expression (43) based on the cash flow CF m at time t (t = t1 to t m ≤ T), discount rate r m and terminal value V d tmn end *

[0049] GV t=t1～tm t / T = Σ tmn {CFi / (1 + r) tmn} + V t ‥(43).

[0050] Terminal value V d is defined according to relational expression (45) based on the cash flow CF tmn at time t, discount rate r t and specified period T

[0051] V t / T = CF t / T × (1 + r) end / {(1 + r) end - 1} ‥(45). <00002​​​​​​​​​​​​​​​The governance value (gv) per token is also displayed. The display may also show trends in the number of token holders, token trading volume, market size, token supply, etc., as shown in Figures 11 to 14, for example.

[0055] As shown in Figure 15, the token supply schedule and token supply method are displayed on the touch-panel displays of the input unit 11 (input interface) and output unit 12 (output interface) in a form that can be adjusted using touch-panel sliders or the like that make up the input unit 11, for confirmation before the simulation starts. Furthermore, as shown in Figure 16, the market size growth schedule is displayed on the touch-panel displays of the input unit 11 in a form that can be adjusted using touch-panel sliders or the like that make up the input unit 11, for confirmation before the simulation starts. In addition, as shown in Figure 17, the parameters including token characteristics, platform characteristics on which tokens circulate, parameters of the platform's productivity stochastic process, and agent parameters are displayed on the touch-panel displays of the input unit 11 in a form that can be adjusted using touch-panel sliders or the like that make up the input unit 11, for confirmation before the simulation starts.

[0056] In this way, by adjusting the parameters that define the asset characteristics of the target cryptocurrency and the parameters of the agent's trading tendencies, it is possible to define a simulation model that takes into account diverse cryptocurrency and market characteristics and treats them as complexities on an agent-based basis.

[0057] (Effects) The crypto asset valuation system with the above configuration can incorporate different crypto asset characteristics (supply volume, supply method, service value, market size, productivity, etc.) for each platform. By using agent-based simulation, these elements can be considered dynamically as a complex system. As a result, the accuracy of crypto asset valuation is improved.

[0058] When crypto assets are supplied by an exchange, they are sold from the exchange to the agent via market orders, as described above. Here, if the total number of buy orders by the agent is less than the number of market sell orders by the exchange, all sell orders by the agent may be discarded. This reduces the computational burden when determining the price of the crypto asset. Specifically, the load of order sorting processing required when determining the price (when searching for an intersection, sorting of prices is performed) is reduced by making the number of orders in the order list smaller. This is made possible because the agent's orders are calculated before the exchange's orders.

[0059] Because the load of order reordering during price calculation is reduced, the unit of cryptocurrency handled in the simulation can be adjusted to any value, which reduces the computational load even when the supply of cryptocurrency in reality is enormous.

[0060] Assuming that each agent has an equal influence on the price, the number of sell orders or buy orders from each agent is set to one unit. This reduces the burden of order reordering during price calculation, even when the number of agents (market size) is enormous.

[0061] Order reordering is only necessary when it is necessary to calculate the transaction price (see Figure 7 / STEP 513, STEP 514). Therefore, the judgment processes in Figure 7 / STEP 510, STEP 511, and STEP 512 are inserted, and unnecessary order reordering is omitted.

[0062] The crypto asset valuation system according to the present invention contributes to the industry by using a computer to evaluate the value of crypto assets according to a model.

[0063] 11...Input section 12...Output section 20...Model calculation section

Claims

1. A cryptocurrency valuation system comprising: an input unit that receives model parameters including parameters that define the trading characteristics of a cryptocurrency on a platform; a model calculation unit that determines information regarding the future value of a cryptocurrency according to a valuation model that defines the trading trends of the cryptocurrency on the platform based on the model parameters input to the input unit; and an output unit that outputs information regarding the future value of the cryptocurrency determined by the model calculation unit.

2. A crypto asset valuation system according to claim 1, wherein the valuation model is a model defined such that the productivity of the platform transitions probabilistically according to geometric Brownian motion.

3. A crypto asset valuation system according to claim 1, wherein the valuation model is a model in which the transaction characteristics of the crypto asset are defined such that, according to a first crypto asset supply form, agents who satisfy the reward receiving conditions are randomly selected and the crypto asset is awarded to the selected agents.

4. A crypto asset valuation system according to claim 1, wherein the valuation model is a model in which the trading characteristics of the crypto asset are defined such that an exchange sells the crypto asset to an agent at market price according to a second crypto asset supply model.

5. A crypto asset valuation system according to claim 1, wherein the valuation model is a model in which the trading characteristics of the crypto asset are defined such that the crypto asset is staked when the agent's holdings of the crypto asset are less than or equal to the optimal holding amount, while the crypto asset is not staked when the agent's holdings of the crypto asset exceed the optimal holding amount.

6. A crypto asset valuation system according to claim 1, wherein the valuation model is a model in which the trading characteristics of the crypto asset are defined such that a limit order by a reasonable agent places a buy order if the optimal amount of the crypto asset held is less than the current amount of the crypto asset held and the expected reward for the crypto asset, but does not place a buy order otherwise, and places a sell order if the optimal amount of the crypto asset held is greater than the current amount of the crypto asset held plus the expected reward for the crypto asset, but does not place a sell order otherwise.

7. A crypto asset valuation system according to claim 1, wherein the valuation model is a model in which the trading characteristics of the crypto asset are defined such that unlocking is performed only for the crypto asset whose holding period is equal to or greater than the crypto asset lock period.

Images