Methodological approaches to pricing and quantitative risk assessment of digital financial assets in Russia

Galyaev Pavel Aleksandrovich Postgraduate Student; Department of Financial Markets and Financial Institutions; Kazan (Volga Region) Federal University 420008, Russia, Rep. Tatarstan, Kazan, Vakhitovsky district, Kremlevskaya str., 18, room 1 galyaev1990@gmail.com Other publications by this author

Introduction

By the beginning of 2026, the Russian digital financial assets (CFA) market had undergone a fundamental transformation, moving from the experimental pilot phase to the stage of a mature segment of the debt market. According to financial information aggregators, the total volume of CFA issues by the end of 2025 exceeded the landmark mark of 1 trillion rubles ^[1]. This milestone marks the end of the "exotic" stage: digital rights have become a systemically important tool for attracting liquidity not only for fintech companies, but also for the largest corporate borrowers from the real economy and state corporations.

With the Bank of Russia's tight monetary policy remaining in 2024-2025, CFAs have de facto occupied the niche of short-term money market instruments, offering investors returns exceeding the rates on classic deposits and exchange-traded bonds. However, the scaling of the market and the entry of institutional investors (insurance companies, non-governmental pension funds) into it have actualized the problem of fair valuation of these assets, which is reflected in the research of L. P. Kharchenko and O. N. Korableva ^[2].

The key scientific and practical problem at the current stage is the lack of generally accepted methodological approaches to pricing digital rights. In the current practice described in the works of D. A. Kochergin ^[3], T. Koller, M. Goedhart and D. Wessels ^[4], CFAs are often considered as a complete analogue of traditional bonds, and their valuation is carried out using the discounted cash flow method (DCF) using the coupon-free yield curve (G-curve) or corporate bond spreads ^[5].

However, this approach seems simplistic and does not take into account the institutional specifics of digital rights. Unlike exchange-traded bonds traded in a single liquid circuit (the Moscow Stock Exchange), the CFA market remains fragmented: assets issued in one information system often do not have liquidity in another. This creates a specific "premium for illiquidity", which the investor demands for the risk of the inability to quickly realize the asset ("risk of platform isolation") ^[6]. In addition, the technological architecture of the market, based on a distributed registry ^[7], introduces a new parameter into the pricing model – the "technological discount", reflecting the operational risks of a particular information system operator ^[8].

Thus, the mechanical transfer of valuation models from the securities market to the CFA market leads to a distortion of their fair value and underestimation of risks, which prevents the formation of an effective secondary market ^[9].

Table 1 shows that the key interest rate and credit risk factors for bonds, while the CFA adds platform risk and liquidity fragmentation.

Table 1 – Comparative characteristics of corporate bond and CFA pricing factors

Source: compiled by the author.

As follows from Table 1, the nature of CFA pricing is dualistic. On the one hand, the fundamental value of digital debt, as in the case of bonds, is based on a risk-free rate and the credit quality of the borrower. However, unlike the classical stock market, CFA investors require compensation for the specific risks inherent in a decentralized infrastructure.

Empirical data from 2025-2026 show that the "illiquidity premium" and "platform risk" form a positive CFA yield spread relative to the bonds of the same issuer. This allows us to formalize the CFA fair return model (Y _DFA) as follows:

where:

- R _f – risk-free rate;

- RP _cr– credit risk premium (similar to bonds);

- RP _lig – premium for liquidity fragmentation (specific to CFA);

- RP _tech – premium for technological/infrastructural risk of IPOs.

It is the presence of componentsRP _lig and RP _tech explain why issuers are forced to offer higher CFA rates, despite lower transaction costs when issuing.

The working hypothesis of the study is that digital financial assets have a pricing structure different from classical debt instruments. A fair CFA return should include, in addition to the risk-free rate and premium for the issuer's credit risk, a specific premium for the information system operator's infrastructural risk and limited secondary liquidity.

The purpose of this work is to identify specific factors influencing the formation of the CFA's market value in the context of the Russian financial infrastructure in 2026, as well as to develop a classification of risks of investing in digital rights for qualified and unskilled investors.

The scientific novelty of the study is to substantiate the multicomponent structure of the required CFA profitability, which includes specific premiums for liquidity fragmentation and the technological risk of the platform. The author has developed an original risk matrix for investing in digital rights, which allows decomposing uncertainty into systematic and specific (infrastructural) factors. For the first time, based on an empirical analysis of data from 2024-2025, a stable positive CFA yield spread was identified and quantified in relation to classic bonds of comparable duration (+0.8–1.5 percentage points) due to the "platform closure" effect.

The theoretical and methodological basis of the work is based on the fundamental principles of modern financial market theory, asset valuation concepts and risk management methods. The main research tools used were methods of comparative institutional analysis, a modified discounted cash flow model (DCF), as well as an empirical method for analyzing yield spreads based on data from Russian information system operators and the Cbonds aggregator.

Theoretical foundations of CFA pricing

The pricing of tokenized assets is widely covered in modern foreign literature. The fundamental works of L. W. Cong and Z. He ^[10] describe the mechanisms of formation of the equilibrium price of tokens in decentralized ecosystems, and the research of R. Moro-Visconti ^[11] focuses on the mathematical modeling of the "illiquidity premium" in liquidity pools (DeFi Liquidity Pools).

However, it is important to note a significant institutional difference between the objects of research. In foreign works (K. Qin, L. Zhou, L. Lazzaretti ^[12] and others), the main focus is on the risks of smart contracts in public blockchains (for example, Ethereum), where there is no central administrator, and the risks are mainly algorithmic in nature.

The scientific novelty of this study lies in the adaptation of risk assessment approaches for private blockchains, on which the Russian model of information system Operators is based. In the domestic circuit, the key pricing factor is not the risk of code or the volatility of network fees, but the risk of a particular operator and regulatory fragmentation of liquidity, which requires a revision of the classical models proposed by foreign colleagues.

From an economic point of view, the vast majority of digital financial assets issued on the Russian market in the period 2022-2025 are inherently debt obligations (monetary claims). This suggests an institutional isomorphism between CFAs and classic corporate bonds, as noted in the study by V. S. Stankevich and A.V. Vlasov ^[13].

The basic approach to estimating the value of such assets has traditionally been the discounted cash flow method. The fair value of the CFA (P _DFA) under this paradigm is defined as the present value of future payments.:

where:

- CF _t – cash flow in period t (coupon payments and nominal repayment);

- r is the discount rate (required yield);

- n – maturity date.

However, despite its versatility, the application of the DCF model in its pure form to the CFA market is fraught with a number of methodological limitations. According to researchers D. Mikael-Ismail and O. B. Digilina ^[14], the key problem lies in determining the correct discount rate of r.

For the bond market, this rate is based on the coupon-free yield curve (G-curve) The Moscow Stock Exchange, which has high liquidity and information efficiency. The CFA market, on the contrary, is characterized by segmentation: the profitability of assets of the same credit quality can vary significantly on the platforms of different information system operators due to the lack of arbitrage.

Consequently, the basic DCF model ^[15] requires modification by introducing a premium for specific risks of the digital environment that are not taken into account in the valuation models of traditional securities.

Based on historical trading data and an analysis of pricing practices on Russian platforms (Atomize, Sber, Alfa-Bank, and Lighthouse), it seems possible to empirically identify four fundamental factors that determine the CFA market price. Efimtseva ^{[16, p. 61]}).

First, the credit quality of the issuer.

This factor is dominant for the CFA on monetary demand. By analogy with bonds, the CFA yield spread to the risk-free rate (the key rate of the Central Bank of the Russian Federation or RUONIA) correlates with the issuer's credit rating.

In 2024-2025, the market developed the practice of assigning credit ratings to CFA issues by leading agencies (ACRA, Expert RA), which reduced information asymmetry. However, for issuers of the SME segment who do not have a public credit history, the premium for credit risk in CFA quotes is higher than in the VDO (high-yield bonds) market, which is explained by the novelty of the instrument for conservative investors ^[17].

Secondly, the urgency of the instrument.

The specific feature of the Russian CFA market is the predominance of short-term instruments. According to the research data of A. I. Goncharov and M. V. Goncharova, the average duration of issues was less than 1 year (3-9 months) ^[18].

This has a direct impact on pricing.:

1. The price of short CFAs is less volatile when the key rate changes than the price of long OFZs.

2. The cost of funding through the CFA is rigidly linked to the current value of money, and not to long-term inflationary expectations.

Thirdly, liquidity ^[19].

The liquidity factor is the most critical difference between CFAs and bonds. On a classical exchange, liquidity is consolidated, which ensures a minimum spread between the purchase and sale prices. In the CFA market, liquidity is fragmented by isolated information systems ("wells of liquidity") ^[20].

In the CFA secondary market (where it is technically implemented), the spread of supply and demand can reach 0.5–1.5%, which is a barrier to speculative strategies. This should be included in the valuation model as an "illiquidity premium" (RP _lig), which is deducted from the fair price of the asset ^[21].

Figure 1 – Dependence of the CFA yield spread on the issue volume and platform liquidity (model)

Source: compiled by the author on the basis of Cbonds data, an expert of the Republic of Armenia and the Bank of Russia ^[22].

As can be seen from the graph, the yield curve for platforms with an implemented secondary circulation mechanism (such as Sber-CFA or A-Token, where more than 1 million retail investors are connected) is significantly lower than the curve of niche operators, which clearly demonstrates the presence of an "infrastructure illiquidity premium"

With the same credit quality of the issuer (conditional ruAA rating), investors on platforms without a secondary market require an additional premium of 125-160 basis points. This is a payment for the risk of being "stuck" in an asset (lock-in risk) until maturity.

As the output volume increases (on the right along the abscissa axis), the spread decreases for both categories of platforms, but the gap persists. Even large issues on illiquid platforms are more expensive for the issuer than similar loans in ecosystems with a developed secondary market.

This confirms the hypothesis that the technological architecture of the OIC is a pricing factor: the issuer can reduce the cost of borrowing by 1.2–1.6% per annum by simply choosing a platform with a more active secondary bidding.

Fourth, the underlying asset (for hybrid and structural CFAs).

A special category consists of digital assets, the value of which is linked to real assets or indexes ^[23]. In this case, pricing is based not on a debt model, but on a model for evaluating structural products or options.

The fair price of a hybrid CFP (P _Hybrid) can be described by the formula:

where:

- P _Base – the current market price of the underlying asset (for example, 1 gram of gold or 1 sq.m. of real estate);

- k is the participation rate (or the amount of rights);

- Opt _val – the value of the embedded option (for example, the investor's right to choose the form of repayment: in cash or in metal).

An example is "digital square meters" (GK "Airplane", "G-group"), where the CFA price correlates with the housing cost index, but is discounted for the risks of completion (development risk), which distinguishes them from direct investment in real estate ^[24].

Thus, the theoretical pricing model of the CFA is multifactorial and requires taking into account specific infrastructural risks that are not typical for the classical stock market.

Empirical analysis of profitability

The period 2024-2025 was a stress test for the Russian debt market, characterized by an unprecedentedly tight monetary policy of the Bank of Russia. The dynamics of the key rate shown in Figure 2 has determined the structural transformation of the borrowing market (data obtained from the official website of the Bank of Russia section "Key Rate of the Bank of Russia").

Figure 2 – Dynamics of the Bank of Russia's key rate (2024-2025)

Source: compiled by the author based on data from the Bank of Russia.

To verify the hypothesis of a specific risk premium in the value of digital assets, a comparative analysis of the yield to maturity (YTM) of instruments from the same issuer was conducted.

The sample included first-tier issuers ("Blue Chips") who have both classic exchange-traded bonds and digital financial assets in circulation. The comparison was carried out on the date of December 15, 2025 (the start period of the PREP mitigation cycle, the key rate of 16.5%).

The results of the paired comparison are presented in Table 2.

Table 2 – Yield spread between corporate bonds and CFA as of 12/15/2025 (key rate 16.5%)

Source: compiled by the author on the basis of data from the Bank of Russia, research by M. V. Yurgelas and O. A. Avdeevich ^[25], an analysis of JSC Investment Company Rikom-Trust ^[26], as well as discussions outlined in the monograph edited by I. M. Stepnov and Yu. A. Kovalchuk ^[27].

The analysis of the data in Table 2 allows us to draw the following conclusions characterizing the state of the market at the end of 2025:

1. Availability of a stable premium. In all the pairs considered, the profitability of digital financial assets exceeds the profitability of classic bonds of the same issuer with a comparable duration. The spread ranges from 0.8 to 1.5 percentage points. This confirms the hypothesis that the market evaluates CFAs as a riskier instrument compared to exchange-traded bonds.

2. The structure of the award. The minimum spread (+0.8 percentage points) is observed among bank issuers (VTB) issuing CFAs on their own or partner platforms with high integration of the customer base. The maximum spread (+1.3...1.5 percentage points) is typical for corporate borrowers (MTS, Aeroplane).

This difference is explained by the liquidity factor: an investor can sell an exchange-traded MTS bond at any time in the T+1 mode with minimal losses. The sale of CFA MTS on the secondary market of the Atomize platform may require time or a significant discount to the price. It is this loss of liquidity that the investor puts into the required profitability ^{[28, p. 26]}.

3. Arbitrage on the betting curve. Based on the empirically proven hypothesis of an infrastructure premium, for issuers, the presence of a 1.5% spread means that CFAs are a more expensive source of funding. However, the popularity of the instrument in 2024-2025 is explained by the emission rate. During periods of sharp rate increases (September–October 2024, June 2025), the organization of bond issuance took 1-2 months, during which the market situation could deteriorate. The CFA was issued in 3-5 days, which made it possible to fix the conditions "here and now", compensating for the increased rate with the speed of obtaining liquidity.

Thus, by 2026, the Russian capital market has developed a two-tier betting system: a "premium" exchange level (low rates, high liquidity) and an "alternative" digital level (increased rates, limited liquidity), which acts as a buffer during periods of macroeconomic instability.

The premium to CFA profitability (0.8–1.5 percentage points) identified in the framework of empirical analysis cannot be exhaustively explained by classical market factors. This requires a systematic transition from the identification of price deviations to an in-depth analysis of the specific risk architecture of digital financial assets. The author's classification presented below is intended to decompose this premium by linking the cost of capital to technological and infrastructural constraints of the market.

Classification of risks of investing in CFA

As the mathematical modeling of the structure of investor losses shows, traditional risk management models (VAR, Stress Testing) used in the stock market demonstrate limited applicability to the segment of digital financial assets ^[29]. This is due to the fact that the CFA architecture introduces fundamentally new variables into the investment equation related to distributed ledger technology and the specifics of information system operators.

As part of this study, we propose an author's risk matrix for investing in CFAs (Table 3), based on the division of vulnerability factors into two groups: systematic (general market) and specific (infrastructural and technological) ^[30].

Table 3 – Matrix of risks of investing in digital financial assets

Source: developed by the author.

The qualitative matrix proposed in Table 3 requires quantitative interpretation for practical application in pricing models. To calculate the specific premium to the CFA profitability, we propose using the integral risk scoring model (R _total).

This indicator aggregates the influence of specific factors of the digital environment and is calculated using the formula:

where:

- Pi ₍Probability) – the probability of realizing the ith specific risk (estimated on a scale from 0 to 1 based on historical data on system failures or liquidity statistics for 2024-2025);

- I _i (Impact) – the degree of impact of the event on the investor's capital (normalized scale from 1 to 10, where 10 is a complete loss of liquidity/asset);

- W _i (Weight) – the weight of the risk factor in the overall structure (determined by expert analysis; for 2026, the dominant weight is the risk of liquidity fragmentation – 0.5).

The resulting R _total value is converted into basis points (bp) of the premium to the RP _spec discount rate through the market sensitivity coefficient (β _mkt), which we empirically determined at the level of 20-25 bp per unit of risk.:

Calculation example.

To demonstrate how the model works, we will conduct a comparative analysis of two hypothetical CFA issues of the same credit quality (ruAA) hosted on different platforms.:

Platform "A" (type: Niche) – low liquidity, lack of integration with banks.

Platform "B" (type: Ecosystem) – high liquidity, availability of a secondary market (for example, Sberbank/Alfa).

Calculation of integral risk (R _total).

1. The risk of liquidity fragmentation (W = 0.5).

Platform "A": P = 0.8 (high probability of "getting stuck"), I = 8. Contribution = 0.8 × 8 × 0.5 = 3.2

Platform "B": P = 0.2 (low probability), I = 4. Contribution = 0.2 × 4 × 0.5 = 0.4

2. Infrastructure risk of the operator (W = 0.3).

Platform "A": P = 0.3, I = 10. Contribution = 0.3 × 10 × 0.3 = 0.9

Platform "B": P = 0.05, I = 10. Contribution = 0.05 × 10 × 0.3 = 0.15

3. Total (R _total) and Bonus.

Platform "A": R _total ≈ 4.1. Estimated premium: 4.1 × 25 bp ≈ 102 bp (+1.02% of the bid).

Platform "B": R _total ≈ 0.55. Estimated premium: 0.55 × 25 bp ≈ 14 bp (+0.14% of the rate).

Mathematical calculations confirm that, all other things being equal, an investor on the "niche" Platform "A" fairly requires a premium to profitability of about 1 percentage point (the difference between 102 and 14 bp) solely due to a higher integral indicator of infrastructure risk. This correlates with the empirical spread data presented earlier in Table 2.

This classification allows investors (especially institutional ones – NPFs and insurance companies) to correctly calculate the risk premium when forming a portfolio.

If the systematic risks (interest rate and regulatory) ^[31] are understandable to market participants and are hedged by standard instruments, then specific risks require detailed consideration, since they form a "technological discount" in the CFA price.

The risk of liquidity fragmentation is a fundamental problem of the Russian CFA market as of 2026. Unlike the stock market, where an asset (for example, a Gazprom share) is traded in a single pool of liquidity on the Moscow Stock Exchange, the CFA market is a collection of isolated ecosystems ("wells").

An investor who has bought a CFA on the Atomize platform technically cannot sell it to a customer of the Sber-CFA platform without the participation of an exchange operator, an institution that is in its infancy ^[32]. This creates the risk of "locking up" capital: in the event of a crisis, the investor is forced to sell the asset at a discount of 10-15% to the only market maker inside the platform, since access to the external market is closed.

Technological risk implies that digital assets are programmable code. For hybrid CFAs (for example, for a basket of metals or square meters), the correct operation of "Oracles" is critically important – external data sources that transmit the price of the underlying asset to the blockchain, which, in particular, is confirmed by the research conducted by S. A. Andryushin ^[33].

For example, if the Oracle transmits an erroneous gold price due to a glitch on the London Stock Exchange or a technical API error, the smart contract will automatically redeem at a non-market price. Unlike traditional transactions that can be challenged in court and canceled, transactions in a distributed registry are often irreversible or require a complex fork procedure, which creates legal uncertainty ^[34].

Operator's risk – by investing in bonds, the investor assumes the risk of the issuer and (minimally) the risk of the National Settlement Depository (NSD). By investing in the CFA, the investor assumes the direct risk of the IPO ^[35]. Despite the supervision of the Bank of Russia, the financial stability of niche fintech platforms is lower than that of systemically important banks. The bankruptcy of an IPO or the revocation of its license may lead to a temporary "freeze" of assets and the inability to exercise claims against the issuer, even if the issuer itself is solvent.

Thus, the proposed author's model proves that the CFA's risk profile is more complex and multi-faceted compared to classic debt securities. A fair assessment of profitability should include a premium for the specific risks described above, the amount of which correlates with the reliability of a particular platform.

Risk management methods

Closing the outline of a systematic study, let's move from identifying and quantifying risk factors to developing mechanisms for their neutralization. The transformation of the CFA market from the venture segment into a systemic element of the financial infrastructure is impossible without the introduction of comprehensive risk management. If risk management was reactive at the stage of market formation (2022-2023), then by 2026 three preventive areas of investor protection had been formed that could stop the threats described above and reduce the "technological discount".

The traditional strategy of diversifying the portfolio by issuers in the CFA market is necessary, but insufficient. As it was shown earlier, the investor assumes the unavoidable infrastructural risk of the Information system Operator. In the context of 2026, when the Bank of Russia's register includes more than 20 IPOs with varying degrees of technological maturity, platform-based diversification is becoming a critical method of minimizing risks.

Holding a significant amount of assets within a single IP creates the risk of an "infrastructure trap": a technical failure, a cyberattack, or regulatory sanctions against a particular operator can lead to a temporary loss of liquidity for the entire portfolio, even while maintaining the solvency of issuers ^[36].

The author believes that for institutional investors, the concentration limit per IPO should not exceed 15-20% of the total digital portfolio. The use of aggregators – Exchange Operators – allows you to automate this process by distributing assets between independent nodes of the distributed registry of different platforms.

To quantify the effectiveness of the proposed limitation, we applied a scenario–based approach to modeling losses of an institutional investor (Value at Risk - VaR).

Two scenarios for the formation of a portfolio of 1 billion rubles are considered.

Scenario A (Basic). The concentration of 80% of assets on one large platform (OIS-1) with a probability of technical failure/ trading stoppage of P = 2%.

Scenario B (Diversified). Asset allocation across 5 platforms with a 20% limit on each with the same probability of failure.

The calculation of expected losses (EL = EAD × PD × LGD) shows that in Scenario A, when the risk of an infrastructure failure is realized, the volume of frozen liquidity will amount to 800 million rubles. In Scenario B, due to the decorrelation of technological risks of different platforms, the maximum amount of frozen funds in case of a single failure is limited to 200 million rubles. Thus, the application of strict concentration limits reduces the infrastructure VaR of the portfolio by 4 times, ensuring continuity of cash flows even in conditions of instability of individual network nodes.

One of the main challenges of the CFA market remains the high asymmetry of information, especially in the segment of medium-sized issuing companies. In 2024-2025, leading rating agencies (Expert RA, ACRA) adapted their methodologies to the specifics of digital rights, starting to massively assign credit ratings not only to issuers, but also to individual issues of CFA.

The conclusion about the critical level of information asymmetry in the CFA market is based on a comparative analysis of information disclosure by debt issuers. According to the analytical review of the Expert RA rating agency "CFA Market 2025: a paradigm shift" ^[37], there is a significant disparity in transparency between the segment of high-yield bonds (VDO) and the CFA segment for small and medium-sized businesses.

Unlike the VDO market, where, as of the end of 2025, more than 95% of issuers disclose audited financial statements in accordance with IFRS and have a credit rating of at least ruB-, in the CFA market in the SME segment, the share of issuers with disclosed high-quality reports is less than 40%. The remaining 60% of borrowers limit themselves to publishing the minimum required data set, which creates a "blind spot" for the investor.

The fundamental reason for this gap lies in the regulatory plane (the legal reason for the asymmetry). A comparative analysis of the requirements of Federal Law No. 39-FZ "On the Securities Market" and Federal Law No. 259-FZ "On Digital Financial Assets, Digital Currency and Amendments to Certain Legislative Acts of the Russian Federation" shows differences in the standards of issue documentation.:

1. Bond market – requires registration of a Prospectus containing a detailed analysis of risk factors, business description, ownership structure and financial analysis (for public offering);

2. The CFA market requires the publication of an Issue Decision, which is often a standardized document (smart contract) describing only the parameters of the monetary claim (term, rate), without in-depth disclosure of the issuer's business risks.

Such a light regime reduces barriers to entry for businesses, but shifts the burden of verification of trustworthiness to an investor who, without special tools (ratings), is unable to adequately assess the risk of default.

The role of rating in CFA risk management is manifested in three aspects:

1. Standardization of valuation – the rating allows the investor to compare CFAs with classic bonds in a single coordinate scale.

2. Reducing regulatory barriers – having a rating of "ruBBB" or higher is a key condition for unqualified investors to access issues in excess of 600 thousand rubles.

3. Verification of the transaction structure – agencies analyze not only the issuer's finances, but also the legal purity of the smart contract, which reduces the risk of technological errors.

Empirical analysis shows that the rating of the CFA issue in 2025 reduced the required yield (spread) by 40-60 basis points by increasing the transparency of the transaction (Methodology for assigning credit ratings to CFA instruments. Analytical bulletin of the rating agency "Expert RA". Extract from the Protocol of the Methodological Committee No. 550 dated 06.10.2025).

Table 4 shows the impact of having a credit rating on the investment parameters of CFA issues.

Table 4 – The impact of having a credit rating on the investment parameters of CFA issues (summary data for 2025)

Source: compiled on the basis of operational reports of the OIC (Sber, Alfa-Bank, Atomize) ^[38], analytical reviews of rating agencies for 2025 (SBER CIB ^[39], BCS Express ^[40]), data from the HSE Center for Financial Research and Data Analysis ^[41] and the Russian Ministry of Finance ^[42]

The data in Table 4 confirm the thesis about the crucial role of credit ratings in reducing information asymmetry in the digital asset market.

By the end of 2025, a "price gap" is clearly visible: the presence of a rating allows the issuer to reduce the required yield by an average of 200 basis points. This is due not only to increased confidence, but also to the expansion of the sales funnel: according to the regulatory requirements of the Bank of Russia, it is the presence of an investment-grade rating that opens access to issues to a wide range of private investors, which creates excessive demand and allows the issuer to reduce the coupon rate during book building.

The effect of saving the cost of funding shown in Table 4 (reduction of the rate by 1.4–2.9%) was obtained based on an empirical analysis of market data. This indicator is calculated by the author as the arithmetic mean deviation of fixed coupon yield rates for initial asset offerings. A representative sample of the 50 largest CFA issues in 2025 (according to the Cbonds information and analytical agency), comparable in duration (up to 1 year) and the volume of the underlying asset, was formed as the calculation base. The sample was divided into two groups: issues with a credit rating (at least ruBBB) and issues without a rating. Comparing the average values in these groups made it possible to mathematically isolate and measure the "premium for information transparency" that investors are willing to concede to the issuer if they receive an assessment from an independent rating agency.

In addition, there is a direct correlation between the rating and the issue volume. Unrated issues remain a niche tool for small businesses with a check of up to 50 million rubles, while rated CFAs in terms of their parameters (volume over 800 million rubles, more than 1,000 investors) have come very close to the market of exchange-traded bonds.

Thus, a credit rating in 2026 is not just a formal requirement, but a key tool for managing the cost of borrowing and expanding the liquidity of a digital portfolio.

Based on the statistics of the structure of new issues in 2025-2026, we can state a statistically significant trend towards the transition from unsecured digital debt to the use of collateralized CFAs. The use of hybrid digital rights (PPP) allows you to legally "bind" a real asset (gold, grain, real estate) to the issue through the mechanism of a smart contract ^[43].

Risk management through collateral in the CFA has significant advantages over traditional collateral:

First, the automation of out-of-court treatment. In the event of an issuer's default, the smart contract can be programmed to automatically transfer ownership of the underlying asset to investors or to launch a collateral liquidation procedure through a specialized Oracle.

Secondly, smart contract insurance. The introduction of insurance products on the market that cover the risk of an IP technological failure or an error in the smart contract code has become an important stage in operational risk management ^[44].

Thus, by 2026, the comprehensive risk management system in the CFA market has evolved from a simple analysis of the issuer's financial stability to a multifactor model that takes into account the technological reliability of the platform, the availability of an independent rating assessment and automated mechanisms for securing obligations.

To systematically solve the identified problems (information asymmetry, fragmentation) and implement a predictive discount reduction model, we have developed a register of events. The ratio of challenges, proposed solutions and the predicted quantitative effect is presented in Table 5.

Table 5 – A set of measures to modernize CFA market regulation and evaluate their effectiveness (author's development)

Source: developed by the author taking into account previously obtained data.

All numerical values of the forecast effects presented in Table 5 are not expert assumptions, but are based on a hybrid calculation methodology that includes extrapolation of historical data from the high-yield bond market, scenario VaR modeling, and a method for transferring the value of real risk management tools ^[45].

The fundamental difference between the author's calculations and the models proposed by other researchers (for example, the algorithmic models for evaluating DeFi tokens L. W. Cong and Z. He or the classical CAPM model proposed by D. A. Kochergin) is to take into account the institutional specifics of the Russian market. If other authors evaluate the abstract risk of blockchain networks or market volatility, then our model isolates and mathematically digitizes the unique "platform closure risk." The values are calculated not hypothetically, but through an assessment of the real transaction costs of closed systems and the actual cost of cyber risk insurance. The details of the author's calculations justifying the data in Table 5 are as follows:

First, the forecast is for a 15-20% reduction in the share of defaults. This indicator was calculated by the author using extrapolation based on default statistics in the lower echelon of high-yield bonds ^[46] (ratings below "B") ^[47]. Applying the scoring model to a sample of historical defaults for 2024-2025, the author mathematically modeled the situation of the introduction of mandatory "risk tags". Calculations show that algorithmic cutting off of unqualified impulsive demand reduces the probability of credit risk realization (Probability of Default) in the retail portfolio by exactly the specified 15-20%, since high-risk assets remain only with investors with sufficient liquidity reserves.

Secondly, the narrowing of the spread by 50-80 bp. This quantitative effect is derived by the author from the analysis of the structure of transaction costs. The monopoly position of the OIC in a closed system requires a hidden commission for maintaining internal liquidity ^[48]. The combination of "glasses" of quotations (cross-chain) modeled in the work mathematically equates the CFA's liquidity to exchange-traded bonds. The difference between the hidden IPO fees and the open tariffs of the Moscow Stock Exchange is precisely the range of 0.5–0.8% (50-80 basis points) calculated by the author, which will reduce the cost of funding for the borrower.

Thirdly, the reduction of infrastructure VaR by 4 times. The assessment of the effect of introducing a concentration limit per operator (the "N6_cifra" standard with a hard limit of no more than 15% of capital) is derived from Markowitz's mathematical law of portfolio diversification and scenario modeling. With a limit of 15%, the investor is forced to distribute capital across at least 7 independent platforms (N=7). According to the diversification rule, the risk of losses decreases proportionally to the root of the number of independent assets (√N), which for 7 platforms reduces the risk by 2.6 times only due to the distribution of the base). However, the main factor is the lack of covariance of technological failures: a server crash or a smart contract error on platform A does not correlate in any way with the IT infrastructure of platform B. Thus, a radical reduction in the volume of assets at risk of simultaneous blocking (from an average of 80% with a monoplatform approach to a regulatory 15%), in synergy with the mathematical effect of diversification, provides a multiple (more than 4 times) reduction in the VaR of the portfolio.

Fourth, leveling the technological premium by 0.3–0.5 percentage points. The estimate of the cost of leveling the technological discount (Smart Contract Risk) was obtained by transferring the cost of real risk management tools to the discount rate ^[49]. According to current data from the insurance market, the standard cost of a corporate cyber insurance policy (covering operational losses and failures of IT systems) varies in the range of 0.3–0.5% of the insured amount. Accordingly, the emergence of the smart contract insurance institute will allow investors to hedge this risk by simply subtracting the cost of the policy (the same 0.3–0.5 percentage points) from the required risk premium, completely eliminating technological uncertainty.

The forecast made in the paper about the inevitable leveling of the "technological discount" is based on the modeling results presented earlier in Figure 1. Comparative analysis has shown that there is an inverse relationship between the depth of platform liquidity and the required risk premium: spreads on highly liquid platforms with a developed secondary market are already 125-160 basis points lower than on niche platforms.

Consequently, the institutional development of the infrastructure (in particular, the emergence of Exchange Operators combining disparate "glasses" of quotations into a single pool) will mathematically lead to a decrease in the Bid-Ask spread for the entire market. Within the framework of our proposed yield formula, this is equivalent to reducing the component of the illiquidity premium (Rp _lig), which means a gradual reduction in the technological discount to the level of standard transaction costs of the classical stock market.

Discussion of the results

The data obtained in the course of the study makes it possible to significantly expand methodological approaches to pricing and quantifying the risks of digital financial assets, which fully corresponds to the stated subject of the study and form a substantive scientific discussion with existing approaches to their assessment.

In contrast to D. A. Kochergin's approach, which suggests applying classical bond valuation models to CFAs, our calculations prove that the direct transfer of traditional tools (in particular, the discounted cash flow method in its basic form) leads to a distortion of fair value. The CFA assessment requires the mandatory inclusion of a premium for specific platform risks, without which the asset looks unreasonably overvalued.

Entering into a controversy with A. V. Dubrovsky and co-authors, who consider digital rights primarily through the prism of transactional efficiency and issue speed, we argue that savings on primary placement costs (convenience for the issuer) are completely absorbed by the demands of investors in the secondary market. The infrastructure premium they put into profitability actually negates the borrower's transactional benefit.

Our conclusions also fundamentally differ from the results of research on decentralized finance (DeFi). If foreign authors (L. W. Cong, Z. He) see the main reason for the spread in the volatility of the token's underlying asset and algorithmic vulnerabilities of public blockchains, then our study proves that the main reason in the Russian market is infrastructural isolation. Within the framework of private blockchains (OIS), it is not the risk of code that comes first, but the "risk of platform isolation" and the lack of cross-chain arbitration.

Thus, the increase in scientific knowledge consists in proving the limited applicability of the classical capital asset valuation model (CAPM) to the digital rights segment. The steady "technological discount" we have identified in the amount of 0.8–1.5 percentage points (see Table 2) is not a temporary market anomaly, but a fundamental payment for market segmentation. Thus, the asset pricing theory is complemented by a new component – the risk premium of a decentralized (fragmented) infrastructure.

Developing a scientific discussion with Russian researchers, whose works are presented in the bibliographic list, it is necessary to directly compare the quantitative results obtained by us with their conclusions. Thus, in the study of M. V. Yurgelas and co-authors, as well as in the work of V. S. Stankevich and A.V. Vlasov, the emphasis is placed on the fact that the CFA infrastructure forms a new reality that certainly reduces the cost of raising capital due to the absence of a chain of traditional intermediaries. However, the results of our paired comparison (Table 2) empirically refute this thesis for the current stage of the market: it has been proven that the borrower's savings on initial placement fees are completely absorbed by investors' demands for increased profitability. Infrastructural isolation creates a stable spread (+0.8–1.5 percentage points), which makes CFAs a more expensive source of debt than classic bonds.

In addition, our approach directly contradicts the conclusions of V. V. Korotkov. If his systematic approach to stock market risk analysis relies on the classical assessment of price volatility, then our scenario VaR modeling proves that applying traditional statistical limits to CFAs leads to a critical underestimation of potential losses. We have quantified that for digital rights, the main driver of losses is not a market shock, but the covariance of technological failures – the risk of a "one-time blocking" of assets within a closed pool of a particular operator (OIS), which requires the introduction of a completely new standard for platform concentration (up to 15%).

Answering the question about the advantages of the proposed toolkit, it is necessary to emphasize that the developed integral risk scoring model (R _total) It is conceptually superior to existing approaches. In contrast to classical models (for example, the approach of D. A. Kochergin or the foreign concepts of L. W. Cong and Z. He), who consider CFA as a complete analogue of traditional bonds or focus exclusively on software code, the author's R _total model directly and mathematically takes into account the effect of "platform closure". The integration of the probability of a specific platform failure _(PI) and the degree of capital loss (I _i) allows investors to avoid critically underestimating the illiquidity risk, which reaches 125-160 basis points on isolated Russian sites. Thus, existing models provide a distorted (underestimated) assessment of risk, while the R _total model transforms a qualitative understanding of infrastructure fragmentation into an accurate quantitative premium to the required profitability.

The practical significance of the work for market participants is as follows:

1. For institutional investors: The necessity of using the proposed scoring model of integral risk (R _total) in asset allocation is substantiated. The application of the developed risk matrix (Table 3) and the limit policy ("N6_cifra") allows reducing the volatility of the CFA portfolio by 1.2–1.5% per annum while maintaining the target yield above the deposit rate.

2. For the regulator: The proposed scenario analysis confirms that the introduction of interoperability (cross-chain operations) is a more effective method of reducing the cost of borrowing for the economy than subsidizing rates, since it mathematically eliminates the premium for illiquidity.

Thus, the results of the study transform the approach to CFA from an intuitive one ("it's just a digital debt") to an evidence-based one ("it's an asset with a complex risk premium structure").

Conclusion

As part of the research, a comprehensive methodological approach to pricing and quantifying the risks of digital financial assets in Russia has been developed. As an answer to the main scientific problem of the article, the author proposed a modified discounted cash flow (DCF) model, adapted to the realities of infrastructure in 2026 by mandatory inclusion of specific premiums for liquidity fragmentation (RP _liq) and operator technological risk (RP _tech) in the discount rate. To quantify these factors, the author's integral risk scoring model (R _total) has been developed, which allows investors to decompose market-wide and unique risks, as well as mathematically calculate fair returns on digital rights, taking into account the effect of "platform isolation."

Based on mathematical modeling and analysis of market spreads, the hypothesis of the multifactorial nature of the CFA cost has been confirmed. It has been proven that the fair profitability of digital law includes a specific premium for the risk of "platform isolation". Mathematically, this is expressed in the need to include the components of the premium for fragmentation of liquidity (RP _liq) and the operator's technological risk (RP _tech) in the discount rate.

In the course of a comparative analysis of Yield Spread Analysis of data for 2024-2025, it was found that the value of the "technological discount" varies from 80 to 150 basis points, depending on the liquidity of a particular information system operator. A strong inverse correlation has been revealed between the availability of the issue's credit rating and the cost of funding for the issuer (the availability of a rating reduces the rate by an average of 2 percentage points).

The developed integral risk scoring model (R _total) allows investors to quantify infrastructure risks.

The use of the scenario approach proved the effectiveness of the diversification strategy: the introduction of a concentration limit per operator (the "N6_cifra" standard) reduces the VaR of the portfolio by 4 times.

The proposed set of regulatory measures (see Table 5), including the introduction of "risk tags" and standards of interoperability, creates the basis for reducing information asymmetry and improving market efficiency.

Given the high demand for the results obtained by the potential readership (primarily institutional investors, risk managers and representatives of the regulator), a promising area of further research is the practical integration of the proposed scoring model (R _total) with the emerging architecture of the digital ruble. Further scientific research will be aimed at modeling scenarios for the complete elimination of the risk of liquidity fragmentation through the introduction of smart contracts with automatic cross-platform calculations. In addition, the cross-border settlement segment requires a separate study and scaling of the developed risk matrix (in particular, the integration of the CFA into the BRICS Pay circuit), which will provide participants in the debt market with universal academic and practical tools for the safe allocation of capital.

The article is published in its final version as approved following the last positive peer review recommending acceptance for publication. It incorporates revisions made by the author in response to prior negative peer review reports that did not recommend publication. All peer review reports, including initial negative reviews, are published in open access alongside the article. All versions of the author’s revisions are archived in the publisher’s repository and may be made available upon reasonable request in accordance with Elsevier’s editorial policies and applicable data availability requirements.
Read all reviews on this article