Comprehensive analysis of synthetic asset mechanism design, application categories and development trends

Synthetic assets are a product of product design, but at the root, they are actually a product of liquidity.

Original Article Title: “Synthetic Assets – Trends and Opportunities
Written by: Cheng Ka Leung, Director of Research, HashKey Capital

Overview of synthetic assets
Two types of synthetic assets
There are two types of synthetic assets within Crypto.

One type is a replication or mirroring of an asset, where the asset itself exists but is chained up or replicated in another chain.

The other type of synthetic asset is the direct creation of an asset that did not exist before, which can be a synthetic index, an assetization of cash flows, such as an arithmetic contract token, or a tokenization of a risk metric.

Asset replication: mainly solves the problem of cross-platform availability, such as the accessibility of stocks, the need for cross-chain assets, etc.

Direct Synthesis: Provides unmet reward/risk exposure for which there is a market demand and which is directly tokenized.

Key Mechanisms
The main mechanisms of synthetic assets are: prophecy machine, liquidator, price pegging, synthesis (mint) and destruction (burn).

The prophecy machine has two roles: first, it is used to quote prices for assets, such as Synthetix’s synthetic asset trading prices. The second is to provide a price indication for debt monitoring, so that when the debt is under-collateralized (e.g., the ratio of lent debt/collateral falls below a certain threshold), a liquidation process can be initiated. For example, MakerDao’s prognostic machine block is designed to determine whether a CDP is safe and whether liquidation will be triggered by price changes.

Liquidators are a class of actors who liquidate problematic lending. Synthetix did not have a clearing mechanism earlier, but in the 2020 Altair upgrade, a clearing mechanism was added. The clearing mechanism provides price assurance and serves as an alert to give CDP collateralizers an incentive to maintain collateralization levels and to clean up missing CDPs (e.g., lost private keys) in order to maintain overall levels. There are also many types of clearing mechanisms, like the UMA protocol which uses a no-feed clearing mechanism, where the prophecy machine is only used when a dispute arises, similar to the final referee of the dispute mechanism, as the tokens are also no-feed, minimizing the use of the prophecy machine.

The commonly used price pegging mechanisms are: overcollateralization, automatic liquidation of faulty CDPs, supply adjustment, emergency shutdown, etc.

The process of synthesis and destruction is relatively straightforward, i.e. pledging minted coins, as well as returning coins and terminating low-pressure positions.

Risk 1- Abandonment of partial power
Asset replication only reproduces the price performance of the asset, while other rights to the asset are largely relinquished. For example, voting, governance, dividends, etc. for stock synthetic tokens, interest, debt claims, etc. for bond synthetic tokens. In the case of stocks, the voting rights are divided into as many shares as a company issues. Synthetic assets create synthetic stocks and the voting rights cannot be divided into synthetic assets. This one is a general problem for synthetic assets in the asset replication category.

At present this is within the acceptable range: 1, the underlying assets themselves are high-risk assets, users only care about price fluctuations; in terms of the stock itself, also few users lie in dividends, governance, after conversion to crypto synthetic assets users care even less; 2, DeFi’s intervention, giving up some rights will be compensated by other mechanisms, such as stable coins can have interest, because he is generated from real lending, and the interest rate is still very high, while the dollar is almost interest-free. Synthetic tokens can also be collateralized, e.g. BTC is a zero-yield asset, but wBTC can participate in mining.

Direct synthetic assets have no such concerns, and the rights attached are bound by agreements, such as arithmetic tokens that demand a return on arithmetic, and risk-graded tokens that demand a return on risk. Direct synthetic assets represent the various interests in the blockchain world that have not been and will be assetized.

Risk 2 – Price Decoupling
Synthetic assets do not truly peg the price of the original asset, such as the decoupling of stablecoins from the dollar, and synthetic stocks from the original stock price. This is addressed by the protocol’s built-in peg mechanism, which is under experimentation.

For example, the peg mechanism used by stablecoins.

Dollar-collateralized stablecoins rely on confidence in the dollar and a two-way arbitrage mechanism for cashing in and out

Crypto asset pledged stablecoins rely on confidence in over-pledging and clearing mechanisms

Algorithmic stablecoins rely on supply regulation and economic incentives

The mechanism is part of the equation, and the confidence and engagement that comes with the mechanism is paramount. The problem that many algorithmic stablecoins face is that the supply regulation function fails when prices are too low, because participation confidence is drastically reduced.

The risk of price decoupling is always present, simply because none of these mechanisms are tightly coupled, so it is normal to be attacked, with the most common dimension of attack being prophecy machine attacks, such as those suffered by Synthetix in 2019. DeFi protocols that combine synthetic assets and rely on prophecy machine quotes are all at risk of prophecy machine attacks.

Risk 3-Collateral
There are two main risks associated with collateral: one is liquidation and the second is the opportunity cost of overcollateralization.

The seller of the synthetic asset, the borrower of the synthetic asset, needs to deposit the underlying assets accepted by the protocol into the smart contract. Because many synthetic asset agreements are plugged into a liquidation mechanism, they will be liquidated when a certain collateralization ratio cannot be completed. Of course, with a debt contract like a CDP, there is also the risk of changing parameters, such as the CDP needing to be held at a stable rate, which is variable and unpredictable through DAO decisions. Or liquidation penalties, changes in debt ceilings, etc.

More risk comes from the opportunity cost, as most collateral is over-collateralized and over-collateralization equals locked-in liquidity. After liquidity mining comes out, the collateral all have the potential to go to mining, all with annualized yields in the tens. As the seller of the synthetic asset, the cost of the meeting needs to be calculated. The parties to the agreement also use mining to give rewards while pledging, for example, in Synthetix, pledging SNX to generate sUSD, you can get a share of the transaction fee from Synthetix.Exchange and an inflation bonus from SNX (similar to staking in PoS chains)

Advantages of synthetic assets

Uncensored issuance paradigm: anyone can issue a specific synthetic asset based on an off-the-shelf agreement

Global liquidity: synthetic assets can be traded directly on exchanges, with decentralized exchanges providing a convenient venue for direct peer-to-peer transfers

Low-friction trading: synthetic assets enjoy the same low costs and constraints as other crypto assets

NFT allows for more types of synthetic assets to emerge: the nature of NFT allows for more granularity in the play of synthetic assets, and the emergence of Uniswap V3 has made this form of synthetic asset more widely accepted.

The role of synthetic assets – meeting specific and practical needs
The role of traditional synthetic assets is very simple and straightforward – to meet the client’s risk exposure needs, which are often not provided by the standard products. The basic design solution is to add multiple derivatives together, or one underlying asset plus derivatives, and most are customized by the investment bank. That is, there is a need first and the investment bank uses the tools on hand to help the client build a synthetic asset. For example, synthetic convertible bonds are built by synthesizing ordinary debt + long options into a fitted convertible bond. A few examples of traditional synthetic assets might be instructive for Crypto synthetic assets.

Alibaba’s synthetic stock
Alibaba went public in the US in 2014 and for various juggling reasons investment bank Merrill Lynch was not able to participate in the IPO process. So the Structured Products Group created a derivative product that allowed clients to “acquire” Alibaba’s stock even before it went public. The product was a synthetic product, with the long side of its components being SoftBank stock and the short side being SoftBank’s holdings in other public companies, such as Sprint, Yahoo Japan, KDDI, etc. The net exposure of the product was Alibaba stock + some smaller shares. This way the client gets the pre-IPO valuation of Alibaba in advance. And the investor also gains a lot, just unwind this product and purchase the underlying shares after the IPO. ML also gains significant income because of the complex operation and product design.

Asustek A50 Index Fund
Prior to the opening of RQFII, international investors had to invest in mainland China indices through the ASUS A50, which is an ETF but also a synthetic product that uses an A-share linked product (CAAP, a type of derivative) to fit the A50 index, filling the gap that overseas investors do not have access to A-share indices. Later, when RQFII was opened, ETFs that could actually invest in stocks became less competitive, but at one point in time, it was a very necessary and unique investment tool.

CDS (Credit Default Swap)
A broader example is a Credit Default Swap (CDS), where a buyer and seller agree on a specific series of bond defaults, and the buyer of the default swap pays a “premium” to the seller, who will sell the defaulted bond to the seller at par, assuming credit risk arises. This gives the buyer an incentive to go long the risky asset and hedge the credit risk. Because of the direct avoidance of credit default risk, CDS was widely welcomed by the market upon launch because it solved a very necessary pain point, which reflected the ability of product design and the precision of market fit.

So synthetic assets can be created without demand, but the real vitality is the creation with demand. We can divide the role of synthetic assets into the following levels.

Liquidity and aggregation of funds: synthetic assets can create standardized products, and standardized products mean that they can carry large amounts of liquidity, for example, by synthesizing a class of indices.

Combinability: The standardization of tokens means that they can be split and combined, and synthetic tokens can be lapped into more diverse products, such as creating a reverse protection token against arithmetic fluctuations.

Cost reduction: Standardized protocols can inherently reduce costs and low friction transactions.

Addressing real needs: Some of the needs here exist now, and some are potential. For example, there has been a great demand for US dollar stablecoins in the past to the present. The use of synthetic assets for risk management will be a big direction in the future, which may now be treated as a speculative product.

The main directions of synthetic assets
The Largest Synthetic Asset – Stable Coins
Stablecoin is the earliest type of synthetic asset, but the anchor asset is more specific to a currency. The mechanism of stable coins is relatively simple, collateralized dollars will hardly need a stabilization mechanism, if it is a digital asset, you have to add an over-collateralization mechanism. If collateral is not used, algorithmic stable coins will need to be designed to curb inflation and deflation. Starting in 2020, a class of resilient supply stablecoins begins to emerge. The mechanism of an elastically regulated stablecoin relies on adjusting the supply, and rebase may cause the entire market capitalization to suddenly skyrocket and plummet.

The problem with algorithmic stable coins is that when prices get too low, the confidence to continue participating is extremely low. Coins like ESD/BasisCash have gone into shock. Frax, which uses progressive steps and combines algorithmic stability with a collateral mechanism, is one of the more unusual algorithmic stable coins in that the collateral is a combination of existing stable stable coins and equity tokens. Under the two-way exchange mechanism, FRAX is freely convertible with aUSDC+(1-a)FXS. 0

Stablecoins have not given up on innovation since their inception, and almost no one now doubts the usefulness of dollar-collateralized stablecoins. Digital asset collateralized stablecoins are also basically fine, but algorithmic stablecoins still have a long way to go. Mining and other models have made algorithmic stablecoins a big hit for a while, but they are still to be verified in terms of stability.

However, algorithmic stable coins have the biggest space: the reason why the community is diligently pursuing stable coins is that in this parallel world, if the dollar continues to be introduced, it will inevitably still introduce the risk of the dollar (increase in issuance, regulation) into the crypto world, and only by creating a completely endogenous (code is law) stable coin, not even pegged to the dollar, can it become a completely endogenous world’s foundation. This was not possible before, but is becoming more and more likely as DeFi grows.

Protocol Classes
Synthetix, the earliest synthetic asset protocol

Synthetix is one of the earliest synthetic asset protocols, coincidentally initially transformed from a stablecoin. 2018 saw the birth of a large number of algorithmic stablecoin projects in the market, many of which fell silent, some like basis quietly emerging in late 2020, and others like Synthetix starting to transform in late 2018.

Synthtix was one of the first DeFi to start liquidity mining: while liquidity mining originated with Compound’s launch of the COMP token last June, Syntheix can be considered an earlier batch (it started online in 2018). Because the liquidity of synthetic assets is low, and mint synthetic assets require a high collateral ratio , which was later adjusted downward (750%->600%->500%), syhthetix gives pledgers a higher return, which is also to ensure the price stability of SNX. Synthetic s-assets are somewhat more attractive than outright purchase of s-assets. About half of the SNX is currently pledged in the network.

Use of dynamic debt ratios: Static debt ratios are generally used for collateralized projects. This means that as long as the ratio between the collateral, such as the price of Ether, and the loaned debt does not fall below the closeout line, the debt will not be in trouble and can be redeemed at any time. But synthetix uses dynamic debt ratios, for example, in the process of casting sUSD, the initial requirement is 500%, then if the SNX price drops and the collateral ratio drops, the platform will ask the user to pledge additional collateral to ensure that the collateral ratio is greater than or equal to 500% at any time. The worst case of this is that the minter needs to have the need to collateralize the SNX at all times (but can increase the buying pressure on the SNX), and also that the minter will not blow up unless it cannot keep up with the pace of adjustment. Some of the other defi collateral defi agreements have closeout line requirements.

Debt pool-based exchange: In addition to mint assets, Synthetix also offers a debt pool-based exchange. There is no counterparty and no liquidity pool, everyone has only one central counterparty – the debt pool. When trading sUSD to SBTC, it is equivalent to the debt pool recovering sUSD and printing an equivalent amount of sBTC. the debt pool actually shares the debt, because the s-asset price changes, and from the initial collateralization process to the price change, all the debt is redistributed across minter, so even if nothing is done, the individual debt will still move due to the s-asset transactions.

This can be used for trader to express a preference for a certain asset class, for example, if he is bullish on BTC, he will print sBTC, so that the sBTC price will rise and the profit will increase. Due to the dynamic adjustment, more SNX will be saved, and if there is an increase in debt, more SNX will be pledged. There is a lot to be said for what kind of s-assets are printed to exchange for sUSD.

MirrorProtocol built on Terra

Mirror is a synthetic asset protocol built on the Terra network that issues synthetic assets called mAssets. mAssets mint coins that mimic real asset price changes. mAssets are fully decentralized, can accept many types of collateral, and are guaranteed to have a sufficient amount of collateral. exchanges on Terra UST pairs for mAsset can be created on Terraswap. mirror also has its own Mirror token, which allows liquidity providers to receive rewards.

There are four types of roles within Mirror.

Trader: A user who buys and sells mAsset on Terraswap, with the portfolio exposed to synthetic assets.

Minter: the person who actually makes the synthetic assets and enters into a CDP agreement with the protocol , the collateral assets must be above the minimum requirement of the mAsset. Therefore, the minter is actually shorting the synthetic assets. Minter can adjust the collateral ratio by burning mAssets or adding more collateral as long as the collateral ratio is above the minimum requirement

LP: LP adding equal amounts of mAsset and UST to the corresponding Terraswap pool will increase the liquidity of the synthetic asset LP taking LP tokens can earn more pool proceeds.

Staker: Pledge LP tokens and MIR tokens to get MIR tokens. LP token holders can get MIR to benefit from inflation and MIR staking to get CDP fees. If users pledge MIRs, they can participate in governance and have some voting rights. Governance is the entry of new mAssets into the whitelist and the change of parameters.

Oracle feeder: Used to provide accurate and precise external quotes, and is the only account that can make updates to asset prices. Very important to the overall operation, the community will be carefully selected and replaced through governance. Currently the Band protocol is used.

The main difference between Mirror and Synthetix is the collateral asset, Synthetix uses volatile SNX tokens for collateral while Mirror uses stable UST tokens, hence the difference in collateral rates. However, UST is not Mirror’s ecological token, so the tradeoff is a tradeoff between functionality and ecological development.

Universal Synthetic Asset Protocol UMA

UMA is a synthetic asset issuance protocol, through which any synthetic asset can be issued. There are currently 11 projects on it, divided into three main categories: Digitally Native Index, Yield Dollar, and Synthetic Asset Exchange.

DigitallyNative Index, which creates tokens with special features, is the most interesting category. Domination Finance tracks the market share of Bitcoin, ETH/BTC tracks the ratio of ETHBTC, uGAS eventually Ether gas fee, uSTONKS tracks Wall UMA has created a series of products for specific needs that make good use of the characteristics of synthetic products.

Standard Product Protocol UMA uses a standard product contract, the ExpiringMultiParty (EMP) contract, which allows developers to quickly bring synthetic asset tokens online that can expire. Once the development contract is complete, a complete set of synthetic tokens can be brought online in 7 steps.

The UMA protocol also has tokens, UMA Tokens, which can be used to resolve its dispute resolution system, DVM (Data VerificationMechanism), and to govern the protocol as a whole. The main uses of holding UMA Tokens are 1) to receive rewards for voting on price demands using the DVM, and 2) to provide governance rewards for upgrades to protocol parameter changes.

Core Mechanism-Solving Prophecy Machine Bribery DVM is used to solve the problem of centralization and prophecy machine bribery, which is itself a prophecy machine. uma believes that blockchain prophecy machine is difficult to avoid being bribed. So v. with economic means. The corruption of the prophecy machine needs cost, i.e. Cost of Corruption (CoC), and profit from Corruption (PfC) of the action, just to ensure that CoC is greater than PfC. You can also calculate CoC and PfC.

Step 1: To measure the cost of corruption, DVM uses a Schelling-Point style voting system and tokenizes the voting rights. Token holders vote at a contested price point and are rewarded for voting honestly and penalized otherwise. As long as there is an honest majority, voters will vote correctly. This means that the cost of corruption is the cost of buying control of 51% of the voting tokens.

Step 2: To measure the profit from corruption, all contracts using the system need to be registered with DVM and report the value that could be stolen if the price source is corrupted (this is contract-specific PfC value). DVM then sums the PfC of each contract to a system-wide PfC number.

Step 3: The CoC > PfC mechanism is enforced by a variable cost policy.

Since the corruptor needs 51% of the token to make 51% of the token’s market value (CoC) > PfC, this requires the DVM to continuously monitor the CoC > PfC relationship.

The UMA protocol itself does not check the sufficiency of tokens, but the token economy gives the clearing house an incentive to identify synthetic token agreements that are insufficient. There have only been six instances.

Unlike the above two protocols, UMA does not use overcollateralization, but rather uses financial incentives to motivate liquidators to liquidate non-healthy CDPs in a timely manner, i.e., actively managing CDPs rather than waiting for the closeout line. UMA uses “no-feed liquidation” to keep the whole body of CDPs healthy through cooperative gaming. One benefit is that some long-tail assets (e.g., no price yet) can also be brought to synthesis without quoting barriers.

Algorithmic Token Protocol
Algorithm-based tokens are a very interesting class of products that meet our criteria: 1) abstracting special resources; 2) using token characteristics to make them tradable, liquid and price-able; 3) combining liquidity mining to make the arithmetic output (yield) also tokenized. Arithmetic token market and is a very promising application of synthetic assets, solving the problem of blockchain underlying circulation, making the original so-called “cloud arithmetic” DeFi and NFT, and having more decentralized concept.

PoW Algorithmic Token MARS

MARS is an arithmetic derivative token project launched by the Coinprint pool, with BTC mining arithmetic credentials pBTC35A, ETH mining arithmetic credentials pETH18C, and the governance token MARS. Each pBTC35A targets 1 TH/s of bitcoin arithmetic, and each pETH18C targets 1 MH/s of ethereum arithmetic. The MRAS protocol is based on Ether.

The MRAS protocol is based on ethereum, and the arithmetic gains are obtained through Staking, pledging pBTC35A for wBTC and MARS, and pledging pETH18C for ETH and MARS. wBTC, ETH and MARS). The parameters of the pool are as follows: Electricity cost: $0.0583/kWh, pool fee (FPPS): 2.50%, power consumption ratio: 35W/T

BTCST

BTCST is similar to MARS, and is also an arithmetic token protocol. Each token targets 0.1 TH/s, and after staking, you get the daily bitcoin arithmetic revenue.

Unlike MARS-based, BTCST is a free-issuing arithmetic token marketplace that any miner can apply to participate in as long as: 1) the contributed arithmetic is owned; 2) the device contains arithmetic that can pass the team’s risk test; and 3) the participating miners have at least 5 PH/s of arithmetic. Then miners can issue their own BTCST arithmetic. The parameters of the mining pool are as follows: electricity cost : $0.058/kWh, efficiency loss : 2.50%, power consumption ratio : 60W/T

Comprehensive analysis of synthetic asset mechanism design, application categories and development trends

Figure: BTCST algorithm tokenization process

HashMix

HashMix is a general-purpose tokenization solution protocol. It can tokenize the traditional BTC/ETH network’s arithmetic, and it can also tokenize a storage mining protocol like Filecoin, i.e. a token transfer protocol that can be used across chains. Unlike the above two products, the standard arithmetic underlay will not be used, so the arithmetic submitted by various miners will become NFT instead of the usual ERC20 tokens. So each miner’s arithmetic can be very different, which is in line with reality.

Each NFT represents a different arithmetic protocol, the authenticity of the arithmetic is verified by a verifier, a multi-signature wallet controls the payment for purchasing arithmetic NFT tokens, and the verifier relies on the protocol token HSM for incentives. In addition, arithmetic tokens can enter the financial market and be attached to lending and trading functions. For example, the Fil class of mining arithmetic has a lending requirement.

Risk Management
The price risk of digital currencies has always been high, and users with trading experience can go straight to shorting to protect their positions. Modeled after the real world, many synthetic assets have been designed with a newer step. For example, they are designed for volatility, or for risk stratification. Risk-based synthetic assets are practical, and are designed for risk-averse needs, so we expect there to be a market for them. But risk-based tokens are inherently risky, and also subject to the risk of attack from prophecy machines, arbitrage, etc.

Volatility Class

Volmex launched an index based on Ether volatility and developed a tradable platform based on this index, which allows users to trade and price the market directly against the risk of Ether (option-based volatility) and can hedge against the risk of market volatility, or the risk of the entire Ether ecosystem. Like the UMA based dVIX index issued by UMA as well, it can offer a similar trading product for ETH volatility (volETH) and also short volatility (ivolETH)

Risk Tiering Protocol

Barnbridge is a risk stratification protocol that divides the pool of lending agreements such as Compound into Seniortranche and Junior tranche, with Senior receiving a risk-free rate and Junior receiving a risky rate but with a higher return. Each type of tranche has a corresponding token, sBONDs, jTokens, that addresses the representation of returns to users with different risk preferences when the returns of the lending pool change. These products are called Smart Yield Bonds.

Another product from Barnbridge is the Smart Alpha Bond. although called a Bond, the SAB itself is not for fixed income, but for any measurable return, such as the layering of price volatility risk for BTC/ETH. For example, the price range of ETH is divided into three parts, each of which can bear different returns, so that an ETH can be split into three different tokens, such as jETH/mETH/sETH, allowing users of ETH to have a larger choice of options.

Saffron uses a similar structure, making three tiers at the fixed income level: S, AA and A.

Asset package class
The Charged Particle protocol can inject any of the erc-20 representatives into the NFT tokens to “charge” the NFT tokens. Because of the cooperation with AAVE, the deposited token can be turned into an atoken, i.e. a token with interest, such as Dai, which can be turned into aDai, and there is no limit to the type of assets deposited, which can be ERC20, ERC721, ERC1155. This turns an NFT into an asset package, which can be all-inclusive. This turns an NFT into a package of assets that can be all-inclusive and is truly “synthetic”.

Charged Particle provides an idea, in fact, synthetic assets is a packaging process, NFT can naturally combine a variety of assets, what kind of assets you want, you can package in. The trick lies in the distributability of the revenue and the verifiability of the underlying support.

Framework and Liquidity
HartLambur, the founder of UMA, has combed through synthetic assets and given a conceptualized synthetic asset formula: collateralized assets + spending function = synthetic assets. We feel that here the expense function needs to have another layer of non-financial (mainly governance) meaning in addition to the financial meaning, as the claim on the asset is effectively moved from the right to the left. According to this framework, two categories need to be considered.

Adequacy of collateral assets: generally high quality assets. and the smoothness of the prophecy machine mechanism and the liquidation mechanism

Stability of the expense function: how to seamlessly transfer the right of return over

The one synthetic asset that is really doing well is the stablecoin. The track comes mainly from demand, and the stablecoin is a demand, and a long-term demand. As long as the dollar system remains the same (about pricing), and the access to and from the channel remains the same (about the relationship between the two worlds), this demand will not change.

What we currently find most promising: direct synthetic assets that solve real problems, such as arithmetic tokens, risk tokens, and complex revenue tokens (e.g., odd options).

One last point, synthetic assets are a product of product design, but at their root, they are actually a product of liquidity. Whether it is traditional synthetic assets, the largest synthetic asset stable coins, and various products that may be created in the future, there is liquidity behind them all. The spillover of liquidity is not only supplying, but it is also allowing demand to rise. These demands, all revolve around two levels of risk exposure and enhanced utility. Assuming there is no liquidity, there is no demand, and without liquidity, no one will create synthetic assets, and no one will trade them when they are made. So in the grand scheme of things, synthetic assets are still thanks to this era of liquidity flooding. it is All About Liquidity.

Posted by:CoinYuppie,Reprinted with attribution to:https://coinyuppie.com/comprehensive-analysis-of-synthetic-asset-mechanism-design-application-categories-and-development-trends/
Coinyuppie is an open information publishing platform, all information provided is not related to the views and positions of coinyuppie, and does not constitute any investment and financial advice. Users are expected to carefully screen and prevent risks.

Like (1)
Donate Buy me a coffee Buy me a coffee
Previous 2021-06-04 06:18
Next 2021-06-04 06:23

Related articles