Recap of Chief research Qiushan(Tom) Liu’s presentation at BRAINS Conference

On Sep 28, 2020, Qiushan(Tom) Liu joined BRAINS, and in Bytom Session he shared Bytom’s research on Stablecoin system, DeFi and how Bytom will implement Practical Risk Management.

Qiushan(TOM) Liu’s presentation of his paper <MovER stabilized, decentralized finance system with practical risk management>.

Here is the recap of his Presentation:

As I’m the speaker, Qiushan Liu from Bytom foundation， DeFi, the decentralized finance is the most popular development direction of blockchain.

As you can find in the picture, the whole Ethereum DeFi ecosystem, including stable coins, insurance, lending, and exchanges. Among them, stablecoins are the most important part of DeFi. However, more and more financial risks come up and threatening the blockchain ecosystem, Defi , and especially stablecoins. So far, there is no powerful risk evaluation framework for these systems.

Nowadays, the current suitable coins, the system are out of date. For a novel stablecoin Systems, it should hand diversified collateral framework, however many projects like MakerDAO rely on Etheruem too much, and have no ability to include diversified a crosschain assets such as BTC into its collateral Framework, which makes the system suffer from a serious single point of risk. And second, their stabilizing mechanism, always. omit traditional financial risk management methodology.

Bytom presents MoVER — a powerful, all-round risk evaluation framework.

First, it can distinguish between market risks and operational risks.

Second, its the first stable coin system based on classical modern risk financial risk management model, such as JLT, LDA volatility models, and VaR calculation.

Third, the setting of system parameters established under the guidance of these models, forming a trinity risk control system of experience, data, and model.

Risk management is a core module, our system consisting of external risk models and internal risk models.

The internal risk models are made up of the queuing unit, Markov chain unit, and the LDA unit.

The external risk models mainly rely on VaR Unit.

For an internal risk model. We expand Jarrow-Lando-Turnbull(JTL) model, which is a kind of Markov Chain based credit risk model widely adopted by modern financial risk management, and treat the loan liquidation process as finite l state-space Markov chain to reflect the quality and trend of loans in the stablecoin system.

Here we adopt our periodic time homogeneous, continuous time Markov chain, and defined financial state space, safe danger, repair, and clean.`

For the external risk model, volatility Forecasting techniques are chosen to evaluate external market risks by establishing a systematical, VaR calculating, and backtesting framework

As the current cryptocurrency market, obvious van tails and all volatility clustering. VaR models need to be calling with complex distribution, or high order stochastic simulation methods, such as Syria. Class family models are built, heard historical simulation, and Monte Carlo simulation in order to avoid underestimating the small probability events.

As we find the T distribution is more suitable for the Bitcoin market.

Let me give a conclusion for our stablecoin system, we finally set up a systematic risk management system for stablecoin, providing all-round financial parameters guidance at all times.

Also, our evaluation shows that classic risk management tools can also play an efficient effect on the cryptocurrency market.

We pay much attention to the current DeFi Industry’s security, we think professional risk management should be adopted by the currently DeFi industry as soon as possible.