The Bank of Canada released a workbench account that examined loan agreements in the decentralized financial sector on the causes of instability and its relationship to the price of data-encrypted assets. It finds that it illustrates the potential ways to improve the DEFI loan service platform, or it may be a specific limitation of decentralization.

In this article entitled "on the vulnerability of Defi Lending" published on February 22nd, the authors recognize the diversity offered by Defi and the advantages of smart contract agreements over people's administrative law enforcement procedures-but they also point out the systematic shortcomings of Defi. Information asymmetry is a key problem for regulators, which is described in the original text, but another turning point is that in DEFI, information asymmetry benefits borrowers:

"the collateral composition of the loan pool is not easy to note, which means that borrowers have a better grasp of collateral quality than lenders."

This is mainly because the borrower knows at least the quality of the assets used as collateral for the loan. In addition, "only representative assets will be pledged loans as collateral, but such assets usually show high price uncertainty." This paper feels that the price and liquidity will also produce a feedback loop system, and the asset price will affect the loan volume and thus the asset price.

In addition, the lack of human resources investment in blockchain intelligent contracts may have an adverse impact. The credit commissioner can change the traditional loan contract based on current information. However, smart contracts are not flexible, and because the provisions are written in advance and "only in a small part of measurable real-time dynamics", even minor changes to the contract may take a long time to explore.

"therefore, DEFI loans generally involve linear, non-recourse debt contracts, which are characterized by excessive pledge is the only risk control."

As a result, compared with traditional finance, high efficiency, complexity and coordination ability are reduced, and "self-worth emotion-driven pricing cycle time" appears. By using high numbers, the author discusses some problems of completing market equilibrium under such circumstances.

Research shows that a flexible optimal debt limit can provide a balance. However, the "simple linear shaving standard" generally set in the blockchain intelligent contract cannot implement flexible limits. Establishing protocols with this capability will be difficult, and they will be highly dependent on the selection of Oracle. As an alternative to this challenge, "debt pledged loans can abandon real decentralization and re-introduce manual intervention to achieve immediate risk control."

Therefore, the author concludes that the three dilemmas of decentralization, simplicity and security have not been rid of.