Getline Whitepaper Article Series:
Credit Risk Management & Loan Lifecycles
We’re pleased to introduce the second of a series of articles based on our whitepaper. The purpose of this article series is to run through the whitepaper in layman’s terms, and make it as simple as possible to read and understand the details of what the Getline Network is, how we’ll function, how we’re structured, and how we plan to move ahead. These articles will be shortened & simplified versions of the whitepaper content. See the whitepaper for the full content.
Just as a short intro, Getline is disrupting a $1 trillion market through blockchain technology.
The Getline Network is a peer-to-peer lending market on the Ethereum blockchain. The platform will allow for instant and direct lending in cryptocurrencies. Getline will initiate a unique architecture for a credit risk prediction market to make lending safer for lenders and more accessible for borrowers. We aim to revolutionize the peer-to-peer lending market and fully decentralize it, making it easily accessible, safe, and compliant to serve as an infrastructure for a new kind of global financial system. For those of you who don’t know, peer-to-peer (or P2P) refers to the ability of two individuals or entities, to exchange value with each other directly, without the need of a middleman.
All Articles:
1st Article — Introduction, Aim, & Our Idea
2nd Article — Credit Risk Management & Loan Lifecycles
In this second article of the series, we’ll cover sections 4–5, which are:
- 4 Credit risk management
*4.1 Introduction
*4.2 Collateral
*4.3 The risk assessment mechanism. Credit scoring market
*4.4 Credit scoring process
*4.5 Metascoring information flow
*4.6 Risk analysis market
*4.7 Calculating metascore
*4.8 Mitigating risks on ARA prediction market
*4.8.1 Aligning ARAs’ interests with lenders — the reserve system
*4.8.2 Preventing ARAs’ attacks against each other
*4.8.3 Preventing ARAs’ attacks against lenders
*4.8.4 Preventing ARAs’ manipulation of their own metascore
*4.9 Preventing strategic defaults
*5 Loan lifecycles
Let’s get started…
4 Credit risk management
4.1 Introduction
Investing in loans always has a risk of loss due to defaults, which can happen because of an inability to pay, or because the borrower has calculated that it’s more profitable not to pay (a strategic default). The risk of defaults is usually balanced by interest rates (default premiums), collateral, or a combination of both. The amount of either is usually based on the borrower’s credit score (their creditworthiness). The credit score & collateral help to discourage strategic defaults. Defaulting on a loan negatively affects one’s credit score. A good credit score allows a borrower to borrow more easily, with lower interest rates & less collateral.
The Getline protocol minimizes fraud risks for lenders, & leverages the above mechanisms in a decentralized manner, by using blockchain technology & the Ethereum network. Section 4 covers the key elements of credit risk management in the Getline Network:
• The risk assessment mechanism,
• The credit score information flow,
• The collateral design,
• Risk of the premium market,
- Out-of-blockchain debt collection.
4.2 Collateral
All Getline loans will use GET tokens as collateral, which assures lenders they won’t lose all their money, & ensures lower interest rates for borrowers. Collateral will be crucial in our bootstrapping phase & in expanding, as new borrowers with no previous loans in the system may need to collateralize as much as 80%, 90%, or even 100% of their first loans. After building sufficient credit history & proving creditworthiness, they might get loans secured by fewer GET tokens, yet no less than 5% of a loan’s value.
4.3 The risk assessment mechanism. Credit scoring market
The Getline Network will assign risk assessment to for-profit, third-party entities called Attestors of Risk Analysis (ARAs), who will serve as credit rating agencies. Their role will be to cryptographically sign a given loan request with a projected default rate, based on risk examination. Getline will provide an automatic scoring mechanism assessing the historical accuracy of all ARAs’ default rate predictions, based on accessible information in the blockchain. The ARAs risk prediction accuracy scores will be available to all lenders & borrowers through the application layer of the system.
These ARAs will create a prediction market & compete for price and accuracy of default rate predictions. The market aspect will hold ARAs accountable for their predictions & stimulate innovation.
Automated & transparent scoring of credit rating agencies, based on openly accessible on-chain data, is a unique feature of the Getline Network. Legacy rating agencies can lack transparency, as their performance is seldom audited or scored. This gives us a huge advantage over traditional credit risk prediction markets, and could be a vast disruption of the financial sector.
4.4 Credit scoring process
To get a loan, a potential borrower will choose an ARA based on its price & accuracy score. Then they must provide documentation that ARA requires, to confirm their identity, determine the purpose of the loan, attain credit history, & other factors indicating their creditworthiness. Aside from basic identification information, the ARA decides what other information it would seek from potential borrowers. To retain a good accuracy score at Getline, the ARA should calculate default rates as accurately as possible. ARAs will likely collect information usually collected by banks & other lending institutions today, such as:
• The borrowers’ FICO® score or equivalent,
• Information on credit history & outstanding loans obtained from on-chain and off-chain financial institutions,
• Income statements,
• Confirmation of equity ownership.
but some might request other types of data that prove to be increasingly useful in credit scoring, such as:
• Machine-readable data about financial behaviors (e.g. leveraging EU’s PSD2 regulation),
• Data from broadband/network providers regarding browsing habits,
• Behavioral data from social media,
- Verification from biometric identification systems (e.g. Indian Aadhaar).
All ARAs must comply with privacy laws and regulations within their jurisdictions, which may prohibit the use of such information to calculate creditworthiness. The credit assessment phase is conducted off-chain & abstracted from the protocol.
After receiving the information, an ARA issues a default absolute probability for the borrower for a given loan, & the investor’s expected capital value at the end of the loan term. The default rate is based on the borrower’s creditworthiness, and is shown as a percentage. The investor’s expected capital value at the end of a loan term will account for the default rate, interest rate, and the expected currency price changes if a loan is denominated in a currency different than the currency of such a loan.
With a predicted default rate, the borrower may issue a loan request to the network, with the amount, a maximum acceptable interest rate, & the time of the request’s termination. The request is immediately visible to potential lenders in the loans browser, where they can request more information or confirmation from another ARA.
After this stage, potential lenders can access all of the credit risk indicators in the Getline browser:
• From borrower: basic personal data, loan goal, repayment plan, answers to extra questions;
• From ARAs: predicted default rate, additional information about the borrower’s credibility, and the subsequent ARA verification results;
• From Getline: ARAs metascore and certification issued by Getline.
Lenders then decide whether to enter a bid in the auction for a proposed interest rate. After the request’s termination time passes & the lenders have committed their funds to the loan, the contract is considered concluded.
4.5 Metascoring information flow
All funded loan information is placed on the blockchain & will be available for further audit. After a loan matures, data about either its default or repayment is gathered by Getline to calculate the ARA’s metascore. During this phase, Getline will also review the current certification of the ARA. Information on the ARA’s current metascore & certification will be public.
4.6 Risk analysis market
The Getline Network will connect investors all over the world with borrowers already scored by their creditworthiness. Investors will also see their expected returns for their investments. By ARAs conducting the risk assessment process themselves, Getline creates a competitive edge compared to centralized lending markets, that by design cannot scale past their primary jurisdictions. For example, the Getline Network enables a Chinese investor to fund a loan for a Nicaraguan borrower, thanks to ARAs that operate locally, but are scored in comparison to their global competition.
Lenders in the Getline Network decide on their capital allocation based on:
• The absolute probability of default for a given loan,
• The value of an investor’s capital at the end of the loan term,
- The attesting ARA’s metascore.
The ARA’s metascore is it’s score of past default prediction accuracy. It is represented by a value between 0 and 1. The better an ARA’s accuracy, the higher their metascore value. The metascore & other loan details will be displayed in the Getline browser. Metascores will allow lenders to assess the trustworthiness of a given loan’s predicted performance and to offset any uncertainty of an ARA’s estimates.
Borrowers will want ARAs with good metascores that offer good prices, thus ARAs will want to achieve the highest possible metascore. The price of an ARA’s services will be predetermined and calculated as a percentage of paid back capital and interest.
4.7 Calculating metascore
Since these articles were meant to be simplified and shortened sections of the whitepaper, in layman’s terms to be precise ;) we’ve omitted any complex equations or explanations of those equations within this section. If you’d like to read the more techy info, you can do that by reading our whitepaper, pages 14 & 15.
The metascore will be calculated outside of the blockchain by the Getline browser application layer, which will display the metascore & other loan characteristics, such as credibility & competence rates of the ARA’s systems. Getline would provide easy access to historical data of ARA’s ratings & scoring inside of the Getline browser.
Transparent, automatically calculated, & easily available scoring of ARAs will lead to a vibrant, competitive, & ever improving credit risk predictions market. We expect the market to eventually experience regional and vertical specialization, as some ARAs might concentrate on certain regions of the world, e.g. the EU countries or India, & specialize in various market segments, like loans offered to individuals or micro-businesses.
Getline is planning to establish a Getline ARA as the first ARA on the network, in order to bootstrap the Getline Network. We are also going to invite other institutions to join the Getline Network as ARAs to help create a competitive predictions market from the very first day.
4.8 Mitigating risks on ARA prediction market
Although creditworthiness scoring is the key element of the Getline Network, necessary for trust, a decentralized lending market allowing pseudonymous users creates new risks & potential loopholes unknown to the traditional financial markets. This section explains how we plan to protect the integrity & accuracy of the ARA prediction market.
4.8.1 Aligning ARAs’ interests with lenders — the reserve system
To protect lenders against failed ARA predictions, ARAs must stack up on GET tokens, a percentage of which get paid to the lender if the borrower defaults on their loan.
MORE equations! Please see the whitepaper, pages 15 & 16.
ARAs will be required to maintain their provisions ratio in order for loans scored by them to be presented in the Getline browser. Failure to keep the required provisions will result in their delisting.
Similar to the borrower’s good reputation giving them the ability to borrow more with the same collateral, a higher metascore will allow for lower provisions and a bigger value of outstanding loans for ARAs.
4.8.2 Preventing ARAs’ attacks against each other
ARAs may attack other ARAs by creating fake or fraudulent identities, obtain good scores from their competitors, and then default on a loan scored positively by the competing ARA to lower its metascore. Such activities would be subject to criminal liability. ARAs are likely to use various identity verification techniques to avoid positively scoring fraudulent identities. The on-chain solution to this potential problem involves ARA’s usage of external trusted identities, such as Boson Identity Management, which the Getline ARA will be using, and which can also be used to exchange information with other entities using Boson.
4.8.3 Preventing ARAs’ attacks against lenders
ARAs could use their position to issue high credit scores to identities under their control & later use them to obtain loans and default on them. However, this behavior would not pay off as each default would lower the ARA’s metascore, and in effect lenders’ trust in credit scores issued by them. Every established ARA would in effect lose its market position, making this type of fraud economically non-viable. The only ARAs that could profit from such a fraud would be new, unestablished, ARAs. The Getline Network will require new ARAs to collateralize part of each positively scored loan with GET tokens. The required collateral will be reduced with a sufficient metascore & the number of verifiable credit scores issued by an ARA.
4.8.4 Preventing ARAs’ manipulation of their own metascore
Finally ARAs might try to trick the metascoring system by creating false identities, scoring them with high scores, obtaining credit from other accounts under their control and paying them back. This would increase an ARA’s metascore and trick potential lenders & borrowers into believing that it is trustworthy. We are exploring various ways to prevent such an attack, including analyzing token flows to detect such frauds. In case of detecting such fraud, the Getline organization would evict such a fraudulent ARA from the Getline Browser. Getline is also going to implement an on-chain solution that will create economic costs for fraudulent ARAs performing such attacks — the Network Trust Fee.
The Network Trust Fee (NTF) is one of the most important features of the GetLine Network. It is the main mechanism preventing ARAs from manipulating their metascore, and thus preserving the integrity & trustworthiness of the network. The NTF is meant to incur economic costs to any manipulation of metascores, unethically leveraging pseudonymity, & concluding contracts between one’s own Ethereum wallets.
The NTF fee is 1% per annum of any loan created in the Getline Network. Every Getline contract burns an amount of GET tokens, paid as collateral.
4.9 Preventing strategic defaults
Strategic defaults are deliberate defaults by borrowers. These can result from economic calculations, when it is found that a default would be cheaper than the loan repayment. There are many costs involved with defaulting on a loan. Costs of defaulting on a loan include: the loss of collateral, damage to one’s credit score, and debt enforcement proceedings.
Furthermore, lenders on the Getline Network would submit information about the defaulted borrower’s identity to the chain, which would increase interest rates required by future lenders for the next loan to that borrower, or collateral requirements would be raised for the next loan, in order to get an acceptable default rate from an ARA. Their damaged credit score will not only remain on the Getline Network, it will also affect their off-chain credit score also, such as the FICO® score, as ARAs will pass information about defaulted loans to credit rating agencies. Some ARAs could even offer reporting services to lenders, making reporting defaulted loans even easier.
Loan agreements using blockchain could be just as enforceable & binding as those concluded using more traditional means. In all researched jurisdictions there is no legal reason why a loan agreement cannot be a purely oral contract. Structuring it as a Smart Contract and putting it on a blockchain would result in a binding and enforceable contract in most jurisdictions. It can be argued that the U.S. Electronic Signatures in the Global and National Commerce Act (ESIGN) and the state laws modeled on the Uniform Electronic Transaction Act (UETA) allows for digital signatures to have the same effect as a physical signature and provides a sufficient legal foundation for a blockchain-based Smart Contract to be enforced under current U.S. law. Ultimately, it will be the ARAs’ responsibility to determine whether or not variations by jurisdiction affect the viability of the concluding loan contract with the borrowers from a given jurisdiction, & account for this factor in the issued credit score. If a loan agreement concluded in the form of a Smart Contract is enforceable, a lender, or a hired debt collector, can obtain a payment order issued by a court or other competent authority and enforce such a contract. Defaulting on a Getline loan could result in personal assets being liquidated by law enforcement to satisfy the lender.
5 Loan lifecycle
The process of obtaining a loan in the Getline Network:
1. A borrower can apply for a loan either through the Getline browser, an Ethereum dApp developed by us, or by using a third-party interface integrated with the Getline Network.
2. If the borrower has no previous identity identified by an identity provider, they would be redirected to boson.me or civic.com to get verified.
3. A verified borrower will be able to specify the desired terms of a loan — its value, duration, & maximum interest rate. Determining whether the loan would be denominated in a cryptocurrency or in a fiat currency will also be possible.
4. The Getline browser or a third-party application will then issue a loan request to the Ethereum blockchain. The request will list all the terms of the loan specified by the borrower. At this stage, the loan request would get funded with collateral in GET tokens.
5. The loan request will then be scored by an ARA picked by the borrower. ARAs will attach their scoring to the loan request and sign it with their unique signatures. The process of credit scoring will vary depending on the ARAs’ policies.
6. The assessed loan request will then go up for public auction for investors to bid for the lowest interest rate.
7. Investors who offer the lowest interest rate will then fund the loan. In exchange for the loan, the investors receive ERC20 tokens representing rights to the loan’s repayment.
8. After the loan matures, the borrower might:
(a) Repay the loan in full — the contract would release the repayment proportionally among holders of ERC20 tokens representing right to repayment, & the collateral would be returned to the borrower, minus the Network Trust Fee;
(b) Pay only a fraction of the money — only full repayment will be accepted by the Getline Network lending contracts, thus the loan contract would return the money to the borrower;
© Fail to repay the loan — in this case investors will be able to extend the repayment period; if they do not, they will be able to fight for the repayment in court or sell their ERC20 tokens of non-performing loans to specialized debt collectors.
We hope you enjoyed this article, keep an eye out for the next article in our series,
Have a great day!
The Getline Network team.
If you enjoy our articles, please like and follow us on Facebook and Twitter.
Also, join our Rocket chat if you’re interested in contributing to our platform.
