Address the following:
What is the core concept(s) in each of the learning materials for this week?
What ideas do you questions/disagree with, if any?
Describe how digital lending is used in various contrived and economies for improved financial well-being and governance.
Chapter 17
Online Marketplace Lending
Copyright 2018. De Gruyter.
All rights reserved. May not be reproduced in any form without permission from the publisher, except fair uses permitted under U.S. or applicable copyright law.
Online marketplace lending refers to loans originated from Internet-based businesses
rather than traditional banks. In this chapter, we will focus on online marketplace
development in two major economies—the United States and China—and compare
similarities and differences in online marketplaces between the two countries, including the drivers that caused the differences.
US
Online marketplace lending platforms, such as Prosper and LendingClub, emerged
in the mid-2000s to provide an alternative source of loans for the underbanked in the
United States. These early platforms matched individual borrowers and lenders on a
Dutch auction model. The rise in online marketplace lending naturally led to scrutiny
by the U.S. Securities and Exchange Commission (SEC), which resulted in the temporary closure of these online marketplace lending platforms, as they were labeled
“sellers of investments” rather than merely lenders. However, operations at Prosper
and LendingClub resumed in 2009 following registration with the SEC.
China
Similarly, China’s first online marketplace lending platform, Pat Loan, emerged in
2007 to match individual lenders and borrowers. These marketplace lenders expanded
rapidly in China as traditional banks focused on lending to large businesses or stateowned enterprises. Alongside the rapid growth of Chinese marketplace lenders,
there was a general absence of standardization and security of products. As a result,
Chinese regulations during the initial period allowed borrowers to bear a very low
cost when they defaulted. This resulted in the loss of trust among lenders and investors, as they were not confident that borrowers could repay these loans.
This changed with the deployment of machine learning as borrower screening
became a more effective process and supervised, deep learning algorithms could pick
borrower features that would result in lower default rates and, equally important,
better identify fraud. Coupled with more stringent regulations that were initiated in
2015, Chinese marketplace lending began a period of fast and sustained growth.
Today, these lending platforms have become recognized as part of the smaller
borrower’s lending options and any default on the platforms will be reported to the
credit bureau and officially recorded.
DOI 10.1515/9781547400904- 017
EBSCO Publishing : eBook Academic Collection (EBSCOhost) – printed on 10/8/2023 2:35 PM via UNIVERSITY OF THE CUMBERLANDS
AN: 2026908 ; Pranay Gupta, T. Mandy Tham.; Fintech : The New DNA of Financial Services
Account: s8501869.main.ehost
298
Chapter 17: Online Marketplace Lending
Institutional Investors
As the online lending marketplace became more mature, it transformed from that of
a cottage industry to one that has gained institutional recognition. Banks and institutional investors are funding loans as compared to wealthier individuals funding these
loans in the past. Now, only 35% of the loan dollars are coming from fractional loans.
In 2017, the other 65% of the more than US$3b loans on the Prosper and LendingClub came from investors snatching up whole loans, which traditionally have almost
always been made by institutional investors rather than individuals. Banks want to
engage with these online marketplace lenders as they are drawn by the prospect of
strong cash flows while avoiding the costs required to underwrite and service these
loans. Furthermore, banks have started evaluating the loans that originated from marketplace lenders and have purchased these loans directly onto their balance sheets in
ever-larger quantities.
As recognition from the institutional space grew, interest for this asset class followed suit. Institutions like the idea of trading liquidity in exchange for getting a
short duration loan portfolio and a healthy credit spread. On top of that, rating agencies are currently looking at this space and have given the green light to these securities, acknowledging that a well-constructed loan product qualifies for investment
grade rating. Thus, the combination of these forces allowed the securitization of these
loans to occur and make it accessible to a much broader set of investors. Moreover, as
most institutional investors can only invest in graded securities, this helps to expand
the market and provide more liquidity as well as validation to the online marketplace
lending ecosystem.
New Borrowers
As lenders profiles changed, so did that of borrowers. The average credit score for
borrowers has increased year over year, which is a result of two factors. First, the type
of borrowers coming to marketplace lenders has changed. Previously, the majority
of borrowers that came to an online marketplace lender tended to be marginal borrowers who could not get a loan from a bank. Now, there are many individuals who
can get a loan from a bank, but prefer the convenience and efficiency of going to a
marketplace lender.
Second, marketplace lenders have expanded into new borrowers, including providing student loans and loans to SMEs. The learnings from risk profiling and fraud
prevention have given marketplace lenders confidence to expand into these markets,
as regulators have grown more comfortable with the business model and are willing
to include new borrower types.
EBSCOhost – printed on 10/8/2023 2:35 PM via UNIVERSITY OF THE CUMBERLANDS. All use subject to https://www.ebsco.com/terms-of-use
Requirements for Online Marketplace Lending
299
Requirements for Online Marketplace Lending
Establishing an online marketplace lending platform requires the usual business
and technological investment of any internet-based business; however, there are four
additional requirements for marketplace lenders: borrower data, historical default
rates, the risk framework, and a machine learning platform.
The heart of the online marketplace lending business model is replacing human
loan officers at physical bank branches with a centralized, programmatic approach
to loan approval and rate setting. This requires the business operators to effectively
combine consumer risk and advanced data analysis. The required components are
both financial and technological expertise as well as the hardware and coding ability
to operate the machine learning platform. The requirement discussions that follow
assume a cold start, where vintage or processed data is not available.
Borrower Data
Borrow data is the raw material for analyzing loan quality and training a machine
learning platform to select and grade potential loans.
Potential data sources include public, proprietary and third-party data. Public
sources of borrower data typically include government or public research figures,
such as unemployment figures for specific geographies or household debt-to-income
ratios. Proprietary data could be drawn from companies that already have interactions
with potential borrowers, such as payments providers, retailers, budgeting apps, etc.
After a sufficient period of operation, borrower repayment history can also be another
category of proprietary borrower data. Third-party data sources may include FICO
scores, information verification databases, etc.
All borrower data, whether structured or unstructured, must be restructured into
a homogenous taxonomy to ensure accurate processing by the machine learning platform. The taxonomy should be designed around the key features that the marketplace
lender selects to use as the determining factors for creditworthiness. These features
will be shaped by the risk framework as well as the available datasets. Data must also
be pre-processed for correlation between features, such as credit score and age.
Historical Default Rates
To better evaluate borrower creditworthiness in an automated fashion, marketplace
lenders must have access to historical loan default data. Ideally, this data would come
from a dataset comprised of existing or former loans; however, this is not always possible. As a result, data on loan repayments and defaults must be acquired and, in the
best possible way, be sequenced against the available borrower datasets to create a
EBSCOhost – printed on 10/8/2023 2:35 PM via UNIVERSITY OF THE CUMBERLANDS. All use subject to https://www.ebsco.com/terms-of-use
300
Chapter 17: Online Marketplace Lending
more detailed picture of binary repayments and defaults. When it is not possible to
correlate the available borrower data to the available loan data, online marketplace
lending operators may consider relying on their machine learning platform to explore
potential correlations and patterns between seemingly disparate aspects of the data.
Risk Framework
Online marketplace lending employs supervised machine learning and classification,
and the risk framework is the starting architecture for that classification. A critical
task is to take the available borrower features, as determined by the available datasets, and assign default probability weightings to the features. The default probabilities are likely to be updated as the machine learning model learns and as the data
science teams improve the model, but a base setting is required to begin this type of
supervised computational exercise.
Machine Learning Platform
Running a machine learning platform that can productively analyze large volumes
of correlated and uncorrelated data requires experienced individuals who can both
program the algorithms and distinguish the features of the borrower. These data scientists may not need a financial background, but it may be beneficial. Backgrounds
that are heavy in modeling and statistical analysis, such as meteorology, engineering,
or mathematics, are needed.
In addition to the people who set up and run the machine learning system, a
variety of technical architectures are possible for the type of high-performance computing that machine learning requires. Graphics processing units (GPUs) have been
the standard for machine learning platforms, but new alternatives are emerging
including tensor processing units (TPUs) and application-specific integrated circuits
(ASICs). For data storage, solid state drives (SSDs) are the standard as they have faster
load times. On the other hand, there are growing numbers of sophisticated cloud
computing services that allow the user to rent computational power, but these may
come with time limits.
Case Study: Applying Machine Learning to Online Marketplace
Lending
Having assembled the required building blocks to run an online marketplace lending
business, the company must develop a machine learning platform from the composite elements. This section provides a process-level overview of the establishment,
EBSCOhost – printed on 10/8/2023 2:35 PM via UNIVERSITY OF THE CUMBERLANDS. All use subject to https://www.ebsco.com/terms-of-use
Case Study: Applying Machine Learning to Online Marketplace Lending
301
training, and testing of a machine learning platform for online marketplace lending.
The process consists of three steps:
1. Clarifying the business case and data model
2. Applying the algorithms to identify patterns in the data
3. Deploying the platform and ongoing iteration
We first discuss these steps and then explain a periodic model refit process.
Business Case and Data Model
The first step is to translate the business case into a data model. The majority of online
marketplace lending businesses have similar business needs, which include the
primary tasks of evaluating loan applications and selecting appropriate interest rates.
For approval/rejection calculations, the answer is a discrete value, but the lending
rate is a continuous value. In either case, most online marketplace lending operators
use supervised learning at the start of the analysis because the end target is already
known. Unsupervised learning can be used to identify patterns once the model has
been active for a period of time.
To build the data model to analyze each business need, the borrower data drawn
from the application must be structured into consistent features (such as credit score,
gender, income, and so on) that are then overlaid with all third-party and proprietary
data. Each feature is then placed in a separate line entry for each applicant profile.
Taken together, the full table of features, often numbering in the hundreds, forms the
training dataset.
Predetermined targets and definitions are then established for each process.
Typical targets are approved or rejected for the application, and typical definitions
include paid off, default but now creditworthy, charged off, and so forth.
Identifying Patterns
The next step is to apply the algorithm(s) to the training dataset to allow the machine
learning platform to identify patterns among the features. Assuming that the marketplace lender is a new company, the process is known as a cold start, which means no
prior analyzed data is available for comparison.
The algorithms typically fall into a few categories. One of these are simple binary
rules—such as whether the credit score is below a certain threshold—the application
is rejected. Another more complex algorithm would be a logistic regression, where the
risk model is used to apply default probabilities for each of the features (for example,
default_probability = 0.5*fico + 0.2*age + 0.1*balance, etc.).
EBSCOhost – printed on 10/8/2023 2:35 PM via UNIVERSITY OF THE CUMBERLANDS. All use subject to https://www.ebsco.com/terms-of-use
302
Chapter 17: Online Marketplace Lending
Table 17.1: Example of application of a risk model to predict loan status
Loan Id
Amount
Age
Gender
FICO score
Bank balance
Loan status
,
,
,
Female
Male
Male
$,
$,
$,
Paid off
Overdue
?
A third type of algorithm is a word-to-vector analysis, where the relationships between
words are quantified based on observed patterns of usage. This can be especially
useful for data that is not readily incorporated into the overarching data structure. In
a word-to-vector analysis, relationships between words are plotted as vectors to reveal
patterns. The classic example is [boy]—[girl] + [queen] = [king], which is one of the
relationships illustrated in Figure 17.1.
Figure 17.1: Plotted vectors for several word relationship examples
Each of these algorithmic strategies can yield different results, but to combine all
of them into a multi-layered analysis, a deep learning neural network is required.
Simply put, a deep learning platform correlates the probabilities between the results
of each algorithm in a matrix, which raises the overall accuracy of the model’s predictive ability.
Deployment and Iteration
Having identified higher probability patterns in the data, the top-performing strategies are selected for live testing. The live application data received by the marketplace
lender will be divided into traffic channels, sometimes as many as ten, to test the
relative strength of each strategy against live data.
Many firms running big data analyses will commonly run what is known as a
champion challenge, where each traffic channel competes against the others over a
EBSCOhost – printed on 10/8/2023 2:35 PM via UNIVERSITY OF THE CUMBERLANDS. All use subject to https://www.ebsco.com/terms-of-use
Comparison of US and Chinese Online Marketplace Lending
303
defined period of time. The lowest performing traffic channels are cycled out for refitting or retirement, and the firm can either increase the traffic allocation of the higher
performing channels or introduce new channels for live testing.
While these are running, the marketplace lender must establish business intelligence monitoring systems to assess the quality of the decision-making as well as
to identify any anomalies as quickly as possible. Hotfixes are performed in situations where the strategy might be working correctly, but where the physical architecture is not operating properly or where the different components of the platform
(the data processing unit, data warehouse, algorithmic engine, and so on) are not
interacting correctly.
Further iteration of the strategies that are currently used with live data is driven
by model deterioration. Model deterioration comes from multiple sources, including
changes in customer behavior, business processes, competitive landscape, regulations, products offered, or business expansion. In each of these cases, high-performing strategies are refit according to the changes without the model being taken offline.
While day-to-day iterations and hotfixes can be done while the model is still
running, periodic reviews are still required for the data as well as the operation of
the machine learning platform. For the data, the infrastructure of the data warehouse
must be evaluated for performance and suitability, especially as the size of data
grows. Checkpoints must be monitored at all the points where data is transferred from
one component of the platform to another. Each step of the extraction, transformation, and loading of data into the data warehouse needs to be reviewed for any errors
or issues with data handling and classification. From an operational perspective, the
communication between components must be reviewed for deterioration, including
examining the API calls between each of the interfaces.
Similarly, when introducing a new product, the first step of establishing the business case and data model must be reconstructed; however, the second and third steps
of identifying patterns and iteration during deployment can be followed as laid out in
the preceding discussion.
Comparison of US and Chinese Online Marketplace Lending
The US was the first market where online marketplace lending developed on a large
scale; however, China quickly followed because of the high level of personal bank
accounts and mobile-first online services. Both have developed into major industries, with billion-dollar champions leading in each country. However, significant
differences have emerged in the way those leading players run an online marketplace
lending business.
Many of the changes come from differences in regulatory regimes as well as customer behavior patterns. Major differences in the terms and interpretation of data
privacy laws in the US and China significantly change the way online marketplace
EBSCOhost – printed on 10/8/2023 2:35 PM via UNIVERSITY OF THE CUMBERLANDS. All use subject to https://www.ebsco.com/terms-of-use
304
Chapter 17: Online Marketplace Lending
lenders gain information for acquiring and evaluating borrowers. US regulation
largely held back the development of online marketplace lending for several years
after its inception, while in China, it exploded before more stringent regulatory oversight was established. In addition, the US has multiple decades of personal credit
history, while in China, although the People’s Bank of China (PBOC) has established
a reliable public credit bureau, historical data is limited and entirely unavailable for
some new borrowers.
Drawing on primary research and interviews with platform operators in both
countries, these primary differences can be categorized into four areas: customer segmentation, customer acquisition, charge-offs, and cost of funding.
Customer Segmentation
US
In the US, customer segmentation has matured from an early stage where marketplace
lenders were mostly targeting borrowers who could not receive loans from established
banks to now competing with banks for prime borrowers as well as boosting returns
through selectively lending to sub-prime borrowers.
US marketplace lenders typically target a spread of prime and sub-prime borrowers, with the majority of the loan book coming from B- and C-grade borrowers. This
approach increases yield for investors, while the platform uses risk mitigation strategies to reduce overall risk. The majority of the borrowers on the US online marketplace lending platforms come from marketing, referrals, and direct traffic. Unlike in
China, making loans to constituents of a wholly-owned vertical is not very common.
China
Chinese marketplace lenders tend to target a more diverse set of customers. One of the
major platforms focuses on younger borrowers with limited or no credit history. As a
result, this platform has developed a proprietary credit-scoring system to supplant
the limited public data available. Another platform targets borrowers with credit
cards, relying on credit card issuer data as the basis of their credit-scoring system.
A third model used by a leading marketplace lender segments their target customers
based on their employment and assigns weightings to different industries to form a
credit-scoring system. As a result, one of the leading platforms had more than 85%
of its loan categorized as C- or D-grade borrowers, illustrating the Chinese platforms’
willingness to take risks on new borrowers in an effort to gain market share.
EBSCOhost – printed on 10/8/2023 2:35 PM via UNIVERSITY OF THE CUMBERLANDS. All use subject to https://www.ebsco.com/terms-of-use
Comparison of US and Chinese Online Marketplace Lending
305
Customer Acquisition
US
Given that US marketplace lenders target customer segments with more established
credit history and demonstrated repayment ability, a commonly preferred and often
the largest customer acquisition method is direct mailing because it allows the platform to directly reach their ideal borrowers. Email is often the second-largest channel
for US marketplace lenders because they have been in operation for longer and have
a larger pool of prior borrowers. In addition, email is a nearly free outreach channel.
Referral partnerships are used selectively by US marketplace lenders to target new
demographics through partnerships. Direct website traffic, either through organic
visits or digital marketing, is a less preferred channel because of the higher cost of
such campaigns.
As the sector is more mature in the US, most US platforms are willing to trade
slower growth for sustainable customer acquisition costs. As the core competitors in
the US are traditional banks, maintaining lower operating costs is a key competitive
advantage that is rarely sacrificed in search of faster growth. As a result, marketing
costs are often less than 4% of the total origination cost for US marketplace lenders.
China
While most Chinese borrowers have had bank accounts for some time, historically,
many have not taken out loans from the traditional banks where they have their
accounts. This is partly because banks have focused their loan portfolios on stateowned enterprises and large companies, but also because there are not many sophisticated retail-focused products available.
As a result, a large portion of customer acquisition happens online for many of the
Chinese marketplace lenders. For some platforms, as much as 70% of borrowers come
through online channels. In addition to customers acquired through digital marketing, many Chinese marketplace lenders use aggregation platforms and partner with
lenders that have different target customer segments, to refer across to each other.
Access to domestic champion platforms, such as WeChat or Alipay, gives marketplace
lenders a large volume of target customers. However, the acquisition cost can vary
depending on the rate negotiated with the platform.
Charge-offs
US
In the US, the maturity of the financial services regulation, combined with the extensive reach of legal recourse means the majority of charge-offs come from credit losses,
which are triggered after four months of consecutive missed payments. Fraud is a
EBSCOhost – printed on 10/8/2023 2:35 PM via UNIVERSITY OF THE CUMBERLANDS. All use subject to https://www.ebsco.com/terms-of-use
306
Chapter 17: Online Marketplace Lending
small portion of charge-offs for US marketplace lenders, as they invest heavily in
borrower verification, sometimes including manual verification. The predominant
model is to pay for fraud upfront in the prevention process. US platforms outsource to
the same collection agencies as the traditional banks. Unlike in China, debtors have
protections, including cease-and-desist orders to bar communications, the use of a
debt settlement company, and fixed hours during which they may be contacted about
repayment.
China
Due to the imperfect Chinese credit system, the default cost of the borrower is very
low and legal recourse is not always available. As such, marketplace lenders in China
have had to work hard to obtain the trust of investors and grow the industry. While
credit loss is still a factor in China, fraud far outweighs it, both in terms of notional
value and overall volume.
Many have re-tasked their AI platforms, in combination with human investigation, to reduce fraudulent losses. The fraud recognition rate reported through the
machine learning of one marketplace lender’s anti-fraud model is around 60%. In
addition to monitoring transactions and payments, Chinese marketplace lenders also
have access to SIM card data, which allows them to map borrowers against known
databases of fraudulent phone numbers and accounts to discover risk in real-time
and raise a fraud alert. In addition, some Chinese marketplace lenders include a provision for fraud in their business model, while they take time to work out their fraud
prevention strategies.
Cost of Funding
US
Marketplace lenders in the US typically maintain a low cost of capital, preferring
sticky sources of capital that will buy loans through economic cycles. They often work
with dedicated funds on long-term deals with caps limiting the funds’ percentage of
the marketplace lender’s total volume. Between 25–30% of loans originated on US
online marketplace lending platforms are typically held by banks. Retail investors are
still part of the funding mix for the two older participants, Prosper and LendingClub,
but retail investors typically contribute less than 25% of their total funding. Since the
SEC changed the licensing regulations, it has become much harder for later participants to access retail investors. The two most common exit strategies for investors in
US marketplace loans are securitization or hold and yield.
In addition to external investors, many of the US marketplace lenders operate a
warehouse facility for loans they originate, which gives them a stake in the profits as
well as emphasizing their investment to external investors.
EBSCOhost – printed on 10/8/2023 2:35 PM via UNIVERSITY OF THE CUMBERLANDS. All use subject to https://www.ebsco.com/terms-of-use
Comparison of US and Chinese Online Marketplace Lending
307
China
In China, retail investors have retained an important part in the funding mix. For one
major player, as much as 85% of outstanding loans are funded by individual investors, with the rest funded by institutional investors. Another leading Chinese marketplace lender recently reported that 100% of new investors in a calendar quarter were
added through online channels, which is another way of saying that these are retail
investors. While the larger marketplace lenders in China continue to receive institutional support in the form of direct investment capital and IPOs, both in China and the
US, funding for loans continues to be dominated by retail Chinese investors.
Challenges and Areas for Further Development
The development of online marketplace lending has been driven by commercial
factors at the borrower level, technological developments at the infrastructure level,
and burgeoning investor interest at the funding level. These factors have and will
continue to be guided, and occasionally restricted, by local regulatory bodies. As the
industry develops, clear challenges and promising opportunities lie ahead.
One of the primary challenges that marketplace lenders will face in China, the
US, and other markets is the limitation on data sources and usage. As many of these
businesses are being built, online data is also growing in volume and depth. However,
scandals about misused data have heightened awareness, among both the public and
politicians, about the problems with data privacy and use of data. Regulations, such
as the European Union’s General Data Protection Regulation, which require higher
levels of explicit consent for data collection and use, may limit the data available to
marketplace lenders as they seek to build more accurate risk models for borrowers.
Marketplace lenders in developing markets are less likely to encounter such regulations as quickly, but the pushback is likely to happen eventually.
Another significant challenge is the economic cycle. The last time the global
economy experienced a significant contraction in 2008, only a handful of marketplace lenders were operating, and even those that existed had different operating
models than the ones we see today. A risk model designed to capitalize on growth
opportunities in a period of economic growth is likely to be ill-prepared for handling
the fundamental changes to economic conditions associated with a major recession.
The online marketplace lending industry will need to demonstrate a sustained capacity to manage credit risk, while still providing sufficient return to continue to attract
investment.
Each of these challenges, if mishandled, are likely to draw unwanted attention
from regulators. This may include punitive rules that push back on the amount and
types of data available to be used in machine learning processes to such an extent
that the models begin to lose the ability to make more accurate predictions than their
competitors, including traditional banks. If significant losses result from an inability to react to increased risk, regulators may restrict the platforms’ ability to loan to
EBSCOhost – printed on 10/8/2023 2:35 PM via UNIVERSITY OF THE CUMBERLANDS. All use subject to https://www.ebsco.com/terms-of-use
308
Chapter 17: Online Marketplace Lending
customers or create more onerous risk-provisioning requirements similar to the Basel
Capital Adequacy Ratio rules that have forced traditional banks to divert significant
portions of their free cash into reserve provisions.
While these issues will provide plenty of challenges for marketplace lenders, the
opportunities on the horizon are also compelling.
As the industry grows, opportunities for scale will begin to attract competitors to form alliances or even merge. While operating costs are lower for many of
these platforms, relative to traditional banks, scale will help merged marketplace
lenders reach a larger pool of borrowers more efficiently. At the same time, larger
loan origination volumes may strengthen the hand of larger marketplace lenders in
the growing securitization market. Another key component of the maturing of the
online marketplace lending industry may be the increasing standardization that
often accompanies such mergers.
Another major opportunity for growth will be in new borrower categories, especially SME lending. US$600b in SME loans were originated as of 2015, and that figure
is growing as the US economy continues to grow. Successfully applying the risk management principles as well as attracting businesses as borrowers will be a major
opportunity for marketplace lenders to grow their transaction volume. On a related
note, trade financing also holds considerable promise, especially for Chinese marketplace lenders, as they can look to fund the capital needs for suppliers throughout
China’s vast, layered network of supply chains.
FinEX Asia is the first fintech asset management firm connecting professional and
institutional Asian investors with high-quality investments. The company maximizes
return for qualified investors through technology, professionalism, expertise, and an
extensive global network. Beginning with our flagship US consumer credit fund, FinEX
Asia has leveraged its technology core to build new bond, hybrid, and leveraged funds
and expand from asset management into private equity. Founded in Hong Kong, FinEX
Asia now has offices in Taiwan and Singapore, with a dedicated technology team in
China. FinEX Asia is licensed under the Hong Kong Securities and Futures Commission.
For more information, please visit www.finexasia.com.
Dianrong is a leader in online marketplace lending in China. Founded in 2012 and
headquartered in Shanghai, Dianrong offers small businesses and individuals a comprehensive, one-stop financial platform supported by industry-leading technology, compliance, and transparency. The company’s sophisticated and adaptable infrastructure
enables it to design and customize lending and borrowing products and services based
on industry-specific data and insights, all supported by online risk-management and
operation tools. Dianrong’s specific offerings include marketplace lending-related services and fintech solutions. Dianrong was named in 2016 to the executive directorship of
the National Internet Finance Association of China, led by the People’s Bank of China.
For more information, please visit www.dianrong.com/en.
EBSCOhost – printed on 10/8/2023 2:35 PM via UNIVERSITY OF THE CUMBERLANDS. All use subject to https://www.ebsco.com/terms-of-use
Order an Essay Now & Get These Features For Free:
Turnitin Report
Formatting
Title Page
Citation
Outline
