Research paper for LendingCalc
Date: September 19, 2017
Marketplace lending is blossoming in Asia, following the trend in China where over 2000 platforms are purportedly open for business. The majority of Asian platforms were founded in or after 2015, a few years after the Chinese boom, and are able to incorporate the lessons learned from the Chinese market. This is also facilitated by the flow of Chinese capital into the region.
As domestic Chinese marketplace lending platforms are virtually closed to international investors by capital controls, this note will focus on South East Asia, where regulators welcome the participation of foreign institutional capital in marketplace lending, as in the United States or the United Kingdom. Nonetheless, marketplace lending assumes a different character than their Anglo-American counterparts due to divergence in regulatory regime and business environment.
The importance of regulatory regime
Platform is the putative strategy and business model of marketplace lending. Lately the word “platform” itself has taken on hallowed meaning in the fintech world, particularly since the Chief Executive Officer of Goldman Sachs proclaimed that “We’re a technology firm—we’re a platform” on Bloomberg Television.1 Unfortunately this newfound halo only obscures the reality that platforms are not equivalent across geographies, because it is their regulators and not the platform themselves that lay down the rules governing their structure.
To review the basics – A marketplace lending platform matches borrowers with lenders individually, hence also the term peer-to-peer (P2P) lending. Each loan may be matched to and funded by many lenders, hence the term crowdfunding; and each investor may match and fund many loans to diversify the investment. Exactly how the matching occurs is up to the regulator. The most relevant aspects are:
- Whether balance sheet lending (and hence leverage) is allowed;
- The responsibilities (and liabilities) to retail investors;
- The restrictions regarding institutional investors.
In the United States, Prosper and Lending Club are governed by SEC rules, and for all practical purposes behave like broker dealers. They originate the loans and sell securities called notes backed by the loans they originate under a program prospectus. They are technically securitizations, but unlike the CDOs of old, the notes do not correspond to tranched portfolios to enhance the credit rating; and no complex derivative is involved.
As for the three aspects above, broker-dealer designation implies that (1) there is no restriction to lend with its own capital; (2) sale to retail investors is allowed but strict SEC regulation applies and (3) institutional (or accredited) investors may participate.
To illustrate the divergence in the other extreme, China does not allow platforms to lend using their own capital, forbids the participation of institutional investors in marketplace lending and forces platform accountability to retail investors not by securities regulation by through strict disclosure policies. Curiously, under Chinese regulation, marketplace lending platforms are not technically financial institutions but are designated as credit information platforms.
Divergences in regulation directly impact the sustainability of the business. While not entirely obvious, marketplace lenders do spawn credit portfolios in the process, notwithstanding that the majority of which is eventually sold. Credit portfolios come with their long tails in loss distribution. How investors cope with the long tail and how platforms share this risk matter a great deal. The economics of a lending business, after all, cannot escape the logic of risk-return and the cost of funding. Peer to peer or not, lending has been about matching loans to funds for centuries. An unstable platform structure, begotten by clumsy regulation or a clumsy reading of regulation, will vitiate the revenues of the entire business. Structural risk is in fact the single most important risk factor in Asian platforms, and sadly, often poorly understood.
Large structural issues aside, smaller nuances, in cumulative effect, also influence the behavior of platforms. Lending cap is an illustrative example. While Singapore and Malaysia do not impose lending caps, Indonesian regulators have set a cap of SGD 200,000 per borrower, at least for now. Hence Indonesian platforms that are able to acquire a large number of borrowers consistently will generate portfolios with far greater diversification benefits.
The emerging market business environment
Small businesses in Asia face significant financing gaps. Such small businesses range from sole proprietors hawking goods in street bazars to more conventional small and medium enterprises (SMEs) with national or even regional supply chains. Measuring this financing gap is no simple matter. According to a study by IFC2, some 92 million SMEs in East Asia remain unserved or underserved.
As for access to personal credit, a large segment of the population remains unbanked or under banked. Some 55% of the word’s unbanked and underbanked population resides in Asia Pacific region3. The rise of marketplace lending (and fintech in general) is often hailed as the digital age’s cure to these chronic ills. The market size is immense.
That technology shall deliver the bulk of humanity from financial penury appeals to fintech idealists, and has rallied countless entrepreneurs under the financial inclusion banner. Reality, on the other hand, remains ever so unfailingly sober. Imperfect rule of law, absence of data infrastructure and deeply entrenched business practice define the business environment of marketplace lending in Asia.
Successful platforms adapt to this reality and thrive. The characteristics of loans generated reflect these adaptations, the most significant of which are negative enforcement mechanisms and partnerships.
Many Asian countries have just started constructing credit bureaus (e.g. Indonesia) or do not enable marketplace lending platforms to provide defaulting records to credit authorities (e.g. China). In some cases, the legal procedure to enforce default to prove financial fraud is too costly or burdensome to pursue. Hence negative enforcement mechanisms that seek to create negative consequences for unpaying borrowers supersede judicial options. These mechanisms need not be brutal or violent collection practices. For example, a notification to an SME borrower’s vendors, or the mere threat of one, could generally force the borrower into repayment or negotiation. Sharing a black list with other platforms or lending entities is also effective.
The use of partnerships is often motivated by negative enforcement, since partnerships generate a specific context where borrowers cannot default with impunity. For example, a partnership with an e-commerce platform that refers merchants as borrowers and freezes merchant accounts in case of defaults offers powerful negative enforcement. An agreement of program financing with a large manufacturer for its suppliers follows a similar logic. Some fintech practitioners coin the phrase “ecosystem play” to describe these mechanisms, but in essence such agreements rely on a partner’s ability to create and implement negative enforcement mechanisms and to bolster loan growth. Partnership loans therefore often feature prominently on Asian platforms, and well executed partnerships will provide platform investors with handsome risk-return.
At this point fintech enthusiasts may feel disenchanted that terms such as big data, machine learning or analytics have not found their way into the discussion. Such concern is partially warranted. The prevalence of data analytics in developed markets have reaped the benefits of data infrastructure efforts spanning decades by the private sector and the public sector. Lending Club could “easily” declare a credit score cutoff of 600 because Fair, Isaac and Company (now known as FICO) began building statistical credit scoring models in 1956, and because US public authorities have kept orderly records social security numbers to identify unique individuals, and because the US Postal Service began introducing the five-digit zip code system in 1963 which facilitated the construction of address databases. Most Asian marketplace lending platforms (with the possible exception of those in Singapore) must cope with lesser data infrastructure at the national level.
This does not mean that data analytics efforts in Asia will wither. The data landscape in Asia is actually more complex. In some areas such as credit bureaus, insufficient default histories simply take time to build up and will lag developed markets. In other areas such as the incorporation of social media driven payment data in anti-fraud modeling and supply chain logistics, Asia may leapfrog developed markets, as China has successfully demonstrated with payment infrastructure. As Asian countries fixate their attention on data infrastructure, marketplace lending platforms must navigate the data landscape as “work in progress”, but may also position themselves to contribute to the national data infrastructure through public and private sector partnerships.
For example, in Malaysia, private credit bureaus proactively solicit default data from marketplace lending platforms. In Singapore and Indonesia, more sophisticated private sector partnerships exist to share logistics, inventory and transaction data across regional supply chains between logistics companies, manufacturers and marketplace lending platforms to facilitate invoice financing. Data does exist in Asia, but who is willing to share what with whom is equally, if not more, important. Claiming technical superiority is inconsequential if the platform cannot sign agreements to unlock key data sources. The risk-return of loans generated by “automated” underwriting is not just about technical prowess.
Conclusion: What this means for investors
Marketplace lending in Asia offers investors fixed income exposures to a region of vibrant economic growth where conventional fixed income products are hard to come by, and without the volatility of marked to market instruments. The risk-return profile can meaningfully enhance any fixed income portfolio, but only for those who know to curate these platforms properly:
- Avoid platforms with structural risk.
- Seek out platforms well versed in regulation – and in dealing with regulators.
- Beware of platforms that boast analytics without explaining how they source data.
- Study how the platform implements negative enforcements.
- Look at how the platform handles partnerships, and who negotiates them.
The list is not exhaustive, but hopefully points investors in the right direction. If it reminds us of the business acumen of old, that should be no surprise. Lending is an ancient business, where risk emanates not from numbers but from human capriciousness. Those who can blend the latest analytics in capturing and measuring risk with the prudence to manage the psychology behind it will likely survive.
1 Bloomberg TV interview, June 2015
2 Closing the Credit Gap for Formal and Informal Micro, Small, and Medium Enterprises, IFC, 2013, p.8
3 See PWC DeNovo Q2 2016 FinTech ReCap and Funding ReView and datatopics.worldbank.org/financialinclusion/home
Terry Tse is a LendingCalc advisor focusing on the company’s strategic direction and global partnerships. He currently oversees global expansion at the largest B2B payment company in China, LianLian Group and serves as an adviser for the Southeast Asian P2P firm, Funding Societies. Formerly, Terry was the Chief Risk Officer at Dianrong, where he executed a ten-fold increase in loan origination for the P2P platform. Terry holds a B.S. in Mathematical and Computational Science and a M.S. in Financial Mathematics from Stanford University.