ZhouPeer-to-peer (P2P) financing has been a very controversial subject in China recently due to several frauds and the resulting arrests of the top management of these companies. Since 2015, many P2P platforms including Ezubao, the Dada Group, the Kuailu Group, the Zhongjin Group and others have been charged with illegal fundraising, involving tens of billions of yuan. This is not confined to China. In May, the U.S. Treasury Department released a report criticizing the peer-to-peer (P2P) lending business and recommended it be more tightly regulated.

According to the industry website WDZJ.com, China’s P2P online finance industry reached 2.036 trillion yuan (about $300 billion in transaction volumes) by the end of May 2016. It took seven years to reach its first trillion yuan and just seven months to reach the second trillion.

“Online financing requires higher professional standards than either the internet business or the finance industry. Companies expand too fast and their risk-control systems can’t match business growth,” says Zhihan Zhou, CEO of Kaixindai Financing Services Jiangsu Co., a Nanjing-based, state-owned P2P online finance company. The 39-year-old Zhou, a veteran of small and medium enterprises (SME), and the credit and microfinance markets, had worked earlier with the China Development Bank (CDB). Kaixindai was set up jointly by China Development Bank Capital (a wholly owned subsidiary of CDB) and Jiangsu JinNong Co., a state-owned company that mainly offers IT systems, industry training and coaching services for microfinance companies.

As at the end of 2015, its accumulated transaction volumes exceeded 15.38 billion yuan. Kaixindai has been among the top 10 players in China’s P2P online financing industry for many years. In an interview with Knowledge at Wharton, Zhou shared his views on the industry including its business model and risk control.

An edited transcript of the conversation follows.

Knowledge at Wharton: Regulators had expected that the P2P online financing industry would position itself as an “information intermediary” and would not be involved in money deals or the credit business. However, we understand the income of most of the information intermediaries is inadequate to cover operational costs. So is this model workable?

Zhihan Zhou: Information brokers in finance have a lot of precedents to follow. Who are the biggest information brokers in the Chinese economy? They are the Shanghai Stock Exchange and the Shenzhen Stock Exchange. They are pure information brokers who deal with large enterprises and listed companies. P2P financing mainly involves personal consumption loans, car loans and mortgages as well as supply-chain finance for SMEs and other financial services, in which the information-broker model can match well.

“Many P2P platforms start from individual consumption credit without step-by-step evolution. This causes trouble subsequently.”

In terms of technical challenges, the most difficult and complicated financial business is the comprehensive service for big enterprises. It has to integrate all kinds of financial products including investments, credit, debt, rentals, trusts and securities. Sometimes it involves foreign exchange and all kinds of complicated derivatives. Traditional financial institutions play a big role in this market.

However, for some simpler deal structures in a more information-open environment, you can help two parties strike deals purely by “information symmetry.” The stock market, for example, has played its role as an information broker just based on a strict information disclosure system.

In the past, due to technical and other barriers, only some very high-quality companies or some good-quality listed companies could get direct financing by information disclosure. However, this has changed since the emergence of the internet, which enjoys a natural advantage of enhanced information symmetry. Direct-financing service now has opportunities to expand downstream to SMEs and even individuals because of the lower cost of information symmetry. Therefore, the notion that P2P online financing companies can become information brokers can be feasible at least in theory.

Knowledge at Wharton: How do you view the predicament of the P2P industry in China now? For example, the non-performing loan (NPL) ratio is very high on some platforms.

Zhou: If you make a list of enterprises that need direct financing based on the challenge level to reach information symmetry from easy to difficult, at the highest end are publicly listed companies. Then come the big companies, the SMEs and the microenterprises. At the lowest level is credit to individuals. The challenge for our industry goes from small to big. Many P2P platforms start from the lowest end (individual consumption credit) without gradual evolution. This causes trouble subsequently.

Take the Lending Club in the U.S., for example. It originally hoped to evaluate personal risk based on data extracted from Facebook, Twitter and other social platforms. That is in America, which has much more sophisticated credit investigation and personal data systems than in China. So you can imagine a large amount of P2P business based on personal credit in China will meet trouble in operation if there is no appropriate risk control system in place.

On the growth path of Kaixindai, we have also gradually expanded from big to midsize and to small enterprises. We have not entered the personal business yet. Our shareholder CDB mainly serves big companies and began to be involved with some SME business later on. When Kaixindai started, we targeted midsize and small enterprises. We think a step-by-step mindset can be more effective to play the function of pure information intermediary.

“There is one feature of the finance industry — the one that grows the fastest, will also collapse the fastest.”

Knowledge at Wharton: How does your platform charge for the service now? Is it enough to run the company?

Zhou: From the operational angle, Kaixindai is sustainable. We charge around 1% annually based on financing volume. The highest might be 1.2%, and the lowest might be 0.7% to 0.8% annually. This charge, though low, covers our operational cost and we are still able to invest on some data accumulation and investor acquisition. So, in practice, this P2P online finance model as information broker is feasible.

Knowledge at Wharton: Kaixindai has an advantage as both your shareholders are big, state-owned companies, which enables you to develop clients at a low cost. If you didn’t have this background, would it add to the cost of your operations?

Zhou: In the early stages, because our platform is supported by the financial office of Jiangsu province, many microfinance companies would recommend high-quality clients to us. Meanwhile, the finance office has background data for all microfinance companies in the province, which enables Kaixindai to use the monitoring system to filter, control and monitor clients’ quality and to control our risk.

It’s been three years since we started. We haven’t had any NPLs as of today; payments are all on schedule. Our performance today is completely based on market-oriented operations now.

In fact, for financial institutions, there has to be different priorities in different industrial lifecycles. Under the current macroeconomic situation, if we focus on SMEs, the risk may be increased. After the economy gets more stable in the future, the proportion of small loan business may increase.

Knowledge at Wharton: So how does Kaixindai differentiate between financing parties who borrow money from it, and control the risk?

Zhou: I always insist on one thing: P2P online financing requires higher standards than both internet companies and financial institutions. It’s not like some people thought that, as long as you develop an IT system and connect it with some credit data, then you can do it. That’s not true. This business requires a lot of data and models and connects them with relevant trading scenarios.

Why do many P2P companies fail to continue their business? It might be because they expand too fast — their operational capabilities including data accumulation and risk-control models haven’t caught up. For a zero-based P2P startup, how can it achieve very low or zero NPL?

Take our shareholder CDB for instance. Its NPL is below 1% for either big or small enterprises. This is based on decades of experience. Kaixindai, which focuses on microloans, can achieve zero NPL mainly because both of our shareholders have accumulated a lot of data on the microfinance industry since 2009. Compared with a company with no databases, we can do much better on risk control.

If there are no synergy advantages, the costs for opening offline chain stores are actually very high.”

Asset choice has to match risk pricing. Many clients of our supply-chain financing business are among the top 50 private companies in Jiangsu province, or industry leaders in the province. Some of them are also doing business with CDB, which has a synergistic effect and relatively lower risk for us. Their high credit level can offer a reasonable and safe return for investors. Only when we know the financing parties well, do we do business with them.

Knowledge at Wharton: What is the major spending of your platform? How do you guarantee the growth in scale that will in turn result in income growth?

Zhou: The major spending of our business is on human resources. We have hired more than 100 people now. The second-biggest spending is on technical systems. (But) we are very constrained on marketing. In online finance, many people think they can make quick money by spending big on marketing to scale quickly. But if the investment is burned up, your profit model doesn’t work and, at the same time, NPL levels are going up, what can you do? If NPL goes up, platforms that make implicit guarantees will blow up.

Knowledge at Wharton: Can we conclude that online finance will not grow explosively like e-commerce companies?

Zhou: There is one feature of the finance industry — the one that grows the fastest, will also collapse the fastest. Anyone who hasn’t experienced an economic downturn can never say he has a successful risk-control model. What is the use of having a rapidly growing business without a reality check? This industry needs time to judge your capability. There are some companies in the market who have only been here for a couple of years and it’s too early to say whether they are successful or not. Some of them will be surely be eliminated given enough time.

Knowledge at Wharton: Many P2P companies have opened offline stores and they look for investment from offline stores. How do you view this O2O (Online-2-Offline) model?

Zhou: If there are no synergy advantages, the costs for opening offline chain stores are actually very high. In the O2O model, if you use it to verify data, it makes sense for risk control. For example, property ownership certificate and equity certificates — you can’t verify them based on one photo online. You have to do some verification procedures for pledges and mortgages through offline channels. However, if you open such channels to sell financial products, or to obtain capital, the cost will be very high.

If we open offline stores, and if those financial products managers have not been trained professionally, there might be big risks out there. Based on our experience in microfinance, it takes at least three years to train a qualified credit manager. CDB introduced microfinance technology from the World Bank in 2005 and many microcredit businesses in China have grown from CDB experience.

My view is that the future of P2P finance is still online. For selling financial products and obtaining capital, we will continue to adopt the online marketing model, rather than offline channels.

Knowledge at Wharton: You have been in P2P industry for three years. What do you think is the key issue to do well in this market, and what experiences do you want to share?

Zhou: Internet finance requires the practitioner to understand both the internet and finance. People from traditional banks tend to stress more on risk control, but less on user experience, especially on Internet-user experience.

On the other hand, it’s very important to have a mindset to resolve problems through data. If you don’t think this way, the costs will be high and the advantages of the internet will not be embodied. The most impressive thing for me is that this industry requires that practitioners should be highly qualified.