A conversation with Zhang Yi, CEO of Quancept, a Fintech startup that develops quantitative and Artificial Intelligence based solutions to invest in financial markets. We talk about his shift from equity trading in NY to starting his own company in China, the overall Chinese Fintech market, and the role of Artificial Intelligence in Fintech services, among others.
I used to work on Wall Street in the 90s, JP Morgan then was Lehman Brothers, in several areas: trading system for credit derivatives, global equity portfolio trading, options among others.
I later started trading equities with different firms in New York. I traded using a custom built quantitative trading system for 8 years. During that time, I developed several systems to facilitate trading. The quantitative trading approach generated well-above-market-average returns, with lower risk. It was appealing for me and for the firms I worked in.
After 2010, I saw a clear opportunity to come to China and adapt the previous models to the market here. With a relatively mature product, the main challenge was to adapt to the local market and to connect to the market data sources and brokerages.
Today we are offering a SaaS product to meet institutions’ needs (Hedge Funds, Brokerages, Pension Funds).
"With a relatively mature product, the main challenge was to adapt to the local market and to connect to the market data sources and brokerages".
I started the company last year. Two months later we were selected for a 6 months' program into Microsoft Ventures Accelerator as one of the 20 applicants selected from over 1000 applicants. They provided us huge amount of resources to help us develop.
We started developing institutional business right away. February this year, we got our first major customer, Guosen Securities. We deployed our solution in a number of their business lines. It was a real breakthrough.
We drive their agency business. To a stock brokerage, the agency commission is their bread and butter. Since last year, there was a dramatic decline in the commission revenue among all brokerages.
So naturally, the priority is to boost revenue. They were looking at innovative technology. They saw our technology as a driver to their revenue.
We drive institutions’ revenue with a more precise investment, trading and timing.
First: Speed. Our clients can build models on our platform, back test them and bring them to production right away. All this process now takes less time, less than an hour, from a rough idea to production in the market. These kind of processes can typically take over a year or longer to achieve. On Quancept platform, it is right away.
Second: Usability. The essence of what we provide is that our users don’t need to write program. They can express their ideas very easily, using simple formulas. We call it: complexity simplified.
We collect a huge amount of data from the exchanges and use historical market data, fundamental data including earning, PE ratios, market caps and many factors that you can add into the model.
Let’s take the Brexit as an example. If you expect the Brexit to happen next week, and you expect the market to go in a certain way, to react in a certain way. With that in mind, you can build a model to try to capture some market moves.
Usually, you don’t have enough time to build all the system, model it, and execute it. On Quancept, you can model the idea in 15 minutes. Perform a simulation right after that.
And you can launch into production on the same day. Time to market is critical.
"Let’s take the Brexit as an example. If you expect the Brexit to happen next week, and you expect the market to go in a certain way, to react in a certain way. With that in mind, you can build a model to try to capture some market moves".
There are three areas where Fintech participants are currently focused on: payment technology, credit risk and capital or wealth management. We are in the third category.
On this specific area, every country has different financial systems. USA’s market is very mature. In China, we are at the beginning stage, still a long way to go.
From a regulation point of view, you can expect different regulations in different countries because participants’ behaviors are different. The measures to allow or ensure stability in each part of the world are different.
Back in 2008, in the US, there were measures put in place to curtail shorting stocks for example. They achieved a specific purpose. Today, other measures exist, different from country to country, from situation to situation.
I would not use the word disrupt. Has been used so much by the startup community that it became a cliché. But I do see where we can add significant value to stakeholders. We like to see ourselves adding value to existing infrastructure.
Big firms are already in this industry. PingAn bank for instance. It is not easy to disrupt anything, but you can add value to people’s current situation.
"It is not easy to disrupt anything, but you can add value to people’s current situation".
In capital management, one thing we want to explore, and I start to see a trend now, is empowering small time traders, or grassroots traders, to take their skills to another level, and potentially to play in the capital management arena.
So this could promote a quite different model from the traditional capital management approach, where a typical Hedge Fund is headed by 2 or 3 people. Now you could have over a 1000 grass-root traders applying their skills into capital management. It can easily bring new talents and skills in the field at a much larger scale.
In wealth management, traditionally we see a more passive style of asset allocation and investment strategy. I do believe that a more active management style will become more popular in this landscape.
At Quancept, we approach the problem from a quite unique angle compared to other players. We want to enable much broader audience to participate by significantly lowering entry barrier.
We started using machine learning two years ago. Our research uses machine learning to build models. We have team members previously working with Intel’s AI team. It’s highly proprietary. Trading is an area that’s still unproven using AI techniques. But we are very hopeful.
Shanghai is the leader from my perspective, in the mainland. HongKong also has a vibrant Fintech culture.
I don’t think it is easy to attract money even if you are a Fintech, especially now. However, we are able to attract interest from a few well-known VCs.
I started the company last year. We are 15 colleagues now.
We like the culture here. A vibrant entrepreneur culture. Feels like home.