QuantConnect Brings Low-Code Principles to Quant Finance

QuantConnect, a crowd-sourced algo trading platform, is striving to bring low-code principles to quant finance.

The term “low-code,” which refers to an emerging area of software development in which people with limited programming knowledge can design and build custom apps using pre-configured modules and widgets, seems, at first, diametrically opposed to the quant world, which is coding-intensive by nature. But the programmers and developers working at buy- and sell-side shops are surrounded by teams of business professionals who fundamentally understand the concepts of quant finance, but lack the programming skills to translate that knowledge into alpha-yielding algorithms.

To help further democratize quantitative programming, the vendor wants to allow users to stitch together the various components that make up a quant trading strategy. 

Today, using the platform’s open-source algorithmic trading engine, LEAN, users can create, test, and sell trading algos in Python. A couple years ago, QuantConnect CEO Jared Broad and his team had the idea to divvy up the different concepts of quant finance and design pre-configured modules, now called Algorithm Frameworks, that represented the coding knowledge needed to write an algorithm pertaining to a concept area: universe selection, such as stock-picking; signal generation; portfolio construction; cash allocation; and trade execution.

With those modules now in place, QuantConnect will next allow users to stitch together the modules to allow a user to build a complete front-to-back quant trading strategy.

“Each of those categories of quant finance have entire teams of people working on them at banks. So you’ll have hundreds of people at a bank doing portfolio management and hundreds of people doing signal generation,” Broad says. “We broke the algorithms into those, and now we’re building a way that you can clip them together in QuantConnect in a low-code way.”

The vendor hopes to release this functionality before the end of the year.

The Unknown Unknowns

In the meantime, the quants who make up the platform’s near 120,000-strong community are maintaining a theme of uncertainty, which has been an undercurrent in many of the algos contributed to QuantConnect’s Alpha Streams marketplace, a network of funds and institutions who compete to license algorithms from developers around the world.

It crept in at the start of the year as the coronavirus spread within China and then beyond its borders, then surged and stabilized in March alongside the US stock market. And as the election in November looms nearer, volatility fears are rising again, along with risk control measures. Some programmers are notably using the Volatility Index (VIX) for position sizing or as a cut-off signal, which would halt the user’s trading and get them out of the market if it rapidly becomes volatile.

Just as this summer spurred an uptick in deploying nowcasting strategies in the marketplace—nowcasting refers to a portmanteau of “now” and “forecasting” that uses unstructured and structured data to predict possible scenarios happening at present or in the very near future—Broad expects to see an uptick in gray-box trading, which involves maintaining human oversight of the algo as it works, and possibly intervention before it places trades.

In QuantConnect, that looks like programmers building in notifications so that when the algo wants to place a trade, the user is alerted via text or email to review the idea, and they can choose whether to take it to a brokerage.

“It’s a bit of a strange time,” Broad says.

In July, QuantConnect announced it was overhauling its subscription model to boost institutional membership on the platform. The new subscription framework, QuantConnect Organizations, allows individuals and trading firms to better deploy their resources via a pay-per-seat and pay-per-service model that acts as “AWS for quant,” Broad says. Shortly after the restructuring, the platform lowered its minimum AUM threshold looking to license algorithms from $100 million to $10 million, meant to facilitate the growth of emerging funds and managers.

Broad says the moves led to user-base growth between 10% and 20% in August and September and that more investment banks, startup quant funds, and established hedge funds have come onboard.

The vendor is also actively seeking out partnerships with other organizations involved in the quant lifecycle, such as data vendors and quant compliance companies.