Andy Kershner built a trading firm around the tools he wanted

Andy Kershner began with $5,000, a friend and a library. By the end of their first summer trading options, the money had become $1,500. Finance has many ways of describing a promising start. This was an expensive introduction. Yet the pair saw a practical problem worth pursuing: getting the prices they could see on a screen. Long before Kershner put his name on a trading firm, access had become personal.

Today, the founder, CEO and chairman of Kershner Trading Group works from Austin, Texas. The business supplies capital, software and support to traders. Its history reaches back through ProTrader Securities, the arrival of electronic trading and the gradual addition of automation to work once done entirely by hand. Kershner's education was in geology at Vanderbilt University. He found another profession in which an apparently solid position can shift beneath one's feet.

A company kept, a platform bought back

Kershner Trading Group dates its present business to 2001, after ProTrader's sale to Instinet. Its longer trading history stretches to 1993. The distinction matters. This was a business carried forward through changes of ownership and technology, with people and accumulated experience surviving the transactions. A founder's story can look suspiciously tidy when every chapter begins with a new company name.

The particularly revealing transaction came later. In October 2004, Kershner bought technology back from Instinet and acquired its technology division. He had retained a proprietary trading business under the name Zone Trading. In 2006, that business announced its change of name to Kershner Trading. Selling the platform had been one chapter; deciding to own the tools again supplied the next.

The sequence gives his career a useful shape. Owning a trading business meant more than assembling people with opinions about stocks. It meant deciding which parts of the working environment deserved to be built and maintained inside the firm. A platform can be sold. The need to improve it tends to stay at the desk.

Software with a seat on the trading floor

The firm's execution platform, Gr8trade, was launched in 1998. Its continuing development places traders close to the people who build their tools. That arrangement is central to understanding Kershner's enterprise. A request from the floor can become a feature, and a feature can become part of a trader's daily routine. Software here has a decidedly practical social life.

Gr8trade accommodates discretionary trading, automated trading and the territory between them. A trader can use scripts for alerts or analysis without handing over every decision. The same script can move through backtesting, forward testing and live trading. The firm also describes tools for examining trades afterward, bringing manual activity and models into a common review process.

There is something appealingly unglamorous about that last step. The order has been sent, the excitement has subsided, and someone must inspect what actually happened. Kershner's business makes room for that work in its technology. Its platform is reserved for its traders rather than sold or licensed as a general commercial product. The tools are part of the employment proposition, with improvements informed by the people using them.

The neighbour with a useful filter

By August 2009, Kershner was sharing trading-filter ideas with Mike Bellafiore at SMB. A developing SMB trader, Joe P, built filters suited to that firm's approach. Bellafiore thanked Kershner publicly and offered help to his traders in return. It is a small episode, and all the more useful for that: cooperation becomes visible in a specific piece of work.

Trading firms compete for opportunities, people and capital. Here were people exchanging methods for spotting opportunities. The relationship later took a more formal shape. Bellafiore and fellow SMB co-founder Steve Spencer joined Kershner Trading in 2013 as partners running its New York office. Austin and Midtown Manhattan became connected through a shared business as well as conversations between traders.

For Kershner, the connection fits the firm's stated belief that knowledge and ideas should circulate. The practical test of such a belief is whether anybody benefits from another person's experience. A filter built at a neighbouring firm is a modest but persuasive example. Corporate values usually arrive in large lettering. This one arrived as something a trader could use.

An apprenticeship that takes its time

Kershner's business also puts a lengthy interval between a successful application and independence. Its published route for new traders starts with screening, a test, a questionnaire and interviews. Basic training takes about six weeks. An apprenticeship with a senior trader follows, with the journey toward independence described as lasting 18 months to two years.

The firm says 80 percent of candidates do not make it through that process. That is its own description of its programme, rather than an industry-wide measurement. Even so, it gives the invitation to join a useful degree of seriousness. Performance determines compensation, and the learning curve is steep. A collection of screens may be installed quickly; a dependable trader takes longer.

There is patience in the structure, alongside a demanding selection process. Terminology, software and basic strategies can be taught in a classroom. The apprenticeship then places a developing trader beside someone with experience. Kershner has built an organisation that asks for persistence while admitting, quite plainly, that persistence will not carry every candidate through. The recruitment pitch comes with arithmetic.

A moat, with an entrance

In 2017, Kershner described three obstacles facing aspiring quantitative traders: data, capital and the knowledge to put them to work. His memorable phrase was “the moat is access to the data”. The image suits finance rather well. It suggests a handsome institution with rather fewer bridges than applicants.

His argument extended beyond providing an API. Traders needed a complete route through research, testing and live execution. Access to one component would leave the rest of the problem waiting. The ambition was to let more people participate in quantitative trading by connecting those pieces, drawing on experience accumulated inside the proprietary business.

CloudQuant gave that argument a concrete example in August 2017. The trading-strategy incubator launched its crowd research platform with a $15 million capital allocation to an algorithm created by an external researcher. The researcher used free simulation and Python data-science tools, then licensed the strategy for a share of monthly net trading profits.

The amount was trading capital allocated to a strategy. For Kershner's wider project, the interesting movement was from an outside person's research into institutional execution. Morgan Slade was CloudQuant's CEO at that launch. The technology drew on two decades of development within its parent, Kershner Trading Group. A toolset developed for the firm's own work had acquired another audience.

The founder leaves his mistakes in the picture

There is a welcome roughness to Kershner's account of himself. In his 2017 conversation about trading, he discussed holding positions too long and giving back gains. He described reviewing mistakes and keeping a journal. The same capacity to tolerate an adverse position could help him stay with a winner or keep him in a trade that needed an exit.

That admission gives the technology and training a human purpose. Rules, review and better tools have work to do even for the person who owns the business. Success does not grant a permanent exemption from one's habits. A trading record can contain both competence and a familiar error, sometimes at an inconveniently close distance.

When he joined a Bear Bull Traders town hall in February 2021, the advice remembered afterward was emphatic: use stops. Andrew Aziz recalled Kershner returning to that instruction three times. Amid the attention surrounding GameStop, Robinhood and newly enthusiastic retail traders, it was a remarkably plain message. There are more theatrical things a guest can say. This one had the advantage of being useful.

“I love talking to traders and helping traders get where they wanna go”

Andy Kershner, 2017

Autonomy brings responsibilities

The firm's quantitative offering makes several specific promises: strategists can retain intellectual property, work with existing setups and enjoy flexibility over location. Capital and infrastructure need not require abandoning everything a researcher already uses. These terms give substance to Kershner's interest in widening participation. Independence is easier to discuss when the ownership of the work is clear.

The company record also includes a regulatory settlement. In May 2024, Kershner Trading Americas, a subsidiary of Kershner Trading Group, settled SEC charges over purchases in 23 public offerings following restricted-period short sales. Without admitting or denying the findings, it agreed to pay approximately $1.5 million in disgorgement, interest and a civil penalty. The respondent was the company; the proceeding did not name Andy Kershner as a respondent.

The responsibilities of building a trading organisation extend across its desks, technology and compliance arrangements. They remain part of the story alongside its recruitment promises and research ambitions. Giving people capital and autonomy creates an institution with obligations as well as opportunities.

Kershner's public professional profile continues to identify him with Kershner Trading and CloudQuant. Across those businesses, the recurring interest is practical: put tools, money and experienced people within reach of someone who can use them. The library experiment supplied an early encounter with the difficulty of access. His subsequent career has involved building a place where the next trader has more to work with.