Challenge

A successful real estate entrepreneur with a healthy interest in the stock market had figured out the perfect set of strategies, that could be run against any number of trading brokerages, to perform orders in low volume, at a high frequency, to gain an incredible ROI with minimal risk. There was only one problem.

This behavior would need to be conducted automatically at a higher rate of action than any human, or team of humans, could perform, against a larger volume of data than anyone could read. Combined with the need to be adjusted on the fly, have smart alerting and failsafe strategies built in, and ultimately make money while the client slept, he sought out a technical solution.

Stumbling upon our first official office building doors, he found a few developers working hard on a Saturday, and was immediately hooked by the idea of having mOOkstr build his vision.

‍

Constraints

Despite having a wealth of knowledge in the Real Estate industry, the client had only theories and research to base his findings on in the Stock Exchange, which meant that whatever we built had to be easily configurable, but able to be changed in the blink of an eye. This versatile engine would need to house a constant stream of data, to include buy/sell orders, price information, and countless other indicators that would be ranked, sorted, and adjusted based on formulas the client could plug in at any moment. There also needed to be the ability to create multiple strategies of this nature, some in a "live" environment, directly against the NYSE (and Nasdaq), while others would remain in an isolated, risk-free environment for testing and analysis.

Additionally, the potential to lose money on anything such as high frequency day trading created the need to allow stop-loss alerts and quick pull outs if anything went haywire. This meant that our representation of the aforementioned order data would need to be shown in real time, as would the results of those trades.

‍

Solution

mOOkstr, having been no stranger to developing multi-threaded, high volume data ETL gadgets, and having recently started allocating research into deeper calculation principals, presented to client an idea of building a fully functioning platform with a Web.API HTML5/JS/CSS/SignalR frontend and .NET backend that included

• a real-time data ingestion process that could pull down live data from a 3rd party that provided all the metrics we needed;
• an analyzing engine that would apply the various strategies configured by the client to historical and real-time data to rank symbols and indicators, then create orders within the clients brokerage of choice;
• a secure web portal that allowed the client and his team to log in, create formulaic strategies, adjust calculations for strategies and indicators, monitor the incoming data and analyzing engine, and view the results of their trades;
• and the infrastructure to allow for future marketing of the platform as a SaaS tool that could be licensed to other potential traders who want to try their hand at running profitable high frequency trading calculations against the Market.

‍

‍

Results

With the client reviewing every step of the way, we were able to build all components and turn a sandboxed version of the platform on within a few months. The client had created several strategies on paper, and eagerly plugged them in, only to be met with disappointment as the numbers immediately started dipping into the red when the market opened. There were a few core issues in the design of these formulas, and several of them had been based on the incorrect assumption that the client would be able to sell at ask - this created a difference in opinion on whether the tool could ever become a fully viable product.

While we were able to successfully complete all requirements, both parties agreed to backburner going live with the app, as the automation of a flawed process was leading to flawed results. We are very proud of the work that we accomplished during the short timeline granted, and greatly appreciate the mountains of information we gathered along the way. Not only did our multi-threaded development skills improve as a direct result of this project; we also formed new theories, ideas, and even stronger team relationships along the way, turning the page on yet another highly valuable chapter in mOOkstr's storyline.