Data Trends for Investment Professionals


WooTrader proves what startups can accomplish when they get the data they need

We recently spoke to Atanas Stoyanov, CEO of WooTraderabout his journey from a software programmer to FinTech founder. In 2007, Stoyanov sold his previous company, an Inc 500 firm that developed software optimization tools. He then turned his considerable programming talent to helping investors keep up with rapidly changing markets. He launched WooTrader, with the vision of becoming the only website you’ll need for asset management. With the recent release of a Wall Street-ready API, WooTrader’s client base is now expanding to hedge funds and investment banks.

You’ve built a number of successful companies, including a software testing company. How did you end up in the FinTech world?
Getting into financial markets was never the plan. I’m a programmer through-and-through. I built a company that tested software performance, regression bugs, and issues in software. At a young age I sold the company (now named Smart Bear), hired some highly-regarded investment managers to handle my assets, and retired. I was content to leave my investments to the professionals. Then, year after year, I saw my portfolio performing three or five points below the market indexes. It was frustrating.

I began to wonder if there was a better methodology. I wanted to provide something to personal investors that incorporated all influencers — social signals, fundamentals, technical factors. We’re bombarded by so much information, and it all plays a part in creating price movement.

What WooTrader does is very similar to my software testing tool in the sense that it tries to find bottlenecks in the system. With software, if we didn’t have the right data, we ended up optimizing something that wasn’t actually the bottleneck. This is very similar to market research. WooTrader tries to accurately find the bottlenecks in current market models.

Can you tell me more about how WooTrader works?
Weighted predictive analytics models have been used successfully for years in the military, sales, and urban planning. WooTrader applies this methodology to the rapidly changing markets.

The metrics that are moving the market are constantly changing, so the information you’re using to evaluate the market should change too. That’s why we developed our dynamic investment strategy. Each night, our system analyzes the day’s data and compares it with past data. Data that is clearly moving the market is given more weight in our calculations while less influential data has less of an impact on a stock’s ranking.

What would your advice to be to other programmers trying to create tools for the market?
Start with the basics. When I began programming thirty years ago, I could have worked in an advanced language, but I chose to work in assembly, just to be closer to the foundational problems.

In general, I try to start at the foundation so I can see how everything works. My advice, especially with data, is to start at the lower level. Right now, the data is not perfect. If you build models on top of bad data, they’ll crumble. Be as close to the data as possible.

Is there something that you wished you knew when you started?
To have faith that a problem will get solved, even if I can’t solve it at the moment. Today, I have three things that I have no idea how to solve, but I know I’ll be ok because I’ve been doing this for thirty years. I will find a way, because I always have. It might not be the greatest way, but that’s also ok. In one month I’ll improve it, and in two months I’ll improve it again. I’m comfortable being iterative.

Also, I used to assume that the industry had already achieved the highest level of investment expertise. I wish I hadn’t. It took me along time to realize that my programming experience gave me a unique perspective on working with financial models. Perhaps also, as an outsider, I was more comfortable questioning conventions.

As a programmer, I was trained to be macroscopic in my analysis and pragmatic in my approach. If something isn’t working, then it isn’t working. Valuation ratios alone aren’t driving the market. It’s no longer about finding a golden ratio or a super complicated model. If the model doesn’t change as quickly as the market, then all that highly sophisticated math is useless. Moreover, even the greatest quant model has a limited shelf life. It might work for two months, stop for three, and then start working again for two months.

What mistakes did you learn from the most?
I had no idea how complicated it was. I thought I would just apply predictive analytics models to a subset of the market — the fundamentals and the technical analysis. The more I got involved, the more I realized how many other influences there were. For option markets, as an example, there are social signals, news articles, data from small companies that you didn’t even know collected data. I had no idea how deep you had to go into the data to really get a picture of the whole market.

Five years ago, before Quandl, it was impossible for a bootstrap startup like mine to get their hands on this much diverse data, at this quality. The cost made the barrier to entry too high. I need to have access to any data that might be driving the markets because I can’t predict what factors will be influential at any given time. So I need access to all of it.

Quandl helped WooTrader get the data it needed in the way it needed it. Besides the lower price point, it was really crucial to have a single point of contact if I needed support with any of the databases. If I have any issues, I just send an email to Raymond and he takes care of it.

The more democratic licensing was also a big help. As a startup, it’s difficult to know at any given point exactly how many registered customers you have It was a relief having a fixed price within a range — I didn’t have to report every single additional user or constantly worry about having the correct license.

It’s a combination of those things that make it possible for a startup like us to compete against the big guys.

What does the future hold for WooTrader?
The next stage is portfolio and risk management — we have released the first iteration of this stage and will continue throughout the year. After portfolio and wealth management, I’d like to add other asset classes, including commodities, real estate and currencies. Long term, I see WooTrader as a one-stop solution to manage all your assets.

WooTrader is one of many startups powered by Quandl data. The cloud-hosted platform downloads new data every day from Quandl.

Offering the tools to become an investing professional, Wootrader enables users to research, buy and sell stock positions all from the same platform. For more information, visit:


Leave a Reply

Your email address will not be published. Required fields are marked *

  • Hi Scott,

    Thank you for the comments. Below are our system specs, just keep in mind that I am coming from a desktop/bative development background, and me being the single developer for initially developing and deploying the system many choices were made to streamline the development process and focus on the core functionality while trying to find the best languages/components for the rest (as every other system we needed user authentication, blogs, payment and subscription processing). Another imperative was to make the system as fast as possible, so I can delay deploying multiple servers.

    For the backend, we are using a mix of technologies, trying to maximize our (very) small dev team.
    1. Php (now updated to PHP7) for the web server (using wordpress for many of the basic functions, also php is pretty fast for serving web pages).
    2. nodejs for our API server (the fastest for a simple REST server, also thinking about the future to go with react native for a mobile app that will use the API server as well)
    3. and Python for the stats, financial calculations (a bit slow for serving web pages, also lacks something like wordpress to speed up the development process of the basic functionality).
    You can see one page in action that uses all 3 servers here :
    It uses the nodejs server for the stock symbols lookup, Python for the risk / returns metrics calculations and PHP/Wordpress for serving the html page.

    Front-end is basic jQuery and bootstrap (we wanted to have a responsive web site that works on mobile while finding the time and resources to develop a dedicated mobile app). Using several components – for the grids (the card view display is our own implementation), select2 for lookups, highcharts for charting and a couple more small components. I didn’t go with larger libraries from vendors like developers or teller mostly because they required loading the entire libraries, even components that I didn’t really need and were slowing down the pages.

    We are running the whole site on a single cloud server on Digital Ocean (they had the best documentation for deployment) and the db is currently MySQL.

    The APIs we are consuming are for downloading data, and integrating brokers – mostly REST, but some vendors are also still using ftp servers . The data is sometimes JSON, sometimes .csv – the industry is still finding its way to standardize on any formats. Even authentication is quite varied among vendors.

    Positioning – we would like to provide a product that gives good returns to our users and as a company to be open and transparent. No hidden fees, no tricks and gimmicks. Not sure what a position that is – there are so many fintech terms nowadays 🙂

    Thanks for the good wishes and I hope my post was helpful

  • Scott Shagory says:

    As someone who focuses on strategy and data it’s always interesting to see firms like WooTrader motivated by market data gaps (and personal frustration) to find a solution.

    A future piece on WooTrader’s infrastructure choices (DBs, APIs, front/backend) and the firm’s strategic positioning would be really nice. The fin tech space is so ripe for continued innovation.

    I wish the team at WooTrader the very best.

Fix This
Created with Sketch.