An Alternative Take On Stock Research

The fintech industry is split between institutional systems with many decades of in-depth data,  and innovative startups with unique datasets.

Both have their strengths. Terminal/dashboard type systems (like Benzinga Pro) aggregate a lot of financial information into one place in an easy to understand way.

Financial startups meanwhile provide unique APIs (or automated feeds) that cover alternative data sets. Some analyze Twitter and StockTwits tweets for sentiment, while others cover analyst price targets and institutional ratings, just to name a few.

FinanceBoards builds a bridge to link the best of both worlds. It combines these unique data sets, which are cheaper than ever, along with the array of financial data that terminals offer and integrates them into an easy to use dashboard platform.

FinanceBoards is made up of “widgets” that cover company fundamentals, technicals, social and analyst estimates, and more.

These widgets can be added to dashboards and resized and moved around. The coolest thing about this is that users can create their own widgets with the APIs that they have access to.

Here is what a typical FinanceBoards dashboard might look like followed by a dashboard filled with widgets that have Benzinga data.

Making your own dashboard, adding relevant widgets to it, and integrating APIs to create your own widgets is all free.

The Predictive Analytics Models

Additionally, FinanceBoards contains a stock ranking algorithm. This ranking is created with a predictive analytics model, which is a branch of analytics that is used to make predictions about unknown future events. It has previously been used to predict missile trajectory, but never in stock picking until now.

The model considers over 400 different stock screeners (including screeners from company fundamentals like the P/E ratio, technical screeners like the Chaikin Money Flow, Options Market screeners like Put/Call Volume Ratios, Sentiment screeners like Twitter and StockTwits Bull/ Bear Sentiment, Analyst Estimates, and more) and check the top stocks picked by each screener every day.

If those stocks outperform the S&P 500, then that screener gets a weight from 1-100. Every screener is measured every day, and then the weighted screeners are used to create a ranking of over 4000 stocks.

Below is a graphic that shows the measured screeners on the outside (with their respective weights in parentheses), and how their weights flow and create the general ranking-

We can see that at the moment, all of the categories have at least some weighting to them, with the Company Fundamentals category having the most weight. This is definitely not always the case, though.

We also provide a chart that shows the time periods from January 2015 to the present day when each category had a weight.

What’s really interesting is that not a single category has been consistently able to pick stocks that outperform the S&P 500. Each has had on and off performance.

What’s unique about this system is that it adapts to these fluctuations due to its dynamic nature, because it changes its criteria depending on what is performing best. A Peter Lynch strategy would have constant criteria (such as picking stocks with a PEG ratio less than 1 and a market cap less than $5 billion). What it assumes is that a PEG ratio will always have an impact on the market. These models show that this isn’t the case, because company fundamentals don’t always drive the markets.

Financeboards simplifies investing to the point where users don’t have to know what a P/E ratio or any other investing metric even is, because all they need to see is the end result: the ranking.

We see that since January 2015, the top 10 stocks picked were, on average, up 0.96%, while the S&P was up only 0.31%. Our model has done a great job of outperforming its benchmark.

With FinanceBoards and it’s ability to let users integrate any third-party data the user has access to, and its outperforming predictive analytics driven stock rankings, we really think we have made something special. As of a few days ago, we’re in open beta testing, and welcome feedback of any sort.