About

Washington D.C. based Bespoke Data and Analytics

Leveraging alternative (‘alt’) data and analytics while often mashing them up with traditional data and analytics, Zettacap provides bespoke services to improve insight and forecasting.  Out data, analytics and/or forecasts have been applied to financial markets, political elections, product demand, fundamental analysis, and macro-economic trends.

Highlights of some of the more interesting projects include:

Political Forecasting using Social Media Influence (SMI)

  • Only forecast of a Trump victory with 306 electoral votes, matching the actual result,
  • Foresaw and specifically called the coming embarrassment of professional forecasters and polls due to their poor analysis of the 2016 US Presidential Election,
  • First forecast of a Macron victory in the 2017 French Presidential Election,
  • Most accurate forecast of actual result in the French election.

Using SMI to Forecast the EURO

  • Forecasted Euro appreciation due to expectation of far-right political wave fizzling out in 2017 European elections in the Netherlands, France, and Germany,
  • Forecasted Euro appreciation to end due to expectation of far-right outperforming in 2018 Italian Election.

Bitcoin and Crypto-Currency Bubble

  • Published a report in February 2018 forecasting the end of crypto-currency’s bull market.  Hindsight shows that bitcoin peaked in December 2017 and many alt-coins peaked in January 2018.  Calling a wild almost decade long bubble to within about a month of its high is fairly good.   Additionally this call was Zettacap’s first bearish forecast for bitcoin therefore it was not part of a permo-bear campaign but a thoughtful analysis of the conditions.

On-Line Job Postings

  • Created a system that tracked on-line job postings of listed companies and then used assumed company-level job demand to improve fundamental investment analysis,
  • First to forecast layoffs in a major blue-chip stock (Microsoft/MSFT) due to a significant drop in the company’s job postings, research report released just before the company’s layoff announcement,
  • First to forecast weaker-than-expected Holiday Season for a retailer (Tiffany/TIF) due to drop in seasonal hiring.  Research report released significantly before company’s announcement confirming weaker-than-expected sales.  Company’s announcement resulted in a one-day 14% drop in share price.

Stock Price Extraction Sentiment Indicator

  • Created an innovative and unique machine-learning based approach of measuring market sentiment. Word-based sentiment is fraught with problems, not least of which is overfitting.  A better way is to focus on asset prices mentioned within communications like social media.  By identifying stock prices mentioned and then comparing these to actual prices, you can more cleanly measure sentiment around any asset,
  • Trends of sentiment derived from stock price extraction forecasted significant peaks in the S&P 500, oil and individual stocks.