Using Mashups for Investment Research


New York – Mashups are web application that combines multiple data sources to form new kinds of information. Some of the most successful mashups have involved overlaying information from multiple sources on top of maps. Consumer-oriented sites like Gridskipper and Yelp combine restaurant reviews and guides to nightlife with Google Maps, providing travelers with an informed geographic guide to the places around them. Kayak is a powerful travel site that pulls price quotes out of hundreds of websites and displays them in tabular format, locating them visually on a Google Map. Nestoria, Zillow, and Neighboroo allow prospective homebuyers and renters to find properties and pull up relevant information about the neighborhood. Although most of these mashups are relatively recent “Web 2.0” creations, one of the most successful mashups of all, in a sense, is a 10-year-old firm called PayPal, which combined two heretofore unrelated domains – credit card payment processing, and email – in order to create a new and ingenious product, allowing Visa/Mastercard payments to be sent directly to email addresses. The e-commerce universe as we know it would not exist without this mashup.

How can mashups be used for investment research? One can imagine a number of useful mashups for financial analysts:

  • An alternative to existing market research methods: a mash-up could crawl across multiple sites which collect user-submitted reviews (such as Amazon, Zagat, Yelp, or in order to get an overall picture of what customers are saying about a company’s products or services. These reviews could be combined with additional data collected by the review sites to break down the feedback by different product lines and regional markets.
  • A mashup that shows a map of a retailer’s existing locations and proposed expansions, with information about prevailing incomes, rents, crime rates, taxes, and competitors in the area surrounding each location, can help an analyst to better understand the retailer’s business prospects.
  • An investor might want to see a map that mashes up a publicly-traded company’s geographic locations with breaking news events. For instance, if the user types in a ticker, this application would pull up a map of the company’s production centers, and alert them to floods, hurricanes, forest fires, riots, political disruptions and other events that may be cause for an unexpected hit to sales or inventory. Going beyond existing physical locations, one could view an insurance company’s exposure to hurricane season on a map, or identify companies that stand to gain from reconstruction efforts.
  • A mashup could search for user-submitted photos on Facebook and MySpace profiles, and analyze them with picture recognition software (such as Riya) to identify brands of clothing or footwear that are gaining popularity – more connected (and influential) users might be weighted more heavily as an indicator of upcoming fashion trends.
  • Believers in the “magazine cover indicator” might want to track the frequency with which an investment theme is mentioned in blogs and media, as a tool for timing the top of the market. Mentions in more widely read publications would be given more weight for timing purposes.

In general, the idea is to take aggregated sources of data which are already available on the web, and to combine them in new and different ways to gain insight. The technology is now advanced enough to develop applications that could only have been imagined a few years ago. Recently, we came across a firm called MapInfo, which supplies businesses and financial services clients with extremely detailed and information-rich maps. Their application could be used for a number of sophisticated mashup-like applications. Another firm, Kapow, which specializes in building proprietary mashups on a turnkey basis, has also begun to make some inroads with financial institutions. Majestic Research is a useful source for some proprietary data feeds that could act as inputs, and tools like Connotate or FirstRain can be used to mine the web for data in ways that go well beyond the capabilities of standard web search engines.


About Author

Leave A Reply