Web Scraping: Another Way to Work with Data
Web Scraping: Another Way to Work with Data: Back in December, Nasdaq stated that they were planning on securing Quandle which is a firm specializing in alternative data. It was at this point that alternative data became normalized. Indeed, in the beginning, it had mostly only been used by hedge funds that were advanced and quantitative as they had the skills and assets necessary to collect the typically unstructured data and then implement them within their investment configurations. As we mentioned, though, alternative data is becoming more prevalent with 50% of institutional investors claiming that they are thinking about increasing its use in the future. The most popular way to collect data today is by scraping it online.
Web Scraping: Another Way to Work with Data
Online Observations
Just as the name suggests, data that is scraped online is collected from websites available publically. Those firms that specialize in web-scraping create programs that enter sites that they are going after, gather, and then hold onto all of that data that had been collected during specific durations of time. Sometimes, vendors, such as Quandl, Thinknum, Savvr, and Yipit, utilize public APIs in order to get information straight from the pages without the need to actually go onto the site. Indeed, there is so much useful data for investors to use from the four billion webpages as well as the 1.2 million terabytes available. Some of the kinds of data that is scraped online includes:
- Listings of Jobs: A business that wants to hire more staff as the firm is expanding.
- Retail Data Located Online: Those items that have good ratings imply that their sales are good on the side of the product manufacturers. However, items with deep discounts could imply sales that are poor.
- Ratings of Company: Websites such as Glassdoor let staff rate the firms. Therefore, should ratings get better combined with a greater number of job listings, it shows that the company may be seeing positive growth.
Case: The Benefit of Information
On the 26th of December, Amazon had claimed that they had a record-breaking holiday season with statistics such as:
- The Echo Dot being the top item that was sold in every Amazon category
- Both the Echo Dot and the Fire TV Stick were best-sellers on Amazon
- Many Prime members used the voice feature on Alexa to purchase gifts, Amazon devices, as well as household basics that are used daily
However, their success had not been input into the stock price, which had increased by 23% since they made the declaration; at the same time, the S&P 500 at a mostly horizontal level. Those, however, who did use alternative data would not have been shocked by Amazon’s success. For example, Thinknum had uncovered data showing the rise in Amazon item sales on Best Buy’s site during the holiday season which began on Black Friday. From that site, it was also possible to visualize how both the Echo Dot and the Fire TV stick would be the best sellers within their department. What’s more is that the success of their AI powered, digital assistant that is controlled by voice, also known as Alexa, could have been foreseen from the information in “Amazon Alexa” job postings on the corporate site. Indeed, throughout 2018, the amount of positions available surrounding the Alexa project saw an increase in 53% as the numbers rose from 500 to 750.
All of this being said, alternative web-scraped data does not provide us with all of the answers. While it may be able to pinpoint which items are doing well, it is not able to put a number on the influence that occurred on a big and varied firm such as Amazon. For instance, even though the rise in job listings did help signify growth, it could have also been the result in a boost in cost, which helped influence profitability. Nevertheless, it is no secret that alternative data like data that is scraped online can offer crucial new data about a firm’s venture and perspective. At times, that information is valuable by itself, but it’s more likely that real value comes from the combination of other derivations of information in analysis that are traditional, alternative, as well as qualitative.
from Chop News http://bit.ly/2LmkO3U
Comments
Post a Comment