4 ms·

Show HN: I made a patent search tool designed for stock research

2 points by greatwave1 6y ago

greatwave1 6y ago
I was inspired to build this tool after reading about Ideanomics, a company that had a 700% stock price surge last month after announcing a pivot into electric vehicles. There are many other examples of this phenomenom, with one of the most notorious being Long Island Iced Tea’s shift to blockchain (https://cointelegraph.com/news/fbi-investigating-long-island-iced-teas-shift-to-blockchain https://cointelegraph.com/news/fbi-investigating-long-island...). Regardless of the legitimacy of these business transitions, I think it's useful for investors to know which companies are doing research into new technological areas. It’s difficult to use existing tools (like Google Patents) for this purpose, because the vast majority of patents are not registered to publicly-traded companies. Companies also often register patents under a wide variety of names. To fix this problem, I scraped patent data from the USPTO website (https://bulkdata.uspto.gov https://bulkdata.uspto.gov) and then used a text-matching algorithm to identify which patents were assigned to Russell 3000 companies (the 3,000 largest publicly-traded companies in the US). From there, it was easy to match each patent to a specific ticker. You can use this tool to search for a keyword, and find all recent patents by Russell 3000 companies (standardized by ticker) with an abstract containing the keyword.
- sawaruna 6y ago
  Can you give a use case for this, or how one could have anticipated the Ideanomics surge you mention? Searching something like 'electric vehicle' will show many recent patents by Tesla, Apple, Honda, et al. Is the idea that you would search for such a term and try to find some outlier company, assume their recent patents would reflect an entry into the market, and buy accordingly?
  - greatwave1 6y ago
    Exactly. It could also be used as a tool to screen for companies that will benefit from growth in certain markets. For example, searching for 'electric vehicle' will show several companies that have developed components which can be used in electric vehicles.
husseiny 6y ago
Feels like if this is going to be used as a stock screener, the search should start at the company level not a keyword. Or show a list of recent patents by Russell 3000 companies that don't match the previous x number of patents filed by that same company.

2 points by greatwave1 6y ago

I was inspired to build this tool after reading about Ideanomics, a company that had a 700% stock price surge last month after announcing a pivot into electric vehicles. There are many other examples of this phenomenom, with one of the most notorious being Long Island Iced Tea’s shift to blockchain (https://cointelegraph.com/news/fbi-investigating-long-island-iced-teas-shift-to-blockchain https://cointelegraph.com/news/fbi-investigating-long-island...). Regardless of the legitimacy of these business transitions, I think it's useful for investors to know which companies are doing research into new technological areas. It’s difficult to use existing tools (like Google Patents) for this purpose, because the vast majority of patents are not registered to publicly-traded companies. Companies also often register patents under a wide variety of names. To fix this problem, I scraped patent data from the USPTO website (https://bulkdata.uspto.gov https://bulkdata.uspto.gov) and then used a text-matching algorithm to identify which patents were assigned to Russell 3000 companies (the 3,000 largest publicly-traded companies in the US). From there, it was easy to match each patent to a specific ticker. You can use this tool to search for a keyword, and find all recent patents by Russell 3000 companies (standardized by ticker) with an abstract containing the keyword.

sawaruna 6y ago
Can you give a use case for this, or how one could have anticipated the Ideanomics surge you mention? Searching something like 'electric vehicle' will show many recent patents by Tesla, Apple, Honda, et al. Is the idea that you would search for such a term and try to find some outlier company, assume their recent patents would reflect an entry into the market, and buy accordingly?
- greatwave1 6y ago
  Exactly. It could also be used as a tool to screen for companies that will benefit from growth in certain markets. For example, searching for 'electric vehicle' will show several companies that have developed components which can be used in electric vehicles.

husseiny 6y ago

Feels like if this is going to be used as a stock screener, the search should start at the company level not a keyword. Or show a list of recent patents by Russell 3000 companies that don't match the previous x number of patents filed by that same company.