Post by rppatel79 on Sept 18, 2016 15:02:41 GMT
Hi All,
Over the last 7+ years, I have used Lending Club and/or Prosper. Initially, I intuitively invested in loans hoping for good return. After some mild success (about ~9% ROI), I wanted to formalize my strategy. This became particularly important as my P2P accounts became a large component of my overall wealth. So, I downloaded Lending Club's historical notes, data mined the population, and confirmed my original strategy was effective (or more accurately, I confirmed my strategy was effective). I then built filters (within Lending Club and Prosper) that found active notes that met the criterion of my strategy. For the next year, I periodically logged into Lending Club and Prosper executed a filter and invested in notes. There were two problems with this approach:
1. I saw myself choosing not to invest in loans because of loan characteristics. This intuitive judgement hurt my ROI.
2. This was a very time consuming processes. I spend almost an 1 hour per day using P2P sites.
So, as Lending Club and Prosper built APIs, I build technology to leverage them! And recently, I have put the code on GitHub as a public project.
github.com/rppatel79/p2p
The code will:
1. Periodically log into Lending Club and/or Prosper
2. Database all the loans that were issued since the last execution
3. Find all the loans that met the criterion of a strategy
4. Purchase those loans
The code is authored in Java and each 'strategy' written in Javascript. The strategies are intended to be simple to write/extend.
Although not require, I have deployed the code in Amazon Web Services.
Feel free to check out the GitHub repository, ask questions, and/or make suggests.
Over the last 7+ years, I have used Lending Club and/or Prosper. Initially, I intuitively invested in loans hoping for good return. After some mild success (about ~9% ROI), I wanted to formalize my strategy. This became particularly important as my P2P accounts became a large component of my overall wealth. So, I downloaded Lending Club's historical notes, data mined the population, and confirmed my original strategy was effective (or more accurately, I confirmed my strategy was effective). I then built filters (within Lending Club and Prosper) that found active notes that met the criterion of my strategy. For the next year, I periodically logged into Lending Club and Prosper executed a filter and invested in notes. There were two problems with this approach:
1. I saw myself choosing not to invest in loans because of loan characteristics. This intuitive judgement hurt my ROI.
2. This was a very time consuming processes. I spend almost an 1 hour per day using P2P sites.
So, as Lending Club and Prosper built APIs, I build technology to leverage them! And recently, I have put the code on GitHub as a public project.
github.com/rppatel79/p2p
The code will:
1. Periodically log into Lending Club and/or Prosper
2. Database all the loans that were issued since the last execution
3. Find all the loans that met the criterion of a strategy
4. Purchase those loans
The code is authored in Java and each 'strategy' written in Javascript. The strategies are intended to be simple to write/extend.
Although not require, I have deployed the code in Amazon Web Services.
Feel free to check out the GitHub repository, ask questions, and/or make suggests.
