A couple of years ago I created a website and an algorithm called Flux Robot that would automatically invest funds in a user’s Lending Club account to maximize their potential return of investment (ROI). In a lot of ways, I succeeded. The average ROI for the very few accounts in Flux Robot was 7.92%, which is much higher than Lending Club’s ROI, and more than most people make in any stock market. The site, however, failed and will be shut down.
I could point out a lot of factors on why the site failed, but I wouldn’t know where to even begin to determine which one was the biggest factor. Before that, let me elaborate on Flux Robot a little.
About 10 years ago, one of my best friends introduced me to Lending Club and Prosper. Both of these are wonderful programs that enable you to invest in other people’s loans. On the other side, it enables people to get loans through crowd-sourcing, rather than a bank so they could save on interest and fees. These sites offer a mutual benefit for both investors and a person who requires a loan.
After a few years, I started to realize the return on my investments were much better than most mutual funds, my retirement funds, and other investments. Lending Club offers a few things that I found beneficial: all of their loan data was available to download, and an API, which allowed me to directly connect to their service.
Lending Club offers an “automatically invest” option where the funds will be automatically invested in a portfolio that you create allowing you to compound interest. However, it doesn’t automatically invest until you had x amount of dollars. I’m not confident in the figure, but you were losing return potential while waiting for that money to invest. I decided to solve this problem.
After downloading over a decade of data, I determined various factors such as how many credit inquiries a person had, their FICO score, the state they lived in, and even their job title affected how likely they were to pay-off their loan. Using this data, I created my own custom algorithm to invest and maximize my potential. Using buckets of $25, the software would start investing as soon as $25 was available into the best loans best on my algorithm.
One day I thought to myself, what if I allow other people to do this too? I launched a server and a website that allowed other people to create accounts on my website, and my algorithm would automatically invest for them. All they had to do was enter their Lending Club account number. Additionally, I gave them a couple of choices for algorithms: high-risk, medium-risk, and low-risk. The few users I had were all high-risk and the ROI was 7.92%. I don’t have a calculation for medium and low-risk because they weren’t used, but based on my projections medium-risk and low-risk were both roughly about 6%.
To keep the server running, I decided to introduce a quarterly-pay system where users would pay 0.0875% of their Lending Club account balance using PayPal. I never received a dime because the users I had were all beta testers and I promised them a free account for life. Thus, why I decided to close the site.
The Factors of why it didn’t succeed:
- Directly after the public release of my site, Lending Club was hit with some brutal news regarding fraud. I don’t want to get into too many details, but you can read about it following this point. Basically, the website lost a lot of credibility due to some internal practices and a lot of my potential beta testers decided to pull all their money out of Lending Club
- Competition. There was another site I was unaware of that existed and has good results. My results were slightly better and my fees were cheaper, but they had a better reputation. It was difficult to grow in a market with low interest and one major competitor – it would be like a local internet company trying to compete with Comcast. When was the last time you heard something like that succeeding?
- Poor marketing. I would be the first to admit marketing is not a strong point. I had several people review my website and they said it was beautiful and easy to understand. I spent a few hundred dollars on Google Adsense to generate clicks, but none of them converted to sign-ups, despite the fact there is a one-month free trial.
- Lack of interest. I don’t think people were too interested in having another system provide returns when the benefit wasn’t enough to account for the costs. I don’t know how much income the main competitor generates, but my anecdotal evidence shows there wasn’t a lot of interest for something like this in peer-to-peer investing forums.
I’m not sure what I’m going to do with the algorithm. I’ll keep it running to generate income for me, but unless people are really interested in having something like this, Flux Robot is shutdown. I’ll also sell the software along with the source code to anyone interested in owning it.
- Visual Studio C#