[twigster]

Per their web site:
Quote

[lascott]

Really too bad they are not at least keeping the Percentile Rank Calculator page up.

[turing]

Yeah, we decided it was time to close and focus on other projects/businesses.

We can't maintain the percentile rank calculator anymore, but here is the info for anyone who wants to make it themselves:

Login:
Send POST request to this url with your email address and password
LoginUrl: https://www.lendingclub.com/account/login.action
login_email=your email address
login_password=your password

Get data:
Send a POST request to this url with the following data points
DataURL: https://www.lendingclub.com/account/benchmarkScatterChart.action
adjOnOff=true
myWeightedAvgIntRate=your weighted average return
similarPercent=1.0
minNotes=100
portfolioConcentration=100

You will get back a JSON file with a bunch of age and anar data points.

Hopefully someone else can recreate this program using the above info.  It has been a fun project to work on, but now is the time to end it.

[jheizer]

I assume you just used your own log in info for every request.  Did LC every say anything about the load from this?

[TravelingPennies]

Yep, I used my own login info.  I never got any complaints from LC about it.

[TravelingPennies]

Well I have an initial version available here http://lc.geekminute.com/Home/ANARCompare  It seems to be working pretty good except the Similar Age and WAIR Accounts number I am getting seems high.  But with out being able to compare directly to current data from the old page I can't say for certain.  Just guestimating based on old screenshots people have posted.  Inclusive vs exclusive bounds, etc mess things up and ANARs that jump all around each day doesn't help things.

[TravelingPennies]

good work!  I worked up a simple php script, the results are very close but not exact (within +- 2%);
probably to do with inclusion on the boundaries.
here is a sample.
--input:
adjusted net annualized return 10.2%
weighted average interest rate 13.96%
weighted average age of portfolio 19.4
--output:
P2PQ % vs ALL   86.05 vs 87.58 <my script vs on-line>
P2PQ % vs Similar  94.41 vs 95.52
P2PQ % vs Similar & WAIR  96.12 vs 96.65

[TravelingPennies]

Thanks for checking it out too.  Before I was being pretty inclusive on the bounds as it seemed like that gave the closest results but yeah they were a bit high overall.  I just changed most of them to be exclusive and it seems to have split the difference between my previous and yours, except the last one matches exactly.  Without knowing exactly what they were doing I think that's the closest I'll be able to get.

[TravelingPennies]

Glad you were able to replicate the percentile calculator in fairly short order.

[TravelingPennies]

Thanks and thanks for providing yours previously.  I'll try to improve on things, add the graphs, etc as time goes on.  I've got a 1 week old at home so free time is going to dry up pretty quickly once they stop sleeping so much each day.

[TravelingPennies]

from: jheizer on November 16, 2015, 01:30:02 PM

[TravelingPennies]

Thank you.  This is #2 so I know I am in that calm before the storm.

[Rob L]

from: turing on November 16, 2015, 01:35:50 PM