Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

A Dive Into The Lending Club Data

100mdeep.com

Post reply on HN
Print view View original

41–50 of 50 posts

Re: A Dive Into The Lending Club Data

#41

Post by ClementM »

akg_67 wrote: post #39
ClementM wrote: post #37

Earlier quoted context omitted.

A bit more courtesy would have been welcome. You sound very condescending. And I hope you talk to your clients in a different way! Let me address your methodology comments nonetheless, which are for the most part unfounded. * I don't have any finding about annual income. I don't think it is mentionned anywhere in my conclusions. * "delinquencies and public records has a big impact on returns as newer loans are not ag…

Sorry for coming across condescending. This was not my intention. I was just trying to guide you in the right direction as you came across someone who is just getting started with data analysis. Once again, I will stress, you need to reconsider your methodology if you want to learn.

I have been doing quantitative data analysis in the investment space for about 10 years.

Happy to hear/talk about methodology if you want to write/talk about it.

Re: A Dive Into The Lending Club Data

#42

Post by jldugger »

paulsutter wrote: post #40
radmuzom wrote: post #22

I build statistical models for banks which help assess the risk of a loan. Effectively, my models will get converted into the grades (A, B, C, D, etc.) mentioned in the article. The strategies (second chance, family guy, safe haven) are generally consistent with experiences from the portfolios of most financial institutions. However, I am skeptical (prove me wrong) of the statement in the article - "Lenders get a ret…

Banks haven't proven too skilled in estimating risk (see 2007). The average Lending Club lender is surely worse than the banks, but the best ones are surely better. Net it all out and hopefully we get a Darwinian marketplace for lenders, lenders who don't need acres of employees preparing powerpoints for each other in class A office space. Of course, the true measure of estimating risk is hidden until a crisis occurs…

> Banks haven't proven too skilled in estimating risk (see 2007).

Arguably, they demonstrated that you can accept a lot of risk and then if things go wrong the Fed will bail you and your creditors out. Not _quite_ the same thing as simply underestimating risk.

Re: A Dive Into The Lending Club Data

#43

Post by gtremper »

existencebox wrote: post #11
rocketcity wrote: post #6

This is great. I have been using LendingRobot lately. I will have to rethink some of my strategies based on what I see in this article.

For my own personal interest, can you speak at all on your experience using LendingRobot? (successes/misses, use duration, gains, etc)

I've been using it for about a month and its one of the better LendingClub auto-investors out there. By far the best interface, but the fee is a little steep, though the fee only takes effect after investing 10k though them. I've rolled my own auto-investor in the past and used the one on NSRPlatform(free for accounts Here's my shameless referral for lending robot https://www.lendingrobot.com/ref/YehMn307/ (you get 10k free, I get 5k more free)

Re: A Dive Into The Lending Club Data

#44

Post by existencebox »

gtremper wrote: post #43
existencebox wrote: post #11

Earlier quoted context omitted.

For my own personal interest, can you speak at all on your experience using LendingRobot? (successes/misses, use duration, gains, etc)

I've been using it for about a month and its one of the better LendingClub auto-investors out there. By far the best interface, but the fee is a little steep, though the fee only takes effect after investing 10k though them. I've rolled my own auto-investor in the past and used the one on NSRPlatform(free for accounts Here's my shameless referral for lending robot https://www.lendingrobot.com/ref/YehMn307/ (you get 1…

Thanks! Any recommendations for someone looking to get into this space, especially in terms of things you had wish you knew starting off? (I'm looking at this as a "long shot" diversification of a portion my savings, allocating a comparatively very small amount and going entirely hands off)

Re: A Dive Into The Lending Club Data

#45

Post by gtremper »

existencebox wrote: post #44
gtremper wrote: post #43

Earlier quoted context omitted.

I've been using it for about a month and its one of the better LendingClub auto-investors out there. By far the best interface, but the fee is a little steep, though the fee only takes effect after investing 10k though them. I've rolled my own auto-investor in the past and used the one on NSRPlatform(free for accounts Here's my shameless referral for lending robot https://www.lendingrobot.com/ref/YehMn307/ (you get 1…

Thanks! Any recommendations for someone looking to get into this space, especially in terms of things you had wish you knew starting off? (I'm looking at this as a "long shot" diversification of a portion my savings, allocating a comparatively very small amount and going entirely hands off)

I have all my LendingClub funds in a Roth IRA so I don't have to deal with any of the tax loss/gains stuff. I've heard the tax accounting can get tricky in normal accounts, so I highly recommend putting your LendingClub funds in a tax-advantaged account.

As for filtering and stuff, you can you do your own underwriting(loan risk analysis) with the data lendingclub provides[1], or use https://www.nsrplatform.com, which has a nice GUI tool to explore the data with you're own filters. LendingClub has a JSON api, so you can an order executer for yourself. (Here's the remnants of the one I was working on https://github.com/gtremper/LoanInvestor. P2P-Picks was a 3rd party underwriter that isn't available anymore). I've noticed that the D and E loans tend to be the best balance of risk and return

Also, be aware that you'll need to continuously buy new notes as payments come in to your account, otherwise you'll build up cash rather quickly. That's why these auto-investing services are so useful. Its best to buy only $25(the minimum) per loan so you can spread your risk among as many notes as possible.

[1] https://www.lendingclub.com/info/download-data.action

Re: A Dive Into The Lending Club Data

#47

Post by edelans »

ClementM wrote: post #18
edelans wrote: post #12

Earlier quoted context omitted.

Hi Clement, very neat dataviz ! How did you get the full history of all payments ? Is it available in the open ? What is 100mdeep btw ? Just a blogging site you are bootstraping or any plans to make a living out of it ? (We met a few months ago @ ToucanToco...)

Funny you mention ToucanToco. They helped me in debugging the viz.

Thanks for the replies, what about the source of the data ?

Re: A Dive Into The Lending Club Data

#48

Post by edelans »

ClementM wrote: post #17
edelans wrote: post #12

Earlier quoted context omitted.

Hi Clement, very neat dataviz ! How did you get the full history of all payments ? Is it available in the open ? What is 100mdeep btw ? Just a blogging site you are bootstraping or any plans to make a living out of it ? (We met a few months ago @ ToucanToco...)

100m is the depth human can free dive to. Without any aid. Without any fins. https://en.wikipedia.org/wiki/Constant_weight_apnea Tribute to people who go deep into things ;).

Nice image! I'm a big fan of Guillaume Nery, the french free diver (4 word records). Actually Constant weight divers are allowed to use fins (monofins most of the time), although they use it only to get to ~30m deep, after that point, the pressure is such that the volume of the body decreases and Archimedes thrust is no longer sufficient to compensate for the weight... so the diver can descend without any mouvement... It's very impressive to see (http://www.liveleak.com/view?i=d4a_1244462128).

Funny to read about this a few days after Guillaume Nery's accident: a line was set to the wrong depth causing the diverto dive to -139m/456ft instead of -129m/-423ft that he had announced the night before...

Re: A Dive Into The Lending Club Data

#50

Post by samstave »

techcreditcard wrote: post #30
radmuzom wrote: post #22

I build statistical models for banks which help assess the risk of a loan. Effectively, my models will get converted into the grades (A, B, C, D, etc.) mentioned in the article. The strategies (second chance, family guy, safe haven) are generally consistent with experiences from the portfolios of most financial institutions. However, I am skeptical (prove me wrong) of the statement in the article - "Lenders get a ret…

Your background sounds interesting! I'm working on bringing a credit card to the subprime market. Would be great to connect with you.

I want to talk to you -- how do I get in touch... sam @ sstave . com
Post reply on HN