Loan default rate by ZIP codes / state

[lascott]

I did not see anyone having commented about loan default rate by ZIP codes since ZIP# was added to the download file. Here is what is observed in my account. It's interest to see some clusters formed. The loans being analyzed are issued between 3 years and 6 months ago. Distribution among grades are: C: 12%, D: 55%, E:26%, etc. 75% were purchased after 1/2014. In the chart, "Current Score Adj" is for adjusting the score that would tighten the filter. "Lost Int rate" is estimated from $ value of total loss, earned interest, and avg interest rate. "Annualized ruin rate" estimates what number of notes were "writen-off", where total loss in number of notes is estimated from number of charged-offs, lates, and FolioFn sales. I intend to use "Annualized ruin rate" to evaluate my filter ignoring note size and improvement due to FolioFn sales. Thanks.

I have info from Interest Radar on mine by state.

[lascott]

I made up the formula to try to get something to sort by so a state with a larger number of notes but with lower loss would sort OK. Happy to adjust the formula based on anyones input.

This is my taxable account with notes in these statuses...
Started in March 2014 but am still purchasing a lot of notes so "Weighted Average Age of Portfolio: 5.3 mos"

http://i.imgur.com/4g4mifx.png


http://i.imgur.com/EYMoqrV.png


My data is from here. The state info gets updated quarterly.
http://i.imgur.com/1JOFtL3.png

[lascott]

PeerCube Lending Club performance - https://www.peercube.com/histperf/perfbyattr/addr_state

http://i.imgur.com/2EhM5YC.png


http://i.imgur.com/O3QQ8rN.png

[craigwbryant]

SO how does LC have data listed for loans in MD even though MD doesn't allow P2P?  One or two, I'd say "ehh, someone cheated", 12,000 loans, which is right at the average for the data listed makes me ask questions.

[lascott]

SO how does LC have data listed for loans in MD even though MD doesn't allow P2P?  One or two, I'd say "ehh, someone cheated", 12,000 loans, which is right at the average for the data listed makes me ask questions.

Great question. I've had it pop in my head as well but never got around to posting about it.

http://i.imgur.com/dL0n0Dg.png -- investors graphic that should have been borrowers.
<snip>
Via: http://www.lendingmemo.com/lending-club-and-prosper-states/

[kbenson99]

MD allows LendingClub P2P borrowing, but not investing.  Hence, the 12,000+ loans listed for MD.  Noted under the "Borrowers: 45 States Open to Lending Club" heading within http://www.lendingmemo.com/lending-club-and-prosper-states/

[lascott]

MD allows LendingClub P2P borrowing, but not investing.  Hence, the 12,000+ loans listed for MD.  Noted under the "Borrowers: 45 States Open to Lending Club" heading within http://www.lendingmemo.com/lending-club-and-prosper-states/

Doh! That is obvious now.

[craigwbryant]

MD allows LendingClub P2P borrowing, but not investing.  Hence, the 12,000+ loans listed for MD.  Noted under the "Borrowers: 45 States Open to Lending Club" heading within http://www.lendingmemo.com/lending-club-and-prosper-states/

Simply one more reason I can't wait to get orders out of this state.  If you're going to allow a form of commerce, allow both sides of the transaction.

[aliengar0]

read this article this morning about mortgage defaults on the decline - http://finance.yahoo.com/news/black-knight-financial-services-first-130000513.html

maybe a good sign in general

[RaymondG]

I made up the formula to try to get something to sort by so a state with a larger number of notes but with lower loss would sort OK. Happy to adjust the formula based on anyones input.

Two thoughts:
1) I would prefer to exclude notes whose age is less than 6 months. They distort the Loss Rate too much.
2) The Relative Loss Rate is not a good measure. It could increase merely because # of notes in the state increases. I would first mark the states whose corresponding # of Notes is larger than a threshold which ensures credibility. Then, the marked states will be sorted by Loss Rate. The interest rate by state is ignored since they very close in your portfolio.

[RaymondG]

PeerCube Lending Club performance - https://www.peercube.com/histperf/perfbyattr/addr_state

It's sad that they do not provide basic filter, such as Loan Grade. To me, only D and E are relevant.