Author Topic: Do I have this right?  (Read 20397 times)

Dennis

  • Full Member
  • ***
  • Posts: 141
    • View Profile
Re: Do I have this right?
« Reply #45 on: September 01, 2013, 03:02:41 PM »
Sorry, but that hasn't been my experience lately.  Most of the notes that fill or mostly fill within a minute or two look pretty good.  Most of the ones that are still <65-70% after a minute or so have at least one wart on them.  I still look at them, but I'm good with the vast majority of ones that go from release to closing in 90 seconds, and the ones that don't I usually don't care for either.

They're D-G notes, over 680 fico, no huge revolving debt (<35k), only credit card/debt consol/home improvement, working for 2+ years, no 'other' home ownership, no major derogs, no public records, make over 5k a month, payment close to 10% of the gross monthly income, 0 or 1 inquiry, credit line > 5 years, dti < 25 (preferably 20), no more than 2 delinquencies, and no 'spouse loans'.

Everything doesn't sell off right away either.  I see ~40-75 notes that are all A or B with some issues or lower grades with lots of issues that are hanging around for a while.  Interestingly, one note that was titled "Dallas Cowboys Rock!!!" filled halfway, then sat there for 2 days as non Cowboy fans were (I guess) turned off by the title, then the title changed and it did fill within the day.  Wasn't a bad loan, but wasn't a great one either.  I doubt the autobuys read and interpret the loan title, so it seems people are still well involved.

In fact (and I think I said this before), an analysis of what DOESN'T sell within 15 minutes is a good example of what not to buy.  Low interest notes, lots of inquiries, high debt, lots of defaults, low fico's, no job length/employer, 10+ years of credit history with a low revolving balance and few lines of credit (spouse loan because the other spouse has crap credit), very low income, and small business loans.

I would agree to most of what you say.  Your criteria is very good and pretty much mirrors what I've used for the last 2 years.  So good job on that.  But it sounds to me that you're buying based on criteria and that's exactly what I said you should be buying on, not what is getting 99%+ funded quickly.  Like I said, I've seen some pretty retched notes quickly fill and for the life of me can't figure out why since they are so bad looking.  So that 99%+ fill rate wouldn't work for me, but we all have different styles of investing.  Also, some of those 99 percenters will return to the board 5 - 30 minutes later having been rejected by the big guys while they sat in their cache during that time.  So you won't know the true fill rate unless you click open each individual note to see how many investors in that note actually completed their purchase (the actual fill rate could be say 4% while the cache fill rate might show 100% for up to 30 minutes).  Hope you didn't get stuck in any of those as many of those look awful.  As long as 2 years down the road you're still happy with your returns, then there's nothing wrong with your method.  There's more than one way to skin a cat - as the saying goes.  Best of luck my friend.

cfb

  • Full Member
  • ***
  • Posts: 132
    • View Profile
Re: Do I have this right?
« Reply #46 on: September 03, 2013, 01:30:20 PM »

This seems to apply to institutional investors as well.  The fact that institutionals held their notes in the cart for 25 minutes before committing the orders seemed to indicate that humans were still involved.

I find some good value in reading the titles and the comments, if any.  Those usually lean on the side of damning.  But if I see a loan with just a small wart on it but the title is all in caps, mis-spelled, and/or says something weird I take that as a sign and pass on it.

Quote
This sounds like a good project to do.  If you have done an in-depth analysis, I'd love to see what quantitative results you found.

Other than reading the past loan data (and I take that like past stock performance, not necessarily an indicator of future performance) I did dig into a few areas that I found little info on or made me curious.  I don't ding people for living in certain states like Florida and California.  Those had a lot of influence from overinflated home prices that popped, and unless that repeats I doubt that the default rate in those states will be much higher than norm.  While cost of living is high, job availability is also high and so are the incomes.  But someone making $3500 a month living in San Francisco would concern me.  I try to think about what economic circumstances may have influenced defaults and late payments, and consider whether those are likely to recur in the next 3-5 years, or if there may be different influences that might turn a source of defaults into regular payers.

I looked into the self employed (employer NA) and did a little research into historic credit card defaults on those, and it turns out that they sometimes pay late, but have a much higher rate of eventually paying.  Cash flow issues.

Most of the rest of my influence in looking at these was derived from the stories of lenders from the distant past.  The wise ones recognized people who were overburdening themselves with debt and tried to avoid being greedy and making their problems worse.  Because the borrowers problems tend to become the lenders problems, eventually.

Other thing that I seem to regularly spot is people who are trying to refi their small business debt as 'major purchase' or 'credit card consolidation' when its really a small business loan.  Sometimes its just something in the title, or its in the comments.  So maybe the institutional/autobuyers like to look at the titles or comments and chuck back a percentage of those since the SBL's default at a high rate.  That would explain why they sit in the cart for a while and some come back.  Probably makes good sense.

Dennis - my filters are pretty detailed, so not much ends up in my basket and what does is the cream.  I was kind of joking about the high speed filling making it easier to pick loans, but I do find that what doesn't fill right away, on further detailed examination, has an issue caught by the eye but not filterable.  Like a credit card loan titled 'buying new equipment for my business'.  But I do find that most of what passes my filters AND sells out right away generally look like very good notes.  Its an extra data point.

LC site was reeeeallly slow this morning.

Rob L

  • Hero Member
  • *****
  • Posts: 2065
    • View Profile
Re: Do I have this right?
« Reply #47 on: September 14, 2013, 06:59:56 PM »
1) I don't see it quite that way. I would agree it would be no worse but I think probably not substantially faster.

But you do see that by necessity the examine-before-lock approach always requires one more HTTP request at minimum, right?  And that's only if your chosen note shows up first in your sorted list... fat chance.  "Substantially" is of course a relative term so I cannot argue with that.   

But what started all this is my saying that all notes can be bitten off with one chew, which I still believe.  I will be happy to prove this, but I will need a different test account.  Possibly from someone else on the forum who doesn't mind having their account locked and having to speak with Stephanie.  Please, all of you do not fill my PM inbox all at once jumping at this chance.  ;)

Mostly (yeah, I know I'm a little slow on the uptake, but I now agree and had some free time to look into the matter). If you already know the loan ID you want to buy you can pop it into your cart with one HTTPS request (not an unlikely presumption for an institutional buyer). However, each and every additional loan you add to the shopping cart requires another HTTPS request (unless the method is well hidden and I have not been able to find it). Once there, one can peruse the cart at their leisure. The API (for a single account and multiple loan purchases) is much faster, but has its own disadvantages.

Rob L

  • Hero Member
  • *****
  • Posts: 2065
    • View Profile
Re: Do I have this right?
« Reply #48 on: September 19, 2013, 12:10:05 AM »
Mostly (yeah, I know I'm a little slow on the uptake, but I now agree and had some free time to look into the matter). If you already know the loan ID you want to buy you can pop it into your cart with one HTTPS request (not an unlikely presumption for an institutional buyer). However, each and every additional loan you add to the shopping cart requires another HTTPS request (unless the method is well hidden and I have not been able to find it). Once there, one can peruse the cart at their leisure. The API (for a single account and multiple loan purchases) is much faster, but has its own disadvantages.
I was so wrong. If you know the loans you want you can put them all into your shopping cart with one HTTPS request. Not hard to do or hidden at all. What was I thinking? :-[

core

  • Hero Member
  • *****
  • Posts: 1784
  • Your loss is my gain
    • View Profile
Re: Do I have this right?
« Reply #49 on: September 19, 2013, 07:32:51 AM »
I was so wrong. If you know the loans you want you can put them all into your shopping cart with one HTTPS request. Not hard to do or hidden at all. What was I thinking? :-[

That is good news.  Hidden in plain sight eh.  Do we know that the shopping cart exploit still works after the 8/28 updates?  We do know there are changes coming from LC in the very near future so it's possible the ability to lock them all up will go away when LC comes up with their Final Solution.

Joleran

  • Full Member
  • ***
  • Posts: 151
    • View Profile
Re: Do I have this right?
« Reply #50 on: September 19, 2013, 08:47:03 AM »
That is good news.  Hidden in plain sight eh.  Do we know that the shopping cart exploit still works after the 8/28 updates?  We do know there are changes coming from LC in the very near future so it's possible the ability to lock them all up will go away when LC comes up with their Final Solution.

Sure, it still works.  I don't see that going away either, it would cripple anyone not using a note selection algorithm even more.

Rob L

  • Hero Member
  • *****
  • Posts: 2065
    • View Profile
Re: Do I have this right?
« Reply #51 on: September 19, 2013, 09:37:20 AM »
One more experiment this morning. I successfully added a $4,200 purchase of a $5,000 loan to my shopping cart (84% of the total loan amount). Did another; $3,000 on a $3,800 loan (79% of the loan). It appears there is no 75% rule enforcement when adding loans to the shopping cart via the web interface (or the max allowed % is > 84%). If you have enough money you can put whatever you want in there. Obviously I didn't buy the things and the 75% rule may be enforced later at purchase time; don't know about that (and don't intend to find out). Purchases using the API may enforce the 75% rule; I don't know.

core

  • Hero Member
  • *****
  • Posts: 1784
  • Your loss is my gain
    • View Profile
Re: Do I have this right?
« Reply #52 on: September 19, 2013, 09:54:55 AM »
One more experiment this morning. I successfully added a $4,200 purchase of a $5,000 loan to my shopping cart (84% of the total loan amount).

Interesting!  Someone sitting on a lot of uninvested cash due to not being able to put it to work all at once (and it sounds like there are a few of you out there) could certainly cause a bit of trouble if they were so inclined.  Lock them all up and release them exactly 2 hours after the feeding times just to screw with the carefully timed scrapers.