Peer to peer lending

[Onward]

People who take out 5 year loans aren't much more likely to have a default event than 3 years

Actually it looks like they are. If this data is to be trusted, 5-yr loans are siginificantly more likely to default over any given time period:

Random Thoughts: Lending Club Loan Length and Default Rate

Since seeing this, I've been favoring 36-mo loans.

On another note, what do you make of the many loan listings in which the employer is listed as "n/a", even when there's a hefty monthly income shown?

 

[cute fuzzy bunny]

Actually it looks like they are. If this data is to be trusted, 5-yr loans are siginificantly more likely to default over any given time period:

Random Thoughts: Lending Club Loan Length and Default Rate

Since seeing this, I've been favoring 36-mo loans.

The writer forgot to model adding two years of a new 3 year loan to the end of the 3 year data to get an equal footing for the two. I'm willing to bet that two 3 year loans suffer more defaults than one 5 year. But there also isn't enough useful data going back far enough to tell for sure.

The one thing I do know is if you exclude 5 year loans on data like this, you're significantly reducing your available notes. So far I'm seeing zero correlation between loan length and default status. In fact, so far I'm seeing very little correlations at all in my late notes.

On another note, what do you make of the many loan listings in which the employer is listed as "n/a", even when there's a hefty monthly income shown?

That means they're self employed. I scanned a ton of data and never found anything really solid about it, except that self employed borrowers pay late more often than regular employees, but they're also less likely to default. Cash flow problems cause the lateness, but the ability to create additional income when needed helps on the defaults. But I'd take their income with a grain of salt, unless its verified. Then you have the funny stat that verified income loans default more than unverified ones. My guess on that was something smelled fishy, calling for the verification. Income might have been okay, but the fishiness remained.

 

[Onward]

His second chart compares default rates quarter by quarter, starting at loan commencement. In all quarters the default rate of 5-yr loans is much higher than 3-yr loans. So I don't see how the extra length of 5-yr loans makes a difference here? We haven't even reached the fourth year of the oldest 5-yr loans.

That means they're self employed.

Or retired?

 

[cute fuzzy bunny]

His second chart compares default rates quarter by quarter, starting at loan commencement. In all quarters the default rate of 5-yr loans is much higher than 3-yr loans. So I don't see how the extra length of 5-yr loans makes a difference here? We haven't even reached the fourth year of the oldest 5-yr loans.

It could be simply that the type of borrower changed or that lending club became more popular and a wider group of borrowers came in, or its because they're taking larger loans. I've seen this thrashed to death by people far more skilled in statistics than I am. For the most part, about 95% of what I see written about the LC process and the data is pretty unscientific and involves a lot of assumptions. From the prairie dog view, I'd have a hard time figuring out how a 10k loan at 3 years would default more than a 10k loan at 5 years, with a lower payment. Since a lot of early defaults come because the borrower took on more debt than they wanted or were able to pay, and its fairly random, I'd think that having to deal with two borrowers over the same time frame, both with larger payments, would be better.

I know that when I looked at it about six months ago, there was no clear cut data either way, and it sort of looked like a 5 year loan would be better. Either way any points offered were leaking badly from several holes. I think its a canard. There is a lot of thin data, a lot of faux correlation.

This also drives the LC people nuts. The loan grade, fico score and a bunch of other evident factors are probably 95% of what we need to know to write a loan. All this other stuff is just trying to snip off small bits of the default rate.

Or retired?

Yep, it just means they don't have a formal employer, and they shouldn't list the name company they own. Probably part of the confidentiality of the thing...if you said what business they owned and where they lived (and we have the latter) you could figure out who they are. As it is, I've written several loans to people working for businesses in the area with a small number of employees, and with their loan app info I could find out who they are.

Otherwise (for other folks reading this), you don't know who your borrowers are aside from where they live and work, and they don't know who you are either. Keeps people from taking the loan sharking collection efforts into their own hands. Thats also a tipoff that you're dealing with a less than intelligent borrower, when they put their name in as the loan description or some other part of the visible application process.

 

[Nords]

I'm still at the $25 per loan level until I reach 800 'active' loans- right now I'm at about 730 loans in portfolio. When I reach 800, then I'll step it up to $50/loan.
I think what I am doing now is scalable by up to one order of magnitude- I would just add a zero to the $25 per loan figure. I have not had trouble applying my filters and moving the necessary amount of money per month (I call it the 'churn'- which is just under 6% of your account value per month that will always be in play when you are funding 36-month loans).

Thanks-- I'm looking forward to seeing how that works out.

Do the P2P companies or the borrowers have any obstacles to getting a $250 chunk of a loan instead of $25? (My typing fingers insist that those numbers should be "$250K" and "$25K".)

I wonder who these institutional investors are and what IRR they're getting.

When I started reading up on the subject, I hadn't even considered the churn rate. Just one more complication.

P2P lending discussions are beginning to sound like the 1990s threads on commodities and New Zealand timber farms. They're not necessarily bad investments, and they certainly offer a measure of diversification, but I'm not sure that the effort is worth the risk/yield. The challenge of figuring out the unknowns and exploiting them for profits? Certainly. Entertainment? Absolutely. A decade or more? Not so sure about that.

What IS worth the risk/yield? A boring ol' portfolio of diversified asset classes where nobody's tempted to chase yield just because one of the assets is returning less than double digits.

 

[haha]

Good that LC has a collections process built in that the lender doesn't need to initiate. Looks like it takes too much day to day babysitting for my taste though, and no collateral for the loans? Not liking that. Our alternative is hard money loans.

An example:

On 7/18/12 we lent $75k on the Summer St. house. It got a known flipper's repairs and staging, and is expected to close and pay back our loan by the end of the month. Pretty quick flip, we'll have collected about $1875 in 10% interest for the three months plus $300 in thanks-for-funding-this-one bucks. We drove out and walked through the house, spent maybe an hour including drive time, maybe a couple hours w/ drive time at the title office, and maybe an hour setting up and entering monthly payments in Quicken.

2746 Summer St SE, Salem, OR 97302 - Zillow

Hi Calmloki, I wondered if this lending thread might call you.

Cute house. I don't see the sale that you funded on this listing. Can you tell us what the flippers paid for it, and when? Did you provide purchase money funds and get your lien in escrow?

That certainly is a good profit. I supose you have learned who the honest and efficient flippers are.

Do you do this as something that spins off earned income? So you could fund an IRA or Roth if you want? At the cost of paying self employment tax I guess. Not sure if that is a good deal or not, but thinkng about it IRS regs might classify you as a business no matter what you want?

Or is a schedule B interest reporting?

Ha

 

[Nords]

- $70k in lending club, which can throw off 0-2500/mo cash flow, rest is as liquid as the secondary market I consider these short-medium term bonds bordering on the junky side with a 5-6% default rate

So are you there yet? How long are you planning to maintain that level of investment?

How many loans in the portfolio, and is the $70K spread out evenly across them or is more invested in certain loan types?

When I divide $2500/month into $70K I get a lot more than the IRRs that people have been discussing. What goes into that $2500/mo figure?

 

[JDARNELL]

Is there any info on if defaults do occur how long they start after loan initiation? I wonder if you could churn notes on a short timeline and reduce default risk? However this is more work of course I guess you could the auto select option. And the assumption would be a strong secondary market to off load loans.

JDARNELL

 

[Lazarus]

Gold flees the man who would force it to impossible earnings or who follows the alluring advice of tricksters and schemers or who trusts it to his own inexperience and romantic desires in investment

(The richest man in Babylon.)

 

[Texas Proud]

So are you there yet? How long are you planning to maintain that level of investment?

How many loans in the portfolio, and is the $70K spread out evenly across them or is more invested in certain loan types?

When I divide $2500/month into $70K I get a lot more than the IRRs that people have been discussing. What goes into that $2500/mo figure?

I am sure he will respond, but I think he is including return of principal...

$70K/36mths = $1,944 without any interest.

 

[calmloki]

Howdy Ha -

The flippers negotiated with the bank directly and hammered the asking price down hard, we provided the $75k purchase price at the closing 7/18; this is a flipper we've been impressed with and I commented on in the past. We did not fund improvements or repairs. While I wanted to claim this was a business to write off a nasty bad loan in one year our tax person smacked me upside the head and we are declaring our lending proceeds as ordinary interest income.

Hi Calmloki, I wondered if this lending thread might call you.

Cute house. I don't see the sale that you funded on this listing. Can you tell us what the flippers paid for it, and when? Did you provide purchase money funds and get your lien in escrow?

That certainly is a good profit. I supose you have learned who the honest and efficient flippers are.

Do you do this as something that spins off earned income? So you could fund an IRA or Roth if you want? At the cost of paying self employment tax I guess. Not sure if that is a good deal or not, but thinkng about it IRS regs might classify you as a business no matter what you want?

Or is a schedule B interest reporting?

Ha

 

[cute fuzzy bunny]

So are you there yet? How long are you planning to maintain that level of investment?

How many loans in the portfolio, and is the $70K spread out evenly across them or is more invested in certain loan types?

When I divide $2500/month into $70K I get a lot more than the IRRs that people have been discussing. What goes into that $2500/mo figure?

Yeah, the 2500 would involve withdrawing capital along with interest, and it'd dwindle down to nothing after ~5 years. But thats its purpose, a cash cushion in place in case my checking account gets low and I need cash flow until the market picks up. As I mentioned, I'd use CD's or short term bonds, but I like 10-12% a lot better than less than 1%.

I can see that you're working hard to make this sound harder than it is, and you're right it isn't for you. If I had two large military pensions to cover all of my expenses and then some, contraptions to create income wouldn't look that interesting to me either.

 

[cute fuzzy bunny]

So are you there yet? How long are you planning to maintain that level of investment?

How many loans in the portfolio, and is the $70K spread out evenly across them or is more invested in certain loan types?

When I divide $2500/month into $70K I get a lot more than the IRRs that people have been discussing. What goes into that $2500/mo figure?

Yes, I'm there. I have roughly 1900 notes, 95% of those are $25, the rest range from $50-500. Yes, that auto invest tool coughs up some large loan segments when your filter is picky.

I'll maintain it until I can get relatively good cash flow and better returns from anything else. Since interest rates are slated to be near zero for another 2 years, at least two years.

As I mentioned a couple of times in the 'ten year' thread you pointed to and in this thread, I only invest in credit card consolidation, debt consolidation, car loans and home loans. Everything else has a higher default rate and some loan types like small business, education and renewable energy are money losers.

Its strictly an interest bearing lump of bonds paying back interest and capital daily, usually in the $100-300 range per day. I can let that accumulate, I can reinvest, or push a button and move it to and from my checking account.

Another random point of interest...lending club was quite handy in that when you link accounts it doesn't seem to matter if the account ownership matches. I had no problems linking it to a joint account with my name on it, sloozing all the money out of that to LC, then pushing it back into a non joint account somewhere else.

 

[cute fuzzy bunny]

Is there any info on if defaults do occur how long they start after loan initiation? I wonder if you could churn notes on a short timeline and reduce default risk? However this is more work of course I guess you could the auto select option. And the assumption would be a strong secondary market to off load loans.

You can do all sorts of stuff. In some states, they can buy secondary notes already issued, but can't buy new notes, and people frequently dump their LC notes due to death/divorce/whatever. Some people never buy new notes but do a nice little business on salvage.

You'll see little default waves for the first 11 months or so until the loans are considered 'seasoned'. You can actually sell a good interest rate seasoned note on the secondary market for more than face value.

You'll see a little blip, in my case about 6-7 notes that went late the first month. Only one never paid, a $5k wedding loan that didn't have any questions asked before I invested, but did after and had I seen those...well...looks like someone was living in moms basement, thought they were getting married, and then their spouse to be saw their comments on LC and ran like hell. A lot of those lates just paid the money back.

Two went into chapter 7 bankruptcy within a couple of months. I'm annoyed that they can do that and not get "eh, thats fraud, so cough up the money or go to jail for a while".

Then things settled with only an occasional late, and I've run at about a 2:1 ratio of people paying all the money back to people who are late. I'm in the 7-11th month window where I'm supposed to get the bulk of my lates, and I've had a bubble of about 10 the last 2 weeks.

By my estimates, if I have 75 defaults this year, I'm just slightly better than average, and I have around 25 right now. Looks like I'll be a lot shorter than 75.

After that, its death, major illness or injury, divorce or significant long term job loss. Over the next 5 years I expect 100-125 notes to go belly up.

Looking strictly at comparing the income to default rate, I've got ~4k in interest right now, ~1K in lates, and they gave me $1000 to try it out. If I keep a ~3:1 ratio of interest to lates, I'll hit around 12%. After recalculating the LC NAR to a real return, it'll probably be around 10%.

 

[cute fuzzy bunny]

Gold flees the man who would force it to impossible earnings or who follows the alluring advice of tricksters and schemers or who trusts it to his own inexperience and romantic desires in investment

(The richest man in Babylon.)

The Gold Lender of Babylon – The Richest Man in Babylon, Pt. 5 – Getting Your Financial Ducks In A Row

This was a good lesson for me, and I took it to heart. While a lot of lenders don't mind helping someone pile on more debt than they can bear, hoping to god they can pay it back...I won't do that. The lesson of "Don't help someone get into trouble, because their trouble will become your trouble" is well met.

If I see more than 30k in debt, or that they're paying a big debt load with a meager paycheck, unless they're consolidating that debt, I don't want them. I see a lot of people making peanuts that want a movie star wedding, and a lot of people that according to their spending have been just having a blast with toys and trips, but then all of a sudden they need a new roof and kitchen! Hey guys, if you need 40k for home improvements, perhaps you should plan that out and spend the 40k in credit card debt you're holding on that future need. Even though some have huge incomes and could wade through all of that debt...I don't really want to be a party to that.

 

[Nords]

I can see that you're working hard to make this sound harder than it is, and you're right it isn't for you. If I had two large military pensions to cover all of my expenses and then some, contraptions to create income wouldn't look that interesting to me either.

Y'know, CFB, I drafted a long response to your points, but I'm belatedly realizing that would be pointless.

I hope your P2P plans work out for you.

I'm glad that your life is heading in a better direction. You have a good one now.

 

[arebelspy]

Thanks for the overview Nords.

I'm fairly meh on P2P overall (speaking as someone who has had over 50k invested in it and no longer does so).

Seems like it comes up a fair amount in the various early retirement communities, this will be a good post to refer them to.

 

[MooreBonds]

Thanks rbmartin for the pointers to some development tools. About all I want is the old school quickbasic where you get a prompt, type in ten lines of old school "open the file, read until end, check 10 fields for criteria and write a line into a second file if I find a hit" code and run it. I'll look into the stuff you provided, as I need to do this sort of thing every once in a while and I'd hate to learn another language or a big time development package to do something simple.

I don't recall from the earlier messages in this thread, but are all of the files available in .csv format? If so, you could simply take turns sorting the rows by one column at a time, then scroll down in that column and highlight (i.e. turn the cell background a certain color) for the range you like.

Then, cut and move the next column to the end of the spreadsheet, and sort the rows by that next column, repeating with setting the attractive parameters with a certain cell background color.

Repeating this for the various columns would only take 1 minute, and then reveal a hodgepodge of rows that have different numbers of variables that meet your criteria. Might be easier than coding it, AND would allow you to easily see at a glance any outliers that might have a variable that's just 0.2% outside of your parameter, but otherwise be an attractive loan.

 

[JDARNELL]

I don't recall from the earlier messages in this thread, but are all of the files available in .csv format? If so, you could simply take turns sorting the rows by one column at a time, then scroll down in that column and highlight (i.e. turn the cell background a certain color) for the range you like.

Then, cut and move the next column to the end of the spreadsheet, and sort the rows by that next column, repeating with setting the attractive parameters with a certain cell background color.

Repeating this for the various columns would only take 1 minute, and then reveal a hodgepodge of rows that have different numbers of variables that meet your criteria. Might be easier than coding it, AND would allow you to easily see at a glance any outliers that might have a variable that's just 0.2% outside of your parameter, but otherwise be an attractive loan.

I was thinking about how to do this also. My approach was to create "if" statements for each column in excel and assign colors. Then use a count function to identify the greatest # of greens and sort the rows. using the same spreadsheet and just cutting in the data would be quick and give you an easy visual of where to focus your time.

 

[photoguy]

His second chart compares default rates quarter by quarter, starting at loan commencement. In all quarters the default rate of 5-yr loans is much higher than 3-yr loans. So I don't see how the extra length of 5-yr loans makes a difference here? We haven't even reached the fourth year of the oldest 5-yr loans.

Thanks for posting the chart. I hadn't seen a clear depiction of default rates from p2p lending sites before. Looking at the chart though, with >10% default rate in 2009 makes me think wouldn't I be better off just increasing my equity portion instead of reaching for yield with these junk bond like loans?

 

[haha]

Thanks for posting the chart. I hadn't seen a clear depiction of default rates from p2p lending sites before. Looking at the chart though, with >10% default rate in 2009 makes me think wouldn't I be better off just increasing my equity portion instead of reaching for yield with these junk bond like loans?

This is not attractive to me, and one of the clearest reasons is that there is a pretty high default rate, even though we are well into a recovery. These loans will be highly correlated when things get stinky. Which they will.

Also interesting is that although quite a few of our members are women, I don't recall any woman posting about doing this lending. A macho undertaking?

Ha

 

[cute fuzzy bunny]

Y'know, CFB, I drafted a long response to your points, but I'm belatedly realizing that would be pointless.

I hope your P2P plans work out for you.

I'm glad that your life is heading in a better direction. You have a good one now.

You too!

I didn't write any long responses though. You were never really interested in this, and you don't need it.

Back on topic...

I did have some thoughts around the 3 vs 5 year data. When LC first offered these, the initial customers were people who didn't qualify for the 3 year loan payment, so LC offered to relist the loan at 5 years. In other words, the first ~6 months or more of borrowers weren't as qualified as the 3 year borrowers.

Of course, if you're filtering your own loans for credit worthiness and whatnot, you're leveling the playing field...if its still off level now that everyone has the option of 3 and 5 year right up front. Remember that ~95% of my loans are for people who ran up 20-30k of debt and just want to make one smaller payment with an actual "out of debt" day. For the people who mean it, the 5 year should be easier to pay.

Looks like the more 'contemporary' the data gets, the closer the two default rates run. I'm seeing less than a 2% variance in the most current, and the interest rate difference you get from a 3 to a 5 year loan in the grades I'm buying is more than that.

 

[cute fuzzy bunny]

This is not attractive to me, and one of the clearest reasons is that there is a pretty high default rate, even though we are well into a recovery. These loans will be highly correlated when things get stinky. Which they will.

Even if the more serious default rate doubles, I'd still eke out 3-4%. Where else can you get something that in the likely worst case scenario still makes money? I keep hearing "add more equities" and I still don't know how to generate more income from those without taxable events and I'm not sure how selling those into a down market to make the income helps me.

Also interesting is that although quite a few of our members are women, I don't recall any woman posting about doing this lending. A macho undertaking?

Same with a lot of other things, mostly stupid. We do let our little heads and testosterone make a surprising percentage of our decisions. Might be kind of boring if we didn't.

My ex wife was constantly skeptical of lending club, yet here I am making a lot of money on it with a history that goes back through the recent recession that says that as long as I'm well diversified (800+ notes) I won't lose money. She was also dubious of my 'eat fat, lose weight diet' and look at how that turned out!

I agree that we might see a second dip and a higher default rate, but this is like the old joke where two guys see a bear and the bear starts charging. One guy stops to take off his hiking boots and put on his sneakers "You cant outrun that bear" says the other guy. "I don't have to! I just have to outrun YOU!". Bottom line is, you write notes that carry an interest rate that exceeds the default rate. My overall average right now is 17% and rising, only pulled down by all the B and C grade debt I wrote to get my capital deployed. That quick deployment of a goodly chunk of cash got me a $1000 freebie from LC. My defaults so far haven't even eaten that up.

That means defaults would have to go north of 15% from their current 5-7% at the grades I'm writing.

There really shouldn't be any correlation between LC loan default rates and the stock market. Most of the people taking debt consolidation loans aren't heavily invested in the market, and few are going to exclaim "Oh my god, the Dow just dropped 5000 points! I'm not paying my debt consolidation loan!"

The correlation is between LC loan default rates and job loss or huge individual financial stress like a house going from 500k to 250k in 3 years. There would be some trailing job loss issues correlated to a stock market plunge, as there would be for any significant economic malaise but it'd hardly be immediate.

I understand that the process of certain losses bugs the crap out of some people. Bugged me too at first, but I realized that even index stock funds have some holdings that tank or go out of business. You lost money there. It was just made up somewhere else.

This is simply high grade junk investing for income. Whats the rule there? 5% spread over treasuries means buy? Sounds like I need to beat .3% for 3 year treasuries and .6% for 5 year treasuries. I think I'll do that handily.

You can buy an index (with the autoinvest tool) or actively manage a portfolio of loans. Someone else does all the work aside from selection and deciding what to do with the money.

 

[GrayHare]

http://si.wsj.net/public/resources/images/BF-AD650_INVEST_G_20121012172200.jpg

 

[haha]

Even if the more serious default rate doubles, I'd still eke out 3-4%. Where else can you get something that in the likely worst case scenario still makes money? I keep hearing "add more equities" and I still don't know how to generate more income from those without taxable events and I'm not sure how selling those into a down market to make the income helps me.



Same with a lot of other things, mostly stupid. We do let our little heads and testosterone make a surprising percentage of our decisions. Might be kind of boring if we didn't.

My ex wife was constantly skeptical of lending club, yet here I am making a lot of money on it with a history that goes back through the recent recession that says that as long as I'm well diversified (800+ notes) I won't lose money. She was also dubious of my 'eat fat, lose weight diet' and look at how that turned out!

I agree that we might see a second dip and a higher default rate, but this is like the old joke where two guys see a bear and the bear starts charging. One guy stops to take off his hiking boots and put on his sneakers "You cant outrun that bear" says the other guy. "I don't have to! I just have to outrun YOU!". Bottom line is, you write notes that carry an interest rate that exceeds the default rate. My overall average right now is 17% and rising, only pulled down by all the B and C grade debt I wrote to get my capital deployed. That quick deployment of a goodly chunk of cash got me a $1000 freebie from LC. My defaults so far haven't even eaten that up.

That means defaults would have to go north of 15% from their current 5-7% at the grades I'm writing.

There really shouldn't be any correlation between LC loan default rates and the stock market. Most of the people taking debt consolidation loans aren't heavily invested in the market, and few are going to exclaim "Oh my god, the Dow just dropped 5000 points! I'm not paying my debt consolidation loan!"

The correlation is between LC loan default rates and job loss or huge individual financial stress like a house going from 500k to 250k in 3 years. There would be some trailing job loss issues correlated to a stock market plunge, as there would be for any significant economic malaise but it'd hardly be immediate.

I understand that the process of certain losses bugs the crap out of some people. Bugged me too at first, but I realized that even index stock funds have some holdings that tank or go out of business. You lost money there. It was just made up somewhere else.

This is simply high grade junk investing for income. Whats the rule there? 5% spread over treasuries means buy? Sounds like I need to beat .3% for 3 year treasuries and .6% for 5 year treasuries. I think I'll do that handily.

You can buy an index (with the autoinvest tool) or actively manage a portfolio of loans. Someone else does all the work aside from selection and deciding what to do with the money.

Sounds convincing. I may be skeptical only because of inexperience.

Ha