Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.

Topics - Fred93

Pages: [1] 2 3 ... 7
I buy notes on Folio, using the API.

When I place a buy order on Folio, the responses I get are a mix of NOTE_NOT_AVAILABLE and SUCCESS_PENDING_SETTLEMENT.

I'm going to be talking about just those orders that resulted in SUCCESS_PENDING_SETTLEMENT.  Those are the ones I think I bought.

A significant fraction of those never end up in my account.  I've been wondering why.

A few months ago I took a day's purchases, from the log written by my software, and checked them against my notes file to see how many actually arrived and how many did not.  I got a 50% failure rate!  Yipes.

I did this check again last month, and got a 38% failure rate.  38% of the time, the note does not show up.  (For the purposes of this little experiment, 38% and 50% are approximately the same number.)

I've written to LC a couple of times about this.  Most recently their response has been that the specific examples I sent them were all cancelled by payment.  I have quizzed them about the payment timeline, and how they decide which orders to cancel.  They say that the window of opportunity for payment after I place an order is only one day.  That would be 1 day out of about 30 days (because payments come about once every 30 days), so that would be 3.33% expected to fail.  But of course they only process payments on weekedays, so one of the 5 weekdays absorbs the hazard of the two weekend days, so maybe that would legitimately get them to 4.66%

Long way from 50%.

So I'd like to hear the experience of other LC folio buyers.  I'd like to know if I'm alone in this.

What fraction of successful orders fail to arrive in your account?

Do many of you see a large fraction of buy orders cancelled, ie say SUCCESS when you place the order, but it never makes it to your account?

LC is asking me questions aimed at a theory that my selection process accurately predicts who will pay on the day I order.  I've looked over what I'm doing, and I just don't think that's possible.

PS: Jheizer made a brief comment in another thread that he had seen "like 50%" failure.  I can send that to LC, but I'd prefer to send comments from several different people, showing that many people have similar experiences.  That will allow us to get past their theories that this is caused by some strange thing I'm doing.

Suggestions / new forum bug - stale "new" indicators
« on: October 20, 2017, 04:54:12 PM »
Starting yesterday or the day before ... Now the "new" indicators on forums are stale.  Sometimes I've already read the recent messages, yet the forum still shows the icon for new messages available.  Note that if I refresh the page, IT IS STILL WRONG.  About a minute later, a refresh will make it right.

Used to work perfectly.  Did you update the forum software recently?

General P2P Lending Discussion / consumer distress
« on: June 03, 2017, 06:27:17 PM »
Article in Barrons today.
For some time now, Stephanie Pomboy of MacroMavens has highlighted the accumulating stress on consumers. “People who save are those who have the wherewithal to save,” she says, “while poorer consumers are borrowing out of distress to fund purchases normally paid for by income.”

The fact that delinquency rates are starting to turn higher “across all segments of the consumer space, despite near record-low interest rates, is a powerful indictment of the strong consumer narrative so widely embraced,” Pomboy says. How can folks have trouble paying their debts with unemployment at 4.3%, mortgage payments low, and net worth at record highs? “One shudders to imagine what delinquencies would look like if rates actually did move up, or—heaven forbid—stocks went down,” she adds.

When Stephanie Pomboy speaks, I  always listen. 

NSR Invest / nsrplatform backtesting error messages
« on: May 21, 2017, 05:32:41 PM »
nsrplatform LC backtesting has been very difficult to use for last several weeks due to these error messages, which I presume are some sort of timeouts within your servers.

When one of these appears, I click "update results" to try again.  Error then often occurs a 2nd time, etc. 

NSR Invest / Prosper back testing
« on: May 21, 2017, 04:33:16 PM »
The Prosper data in nsrplatform seems to end 2016-06-30. 

Do you plan to update the data? 

I've never really done a long term Prosper vs LC comparison of returns before, and its about time so...

I've computed my monthly returns the simplest possible way.  I took month-end account value divided by prior month-end account value, minus one, and multiplied it by 12 to annualize.  When I added money during the month, I treated it as if it was present the entire month, which adds a small negative bias to the resulting numbers.  This occurred during several months in 2015.

First thing I notice is that the Prosper returns are more volatile.  They really jump up and down.  Some of this is expected, because my Prosper account contains fewer loans than my LC account.  On the other hand, that odd month up, even month down pattern in the last year is very strong, which may indicate something about how Prosper processes chargeoffs.  It is what one might expect if they processed chargeoffs every other month, for example. 

For comparison, I've shown the Orchard Index on the same chart.

Some big picture thoughts...

I was amazed to see how closely my returns matched between Prosper and LC.  I use a similar loan selection strategy on both platforms.

I am happy to be beating the Orchard Index consistently.  The limitation of the Orchard Index is that they don't tell us what P2P companies are in the index, or their weightings.  I presume that both LC and P are in the index, but there's no way to be sure.

We're all moving to lower returns.

Suggestions / web server problems
« on: May 06, 2017, 12:20:18 PM »
During the last 24 hours, I've had quite a few forum pages intermittently take >30 seconds to load, and also some error messages, including...

Resource Limit Is Reached
The website is temporarily unable to service your request as it exceeded resource limit. Please try again later.


Service Temporarily Unavailable
The server is temporarily unable to service your request due to maintenance downtime or capacity problems. Please try again later.
Additionally, a 503 Service Temporarily Unavailable error was encountered while trying to use an ErrorDocument to handle the request.

Investors - LC / When will LC next raise rates?
« on: April 07, 2017, 06:43:46 PM »
I'm gonna run this for 23 days, which I presume makes it end on Apr 30.  I will end it sooner if LC raises rates before then.

General P2P Lending Discussion / stacking
« on: March 19, 2017, 09:53:22 PM »
We've discussed stacking before.  (This is the practice of applying for multiple loans within a few days or even hours, so fast that the credit bureaus aren't updated, and the 2nd, 3rd, 4th... lender is unaware of the prior loans.)

Nice talk on stacking at Lending 2017.

I've seen articles on this subject before, but this is the first I've seen someone present actual statistics.

PS: Like most talks, this guy is "selling", so make use of the speech speedup feature.  Click on the little gear, and then "speed" and select 1.25x.

Investors - LC / The identiy fraud repayment promise
« on: March 19, 2017, 09:38:50 PM »
Lending Club promises to buy back loans (notes) "where the loan was obtained as a result of identity theft or fraud".

From the LC investor agreement...
4. Limited Repurchase Obligation for Identity Fraud. If the Member Loan designated for the proceeds of your purchase of a Note was obtained as a result of identity theft or fraud on the part of the purported Borrower Member, we will (a) provide notification either directly or indirectly to you as soon as reasonably practicable upon our becoming aware of such a situation; and (b) repurchase your Note by crediting your account for the outstanding principal balance of your Note. For the avoidance of doubt, (i) if you purchased the Note indirectly through a third-party platform or entity, we will work with such third party to credit your investment account through that entity or platform for the outstanding principal balance of your Note, and (ii) if you purchased the Note through FOLIOfn Investments, Inc., we will credit the Lending Club account that you established in order to purchase Notes through FOLIOfn Investments, Inc. We may, in our reasonable discretion, require proof of the identity theft, such as a copy of the police report filed by the person whose identity was wrongfully used to obtain the fraudulently-induced Member Loan, before we credit your applicable account and repurchase your Note. You agree that you will have no rights with respect to any such Notes except the crediting of the purchase price to your applicable account

Here's the thing that bothers me.  Since 2009 I have bought 8399 notes from Lending Club.  (I bought some before 2009 too, but I don't know how many, because they aren't included in my notes file any more.)  Given that I have bought a very large number of notes, how come I don't recall LC ever notifying me of an identity theft note?  Not even once!

How can this be?  I figure LC is pretty good at weeding out identity theft, but there are people who work at it all day long every day.  There was one talk at Lendit where the speaker talked about massive amazon cloud driven identity theft attacks on lenders.  Wow.  So even if LC is very very good at weeding out ID theft, one would imagine that a few still get thru.  If so, then some fraction of notes we buy from LC must be ID theft notes, and ID theft notes will surely charge off, which concerns us, right?

If even 0.1% of LC notes are ID theft, then out of the 8399 notes I have purchased, there would be about 8 ID Theft notes, right?

Why have I never seen one?

Perhaps LC has been notifying me, and I've been ignoring the notifications somehow.  So I'll ask the rest of you...

Have you ever been notified by LC that a note you have purchased is an ID theft note? 

Ever had such a note repurchased?

If none of us have ever been notified or reimbursed, then there is something wrong, because I refuse to believe that 1.3 Million loans have been issued without a single case of ID theft.

Foliofn - LC / phone call from Folio
« on: March 16, 2017, 05:45:16 PM »
Phone rang an hour or so ago.  It was a compliance guy from folio, with some questions about my LC folio trading. 

He said they run some screens, and talk with a few people who pop up on the screens from time to time.  Seemed reasonable.  He had a few questions about my strategy, and so forth.  Very general stuff.  He seemed to know my trades, and didn't ask for anything I would consider confidential, so I presume he was legit.

He asked if there were things that concerned me, so I told him about some of the software issues with FICO score updates being shown to some people days before being shown to others etc (as we've discussed in other threads here).  Not surprisingly he said that was LC's department, and outside of his purview.  He went on and on with that answer and I had to cut him off.  Told him "Look, just sayin' thats not my problem over and over in different ways doesn't help, so lets move on."

Was a pleasant conversation.  I told him I was glad to see somebody was watching.

General P2P Lending Discussion / Lendit 2017 questions
« on: March 11, 2017, 05:21:17 PM »
The website has videos of a number of the talks now.

So was it all silicon valley happy talk? 

... or did anyone address the serious issues of loan quality and/or performance degradation?

Investors - LC / a view of LC's deteriorating investor returns
« on: February 19, 2017, 04:45:11 AM »
We've all looked at the deterioration in LC loan performance over 2015 & 2016 in a variety of ways.  We've charted delinquencies, chargeoffs, by time, by vintage, used different return measures, etc.  Here's something really simple.

I used numbers from LC's chart with all those dots,

This chart shows the performance of INVESTORS rather than performance of loans, so is therefore close to investor's hearts.  One downside is that this chart uses ANAR to measure investor performance.  I won't detail all the characteristics of ANAR here, but one is important enough that I have to say it.  ANAR measures your performance since you opened your account.  Those of us who have been investing with LC for a long time, have higher ANARs, just because ANAR averages over all the time our accounts have been open, and returns were higher in past years than they are now.  My account is 9 years old, so this is painfully clear to me.  I suspect that many accounts have been open just a few years, so perhaps this effect isn't so bad on the average.  Just be aware.  If Joe's ANAR is 7%, he may be earning only 4% now, for example.

This chart has been changing vs time.  Lets look at what this chart NOW says. 

I made the following selections...
Adjustment for past-due notes: ON
Min number of notes per account: 500  (Lets look at well diversified accounts, so we know the numbers are not noise)
Max note size less than 0.5%  (Same thing, ie well diversified accounts)
Now, there's one more selection, the weighted-average-interest-rate WAIR.  This, like credit grade, is just a measure of the riskiness of the loan, as LC sees it. 

In the middle of the graph, LC highlights accounts with average age of loans in the portfolio of 12 to 18 months.  This is a reasonable definition of a steady-state account, so lets use that selection.  LC displays 10th percentile, median, and 90th percentile ANAR for accounts in this range.

Finally, here's how the median ANAR varies with WAIR.
Code: [Select]
  WAIR     10%   Median    90%
 0- 9%    4.8%    5.3%    6.0%
 9-12%    4.2%    5.2%    6.4%
12-15%    3.4%    4.7%    6.3%
15-18%    2.2%    3.6%    6.5%
  18+%    0.5%    3.6%    6.5%
  ALL     2.7%    4.8%    6.3%

A lot of numbers.  Just look at the Median column.  This shows ANAR for the median account in the selected group.  As you can see, as WAIR goes up, ANAR goes down.  Its monotonic!  The best performing accounts TODAY are those who invested over the past few years in the low interest rate (ie safest) loans.  The worst performing accounts are those who invested in the riskiest loans. 

Among today's accounts, you can see that investing in higher risk loans did not give most investors any advantage.  This is true for the median account, and also true (within one tenth of a percent) at the 10 percentile and 90 percentile account! 

The median LC investor today holds an account with an ANAR of 4.8% .  My account's ANAR is much higher, and yours probably is too, but I opened that account about 10 years ago, so my ANAR averages over times when returns were considerably higher.  Doesn't mean I'm doing better than average lately.  In fact, I have very little idea how I'm doing relative to other LC investors lately.

Meanwhile, the facts for today are that accounts TODAY have a median ANAR of 4.8%, and accounts TODAY show higher ANAR if they have portfolios of lower interest rate loans.  (Those accounts picked those loans over the last few years, so this result tells us about the performance of those vintages.  Unfortunately doesn't tell us what loans picked today will do.)

This is consistent with facts we know about loan performance degradation during the past two years.  Performance and returns on the higher risk grades have gone to crap.

Historically, we have a name to categorize times when loans perform this way (ie when investing in the safest loans was the best strategy, and investing in the risky loans was the worst strategy).  We call those times recessions, or credit crises, or some such name.  We haven't been in either a recession or a credit crisis during 2015/2016, so the explanation is elsewhere. 

I blame LC.  I believe they made considerable changes to their loan underwriting in 2015 to accept more loans, in order to drive volume.  Also possible of course that they just screwed up.  Although they have (multiple times) given lip service to improving loan quality, I see no evidence of same.  I will continue to look.

Among other things, these changes have invalidated all the wonderful historical data.  A lot of back-testing is now simply misleading.  This is quite a loss.  We now have nothing to guide us.  Most of us still use filters we derived from back-testing.  I do.  However, over the past year I've changed how I think about back-testing quite a lot.  I'm suspect this change in attitude is widespread.  In past years, careful attention to historical data gave me an edge of several percent.  Now it looks like that edge is gone.  That's just the way it is.

Pages: [1] 2 3 ... 7