The following loan was fully funded in the first 10 seconds after new loans were released today at 10am PDT (somewhere between 10:01:30 and 10:01:40):
https://www.lendingclub.com/browse/loanDetail.action?loan_id=6816247If you believe 4 different individual investors chose this same loan and executed a buy within the first 10 seconds I have some swamp land in Florida I'd like to sell you. It was, almost certainly, a single API transaction / purchase of the same loan, using four different accounts. This is a $5,500 loan, so each note was probably $1,375.
Does this break the 70% rule? Could be a LLP with a dozen or more wealthy members. Would these be considered legitimate separate and individual buyers, or a simple way to skirt the 70% rule? What if it's one individual with multiple accounts? If this trade doesn't break the 70% rule then what does?
This is looking more like Salem 2013 by the hour. Let's press him with stones!
Keep the defendants coming. This makes for some good reading.
I don't think it's a witch hunt to ask these sort of questions. I've also thought that the 70% rule can be easily overcome with two accounts, and it would be nearly impossible for LC to detect unless both accounts were hosted on the same IP, and the operator made no effort at all to mask the IP.
The point is, LC took some actions to help the "little guy", but the 70% limit doesn't work -- it doesn't even slow things down. The "Whole Loan Program" didn't help -- in fact, it's done quite the opposite by limiting loan access to the small investors.
If LC is serious about keeping the small investors in the game, then they need to find a serious solution.
Imagine if Rev came up with a super-duper invite-only version of Interest Radar, and had only 4 users at the time. A note came in and this note met the filter requirement of all 4 users and the note was purchased from the same IP for the 4 of them. Would that be an abuse? Heck if I know, but it's damn near impossible to police these things.
A bulletproof way to get around such limitations would be to have 4 friends in different cities, 4 proxy IPs, and then you just give each friend a small cut for being able to use his identity. There's no way the 70% limit will prevent people from buying a large percentage of notes. It just adds a miniscule amount of time and trouble.
Why even talk about the 70% rule when they are freely handing out 100% pieces to whoever wants them???
It didn't take long to get to what seems to be complete agreement on this one. Not only did the loan I cited originate from a single IP, I have little doubt it was a single API buy order. Was the 70% rule broken? Ambiguous. Core's super-duper-IR example could easily result in a single API transaction, and FWIW I personally don't think that's rule breaking. Bottom line, the rule is unenforceable, even if there was a desire to do so (which is at best unknown). The cat is out of the bag on this one too (but this game is an obvious one). Its affect on individual retail buyers is considerable.
So the theory was to have the whole loans program for large investors and a 70% rule to protect the retail investors. The unintended consequences resulted in retail investors being denied all access to loans set aside for the whole loan program, and a 70% rule that isn't working and probably can't work.
"The unintended consequences resulted in retail investors being denied all access to loans set aside for the whole loan program, and a 70% rule that isn't working and probably can't work."
Okay.....so now what?
Several possible solutions have been discussed before. My favorite would be limiting the maximum investment to $100 or less for a subset of loans, for a limited time ... something like, for the first 24 hours, for a randomly selected 25% of new issues. After 24 hours, whats left gets put up for grabs. That would be tough to creatively overcome. I supposed the institutional investors could create several accounts and fund $100 each, but I doubt it.
I didn't read most of this, and the other thread on the same/similar topic, but the only "answer" to this is less an option than an eventuality:
Prosper had the proper at-scale model when it launched public in 2006 with RATE AUCTIONS built into the primary market. They
simply built the model that would work at scale before scale was reached, so, they got leapfrogged. Auctions will be back soon.
-t
Bidding on unsecured loans is just a recipe for disaster. I'm pretty such the industry figured that out the hard way.
IMO, there is something else at work here! This is NOT your pristine note. Income too low; DTI too high. I would expect to see this note paid-in-full when the first payment is due.
My idea is to get rid of API altogether. LC and Prosper should come up with their own Auto-Invest tools that let everybody compete on a level playing field. At "feeding times" some algorithm allocates the loan to every investor who wants it and if oversubscribed will allocate on a pro rata basis with even the smallest investor getting at least a $25 piece. The system may need to delay completing the investment by 10-15 min after loans are released so that it can determine how much everybody receives.
Either that, let the retail investors get first crack, letting them buy as much as they want for the first few hours. Investors can get whatever they want after that. The government does something like this for bonds and munis.
I don't think any of these approaches would reduce loan demand in any way, which is all LC and Prosper care about anyways.
Here's an account of something I experience all the time in real life which touches on all these issues. Investor-to-loan ratio, bidding, and institutions:
In my state we have tax lien sales once a year. If a deadbeat doesn't pay his property taxes, investors go to the tax sale and pay the taxes and in return receive 24% APR if the deadbeat eventually pays up. If he doesn't pay you get his house after 2 years. Assuming he doesn't torch the place first.
Even in the smallest counties, there's an auditorium crammed with investors and sometimes only a few hundred liens to go around. You are lucky to get $2k invested if you walk in there playing by the rules like an Average Joe. With the liens in such short supply, here's how the counties deal with that:
For each lien, the computer picks an investor at random and his name is called. He can either take it, or pass, in which case the process continues until someone takes it. If you want more liens, the only way to increase the probability of your number being called is to have more accounts. But that requires bodies in seats, and the names have to be registered well ahead of time. So what the large institutions do (I'll pick on Wells Fargo here) is create tons of LLCs. "WF Sharks 1 LLC", "WF Sharks 2 LLC" etc. Sometimes hundreds of shell companies. The purpose of the LLC is so you don't need to know what individual will be holding that bidder card ahead of time. Then they hire drug addicts off the street for the day for $20 or whatever and give everyone a sheet of instructions with the properties to skip. You should see the parking lot the hour before the event starts, with all these institutions trying to corral their hired hands and toss out the ones that showed up drunk, etc. Quite a sight!
Here's the wrinkle:
The law allows a bidding process where you can bid down not the interest rate, but the percentage of equity in the deadbeat's house. It starts off at 100%, but any individual can shout out a lower amount, say 75%, and override the random computer selection. This means that after the 2 year period you don't get the guy's house; you can only force a sale and take 75%. I imagine that gets quite messy but I've never been through it.
Every year there are lots of new eager investors who have read the law and show up all prepared to do it up right, but without ever having been there before. Invariably, the very first lien to be called, one of those new guys will shout out "99%" to start the bidding. Well, what happens then is one of the Wells Fargo guys will stand up and holler "1%!!!" and take the lien. This lien's now all but worthless if the deadbeat doesn't pay and they all know it. The purpose of this is to make it very clear to everyone in the auditorium that if you want to be a "jerk" and bid things down, then you're going to walk out of there with worthless pieces of paper. It takes about 5-10 more liens like this before folks figure it out and shut up and take what is assigned to them. Obviously the county treasurer / auctioneer has to endure a lot whining about institutions during the break periods.
Core....I got a great laugh out of that story!
What you need to do is show up the night before with like 40 bottles of really cheap whiskey. Make several stacks of like 5-6 bottles in different areas where they can easily be found. Once one or two people find a bottle, word would travel quickly, and nature would take it's course. It would almost be like an Easter Egg hunt.
Main points from the story:
1. Bad bots
I am wondering whether the big guys themselves have complained about this to LC. Presumably, they are also trying to frontrun each other to get at the best loans at the quickest possible time. They can't possibly not notice what we have been noticing--that some folks with deep pockets are gaming the system.
One thing that Rob L did not consider is that there are many investors using the API now and most of these people have sophisticated credit models. The data set to build a model is finite and so many investors will be choosing the same notes.
A much more likely explanation is that 1-2 large investors came in and 2-3 small investors using the API filled the note.
Well, it wouldn't be the first time I was wrong and Peter that's certainly a plausible explanation. When people see me wearing my tin foil hat I do get the feeling it damages my credibility a bit.
Maybe I don't understand well enough how the API works. A couple of large investors love the loan and each want play by the rules and get their legal limit of 75%. They both put in a buy order for $4,125 at lightning speed. The first one there gets the $4,125 and the second one gets what's left; $1,375. The LC server does the subtraction and only gives the second guy what it can. End of loan. Okay, I can understand the two large buyer scenario, but we have 4. Maybe one large and three small, but with $1,375 left after the big guy the little guys are still looking at notes about $450 each (not small to me). I have trouble figuring out how this pie was sliced into 4 pieces in 10 seconds by 4 "people" that don't know how much the other is buying. Additionally all four buyer's loan evaluation models had to agree this was one terrific note; something that rlv99 doubted in an earlier post in this thread.
It's easier for me to believe this loan was carved up into 4 pieces and bought by a single entity in a single API transaction. Imagine you are managing money for a few dozen wealthy clients. You have a LLC set up for each client, each with its own EIN, brokerage accounts, etc. You think P2P lending is a dandy investment for some of your clients so you open LC accounts for each of them, get a computer, sprinkle with software and begin purchasing notes. One computer, one API transaction, four notes for four client accounts. Nothing sinister going on here. Everything's on the up and up. Now change things just a bit. Rather than managing money for four individuals, you are a single wealthy individual or corporation with four LLC's (I know people who have had dozens). What if you're a boutique hedge fund with 100+ LLC's. We've gone all the way from everything's on the up and up to what I guess most would say is an end around the 75% rule. Either way no individual investor had a chance to buy any of this fractional loan. Even if I'm wrong about this specific loan, the scenario I described isn't one that's only for us tin foil hat guys. It renders the 75% rule totally ineffective.
In my original post I questioned whether or not this loan, if purchased by a single entity as I posit, breaks the 75% rule. A few days later I put the following on Bryce's "Fairer Solutions to Excess Investor Demand" thread:
The first order of business is to make LC aware that the 75% rule is not working as they initially intended (promised), and find out if they are interested in fixing it
I think we are having this discussion because we believe the 75% (70%?) rule as it was envisioned by LC does not work. LC may disagree, but several ways around this rule for larger investors have been discussed. Many of us believe these methods and perhaps others we haven't imagined are being used today to skirt the rule. My understanding is that LC put this rule in place specifically for the benefit of the small retail investor in the spirit of compromise at the same time they implemented the whole loans program for the benefit of large investors. All we are asking LC to do is to fix the 75% rule to prevent its circumvention and abuse by larger investors. We are not asking for anything new; only something that works as LC intended in the first place.
LC is drinking from a fire hose of investor demand and growing like crazy. There's that IPO thing out there and I agree that if anything is to be done LC must see it as very low risk and not a chance to mess up the good thing they have going. Only if LC is even willing to engage in a discussion concerning the broken 75% rule need we take the next step and propose solutions. I do believe that a few simple changes as suggested in this thread and perhaps others, would go a long way toward restoring the operation of the rule as LC intended in the first place.
There is one positive aspect to all of this. You can look at the note inventory about half an hour before one of the 'drop times' and see exactly what notes you don't ever want to buy look like, because they're pretty much what everyone else passed on. You might even be able to develop a nice little filter that excludes notes with those characteristics.
For a while I bought notes that were recent drops that were close to selling out, figuring I might be getting the collective wisdom of many different investment approaches. But those notes are (so far) not doing much better or worse than average.
2 weeks of idle cash? Aren't there leftover notes each and every cycle? If you checked 5 mins before _and_ right after feeding times, I would think you'd be able to get 100% invested all the time. Unless you're talking about 2 weeks of LC keeping the loan in review.
Me, I have never been able to sell an issued note for more than 1.07% markup. But then again I don't hold out at a fixed price for weeks because I don't want to get stuck with it. Taking into account the 1% fee I'd be barely breaking even best-case. My mean price is -0.32% so I would lose money each week if I tried this. Or wind up getting stuck with the notes if I was firm on my asking price.
If one wanted to take a tiny bit of risk it would be wise to hold onto it until one payment has been made because then the price really jumps. These are A-B notes anyway we're talking about, so I assume the number of 1st month defaults is rather low.