Scandal At Lending Club: Caveat Emptor

New allegations of scandal at Lending Club are deeply troubling and have implications for Luchadores investing in P2P loans.

I’ll discuss the allegations, outline the basic economics at play and talk about how I’m handling this new information in my own portfolio.

A Lending Club To The Head

For those not familiar with what has become a saga of deafening stupidity over at Lending Club, let me give a brief overview.

Scandal At Lending Club: Caveat Emptor
Scandal At Lending Club: Caveat Emptor

A few months ago, LC’s CEO resigned in the wake of a scandal involving less-than-crystalline disclosures about stuff that should have been disclosed very clearly.

That cratered the stock price, led to a slew of difficulties for the company and shook the foundations of trust LC had been building for “marketplace lenders.”

As a public company and the largest of the marketplace lending platforms, LC had become a champion for the entire industry and in many ways the industry’s bellwether. Its (first) scandal accordingly bled out to affect other, similar enterprises, some of which were forced to alter growth and profit projections and to lay off employees who pretty much just got caught in the crossfire.

Nice work, LC douchebags.

But the new CEO stepped in with a clear strategic imperative to shore up trust and eliminate shoddy internal controls and restore order to operations. And it appeared he’d at least kind of steered the ship out of the choppiest waters.

Until a few days ago.

Sims City

On August 18, 2016, Bloomberg Businessweek published an article outlining the work of a “fanboy” of Lending Club who’d been mining LC’s loans data.

The guy, named Bryan Sims and hailing from the Rose City, was an LC investor who developed a software package to help rate loans offered by LC so individual investors would have independent data about the likelihood of earning profit from each loan.

In the course of his work, he discovered something not only curious but deeply troubling, if indeed true.

He allegedly found that some borrowers on LC’s platform weren’t just signing up for a single loan. They were applying for and receiving numerous loans (which is a violation of LC’s policy to allow a max of two loans per borrower). And it really wasn’t just “some” loans. It was more like 30,000 of them.

Worse, this multi-loan borrowing was not disclosed by LC to prospective investors, despite obvious risk implications for the loans on offer.

As illustration, consider the simplest case of a single borrower taking just two loans. That single borrower might apply for a $25k loan and only get funded $20k. So the loan would be made for $20k. But that same borrower would then come back shortly thereafter and apply for a separate $5k loan (which violates the same LC policy linked above, requiring 12 months of payments on a first LC loan before a second can be taken). It would not be stated on info available to prospective second-loan investors that this individual already had taken one LC loan, and, in many instances, that second loan would get funded.

This same pattern also was detected for multiple loans provided to the same borrower.

So far, not real good. But it gets a little nastier.

The Price Is Risk

These individuals’ multiple loans were likely to be “priced” by LC as possessing very different levels of risk. But not the way you’d think: That second $5k loan might well reflect a lower interest rate than the first $20k loan. That is, investors not only would be unaware there was a first loan in play when investigating the second (or third or fourth) loan and determining whether to invest. Investors would also be compensated with lower interest payments on the second+ loans.

All this despite the fact that:

1) two loans being made to the same individual would be subject to the same risk factors for default;

2) by any reasonable economic standard, the second loan would be riskier to invest in than the first at the time of each loan’s offering (barring positive changes in the borrower’s “quality,” which would be unlikely); and

3) both loans’ economic riskiness would climb if the second loan were funded.

These issues pertain, and are magnified, in instances of more than two loans made to the same borrower.

Hmmm…kinda weird, and more than a little troubling, isn’t it?

Now, Sims was only able to detect occurrences of the single-borrower-multiple-loans thing by using proprietary software to slice and dice and thrice-more-knife LC’s public data. These are things that would not have been discernible to reasonably prudent and cautious investors in LC’s notes.

Indeed, some investors in LC notes have been institutions with substantial computing and analytical resources, not just individuals. It’s unlikely that even these high-powered institutions observed the issues Sims found. After all, if they had, the discovery would have cast doubt upon the LC platform and caused investments via the platform appear unduly risky from the start; the institutions never would have invested in the first place.

To really hammer the point, consider this. LC data has been subjected to analysis by IBM’s vaunted Watson AI supercomputer thing. Watson might be sweet at Jeopardy! But at finding something that seems to substantially distort loans’ true risk profiles? Not so much, it appears.

Notwithstanding the difficulty of finding these multi-loan borrowers in LC’s public data by outsiders, their existence absolutely should have been obvious to Lending Club. They should have been obvious before loans were made. And, since there are big implications for the riskiness of loans funded by investors, they should have been disclosed and pricing should have accurately reflected the risks (if the second+ loans ever should have been allowed at all).

That implies LC has sieve-like internal controls and didn’t know this stuff was happening. Which is bad.

Or it implies LC did know about at least some of this stuff and didn’t disclose it. Which is worse.

Or it implies LC knew about this stuff and concealed it (or even facilitated it) to promote its larger business interests. Which would be way, way worse.

Let me digress for a minute to explain this “concealment” notion. But before I do, please note that this theorizing is mere speculation. I nevertheless think it’s speculation worth making as we navigate the emerging facts in this alleged scandal.

First, let’s understand LC’s motivations. It wanted to grow its business by making more loans of more value and by having those loans funded by more investors. It also needed to maintain a promise of quality to its investors by winnowing loan offerings to include only reasonably good borrowers and reasonably compensable risk. And so, to sustain this quality goal, LC instituted policies about the maximum number of loans an individual could take and the requirements for second loans, etc.

So growing the business rapidly conflicted with some of LC’s own internal controls. If a single borrower wanted a verboten second (or third) loan, it could have benefited LC to enable that loan to occur despite its own rules.

But allowing a third loan to a single borrower would introduce a risk-pricing problem for LC, particularly since it couldn’t disclose that a single borrower was getting a second+ loan in violation of its quality controls.

“Concealment” could explain why second+ loans might pay lower interest rates than first loans. By pricing loans this way, LC could mask differences in risk across ostensibly similar loans seen by prospective investors on the LC platform.

Think about the questions that would be raised by prospective investors if two outwardly similar loans (one second+ loan and one first loan) both offered on the LC platform at the same time offered investors vastly different interest rates. It wouldn’t be long before investors realized there was much more to the risk story for some loans than was being disclosed in the publicly available information.

Meaning that, for LC’s second+ loan offerings to go undetected by investors, those loans’ interest rates couldn’t reflect their true risk; the rates had to be held artificially low to not raise red flags among investors. Those low rates would make LC’s second+ loans outwardly appear to be in compliance with LC’s own lending policies. Which would allow LC to grow its business aggressively (by making more and more loans, including second+ loans), all while keeping investors in the dark. Which goes a long way toward explaining the economically unreasonable pattern observed by Sims where first loans garnered higher rates than second+ loans.

Again, this concealment theory is mere speculation. But it’s not unmoored speculation. And I personally don’t like the constellation of factors that tethers this speculation to Earth.

Regardless of which of the foregoing possibilities is correct, if any, this latest scandal places in doubt whether any trust should be afforded Lending Club and its loans by foxy Luchadores.

All of which means:

Until these allegations are settled, Lending Club’s place in the Luchadorian portfolio is placed on probation. I will not be making any new investments in Lending Club notes, and I’ll implement a program of withdrawing cash as it spins off the loans I already hold.

Prosper, which operates a platform similar to LC’s, has not been implicated in this mess. And Prosper does affirmatively disclose whether borrowers are repeat customers. For now, I’ll keep my Prosper program running as it has been.

But I would like to see Prosper open the kimono to independent auditors. Prosper should allow them to review its loans data and provide an opinion of transparency and validity of the publicly-available disclosures made by Prosper about its loans to prospective investors. If Prosper’s management is smart, it has already begun this process as a step in fortifying investor confidence and, competitively, as a means to grow its business at LC’s expense. If I were helming Prosper, I’d see this LC scandal as a golden opportunity wrapped in diamond-encrusted silk, with plenty of threats strewn all around, and having just one obvious way to make the most of it: Get that audit underway ASAP, Prosper, and provide an A+ report card as soon as you can.

For P2P/marketplace loans generally, I continue to see economic merit in their role as part of a properly balanced portfolio. But these troubling allegations are a concrete reminder of very real “platform risk” that was discussed in prior FinanciaLibre posts in connection with P2P and other “marketplace” loans.

I expect to see more about this latest LC scandal in the coming days, and I think there’ll probably be lots of work for the plaintiffs’ bar as it assembles info and starts thinking about what comes next for Lending Club and investors in its notes…if there’s any substance to the findings of Bryan Sims, that is.

Luchadores, are you troubled by the latest issues facing Lending Club? Or are you of the mind that these allegations, even if true, are immaterial and shouldn’t affect investing plans?

8 Comments

Libre Your Mind