Monat: Februar 2017

Even Good-Guy Student Loan Startups Still Favor the Rich

Last February, the online lending company SoFi paid $5 million for a 30-second ad during the Super Bowl. The spot begins at a busy city crosswalk, panning from person to person as the narrator assesses their worth. “Jim is great. Sarah is not great at all,” a male voice intones as the focus swoops from a white dude to a white woman. “This guy? Never been great,” the narrator continues, as the camera settles on a smiling bro, who has no idea he just failed a financial test. The commercial ends with an order: “Find out if you're great at SoFi.com.”

That wasn’t where it always landed. The original version of the ad included three more words: “You’re probably not.” But at the last minute, SoFi cut them. The message, a spokesperson told Adweek, wasn’t “authentic” to the company’s image.

youtube.com

The line may have sounded too crude for national TV, but it was actually a perfect encapsulation of SoFi&039;s brand. Most people aren’t in great financial shape, and SoFi was built around identifying the best and rejecting the rest. Other Silicon Valley-backed startups specializing in student loan refinancing, including Earnest and CommonBond, followed in SoFi&039;s wake. All of these online lenders promise a faster, easier, more data-driven way to refinance student loans online, but they can only offer cheaper rates by courting the kinds of customers who have no trouble paying off their loans in the first place: doctors, lawyers, pharmacists, and MBAs in their early to mid thirties, who went to good schools and have good jobs but are still paying off major loan balances. SoFi even has a catchy nickname to describe its ideal demographic: HENRYs (high earners, not rich yet).

Last spring — after biting its tongue at the Super Bowl — it plastered public bus stops in San Francisco, a city still suffering an affordable housing crisis, with posters that said, “10% down. Because you’re too smart to rent.” In New York, it greeted subway riders with an ad that said, “Six figures is for your salary, not your student loans.”

Marketing for student loan refinancing on Earnest’s website. / Via earnest.com

These high-end customers base (and their very attractive debt) have attracted big-league investors like investment banks (who give startups like SoFi lines of credit to support their lending operations), and the insurance companies, long-term mutual funds, and hedge funds that ultimately buy the loans. (Just like mortgages, student loans can be pooled together so that they are easier to sell. Investors are buying the future income stream of monthly repayments and interest.) In May last year, SoFi received Moody’s highest rating, AAA, for a $380 million bond backed by student loan repayments. On Thursday, SoFi became the first student lender to get a AAA rating from S&P, this time for a $561 million bond. SoFi has refinanced $9.76 billion worth of student loans since it launched and offered $2.7 billion in bonds in 2016 alone.

BuzzFeed News

President Donald Trump has hinted at potential changes to the student loan system, including increased reliance on private lenders over the federal government. Ben Miller, senior director for postsecondary education at the Center for American Progress, told BuzzFeed News that he doesn’t “expect much leadership” from Betsy DeVos, Trump’s pick to run the Department of Education, who also happens to be an early investor in SoFi. “There&039;s no evidence Betsy DeVos knows a single thing about student loans and any major change would have to come from Congress,” said Miller.

But Silicon Valley has not been waiting around for Washington. In their buzzy ad campaigns, SoFi, Earnest and other fintech startups say they want to help fix the student loan crisis by bringing Silicon Valley-style meritocracy to one of the oldest financial instruments in the world, the loan. In practice, however, private student loan refinancing looks more like an updated version of the same wealth management services that have always catered to the rich — except these startups capture customers while they’re still young.

In some ways, especially in the eyes of critics, these companies&039; closest analogue isn&039;t a traditional bank but a startup like Uber — but instead of on-demand private drivers undermining the public transportation system, it’s refinancing for the rich masquerading as a solution for the dire student loan crisis that has left most graduates burdened with debt for decades to come. Last year the Wall Street Journal published a profile of SoFi called “The Uberization of Banking,” which asks and answers its own question: “But isn’t SoFi cherry-picking loans? Absolutely.”

The problem these startups purport to solve is, inarguably, a huge one. Forty-four million Americans currently owe more than $1.4 trillion in student debt. That’s $1.4 trillion dollars hanging over 44 million heads, and, for those who can’t repay their loans, it’s a lifetime of ruined credit scores and dodging collections agencies.

Most of that debt comes from federal student loans, which all have the same, relatively low, interest rate, assigned by the government each year. SoFi, Earnest, and CommonBond’s innovation is promising a rate even lower — as low as 2 or 3% — though only to the kind of borrower who can reliably be trusted to pay off such a loan. SoFi CEO Mike Cagney got the idea for the company when he learned that Stanford MBAs almost never default on their student loans. When the company launched in 2011, its services were only available to this group.

But though sorting would-be borrowers by school alone may have been a simple way to identify HENRYs, it was a controversial one. Consumer advocates, regulators (and the aforementioned Journal article) have called this selection process “cherry-picking” or “cream-skimming” the best borrowers from the federal loan program. So these days, the pitch to both consumers and the press is less about borrower pedigree than data science, algorithms, and detailed new metrics for credit scoring (such as replacing “debt-to-income ratio,” a benchmark used by traditional lenders, with “free cash flow,” a more granular criterion that shows money available after expenses like rent). Take Earnest, which is sort of like the Lyft to SoFi’s Uber — smaller, friendlier, and less likely to dominate the world. The company says its “merit-based” lending algorithms analyze 80,000 to 100,000 data points per client. The average Earnest borrower shares read-only access to six accounts, including their bank, investment, and retirement accounts.

According to the company’s CEO, Louis Beryl, this allows Earnest to price more accurately and ultimately lend to more borrowers, identifying worthy candidates across a wider credit spectrum. By scanning tons of data, the logic goes, online lenders can see past a blip on your record. Earnest&039;s website, for example, claims that profile information “is run through a series of predictive analytics and algorithms to find people who show great financial responsibility and potential.” And Beryl is fond of saying his company has the least biased lending algorithms in America. “You could be low-income, but if your expenses are very low and you’re a good saver that’s fine,” he told BuzzFeed News. “You could have gone to any school, you could have any type of employment. What we think is that our system is actually much more fair.”

CommonBond, which introduced a “multi-variate underwriting model” in 2015, told BuzzFeed News a similar story. The company’s goal “has always been to have the broadest impact possible on the student debt crisis,” said Phil DeGisi, CommonBond&039;s vice president of marketing. “We want to help as many people as possible save.”

BuzzFeed News

But although the marketing has changed, the demographics have not. Ratings reports from the past four months show that the average Earnest borrower is a 32-year-old with an annual income of $143,447 and monthly free cash flow after expenses of $4,524. CommonBond’s average borrower is 33 years old with an annual income of $159,028 and $5,996 in monthly free cash flow. SoFi&039;s average borrower, in the new bond with the AAA rating from S&P, is 34 years old, with an annual income of $170,260 and free cash flow of $7,088. (Most graduates saddled with student loan debt don’t fit that description, which is why applicants for private refinancing often need their creditworthy parents to cosign, a caveat that doesn’t get mentioned in the ads.)

In fact, one former Earnest employee, who spoke to BuzzFeed News on the condition of anonymity, said that promoting detailed data analysis made applicants with a mark against them feel like they had a better chance. “Our team was coached to be deliberately obscure when it came to explaining the eligibility requirements,” such as FICO scores, the ex-employee said. “Not being able to be completely honest with the person on the other end of the phone was the hardest part of my job, especially when it meant giving them a sense of false hope, wasting their time, and further damaging their ability to obtain credit.”

As it turns out, building a more equitable way to refinance student loans — while still turning a profit — is harder than it sounds. All lenders are in the business of being paid back, of course. But unlike banks that can easily use money from deposits, refinancing startups need to be able to borrow money cheaply in order to buy the loans in the first place. And those traditional financiers expect traditional metrics, not “free cash flow.” To hear critics and consumer advocates tell it, new metrics are more about window dressing to differentiate fintech companies from banks — and from each other — than about substantively changing underwriting.

Kevin Reed, chief operating officer at Peer IQ, a risk analysis firm focused on online lending, said the emphasis on new metrics is aimed at venture capital investors, not institutional investors. “When you’re pitching Silicon Valley, you need an angle, some competitive differentiation,” he said.

Data via Consumer Finance Protection Bureau.

BuzzFeed News

Others are skeptical that testing different metrics can be meaningful when the borrowers are so high-quality to begin with. “When you’re only playing in the very very top of the market, a lot more data really is not going to influence your decision,” said Brendan Coughlin, executive vice president of consumer lending at Citizens Bank, a Boston-based bank that is quickly gaining ground on SoFi in student loan refinancing. “It’s going to be hard to get to a point where you’re not going to approve [applicants] based on other information.”

In fact, one of the strongest signals that these online lenders are focusing on elite borrowers is the fact that banks like Citizens, which jumped in to compete in the refinancing ring, have a similar customer base but don’t use cutting-edge technology to find them. On an earnings call in July, Citizens’ CFO, Eric Aboaf, reassured an analyst about the quality of refinancing customers. “We&039;re talking about undergraduates to colleges that you and others and we all went to, MBA graduate degrees, doctors, lawyers, business degrees, that kind of thing,” said Aboaf, who graduated from Wharton and MIT.

Mark Huelsman is a senior policy analyst who focuses on higher education at the think tank Demos. “In an era of entrenched inequality and lack of upward mobility,” he told me, “the same things that would ding a borrower’s credit — a bout of unemployment, an inability to pay a student loan, an unlucky medical history — are the same things that any private lender would be looking at in approving a new loan.” (Refinancing federal loans with lending startups can also make borrowers ineligible for federal loan forgiveness programs, as well as the ability to discharge the debt in case of death or permanent disability.)

The way Huelsman sees it, these private lenders could end up compounding existing disparities, “if we allow a system to perpetuate where high-income students, primarily white students with graduate degrees, are receiving better rates.” For example, those borrowers may be able to get more favorable terms on a mortgage, whereas working-class students can’t access the same financial instruments. “That’s sort of the American story — wealth begets wealth.”

In Silicon Valley, the Uber approach — starting with town cars then moving down-market — is a respected means of disruption. When it comes to lending, however, the idea of catering to the luxury sedan sector of debtors undermines vital equal-opportunity protections, especially if the mass-market option never arrives. (Imagine if a new neighborhood bank advertised itself as serving no one except thirtysomething doctors and lawyers.)

Online lenders are not as restricted as banks when it comes to asking for information about their borrowers. The startups are regulated by the Consumer Financial Protection Bureau, which enforces fair lending laws to prevent discrimination based on age, race, gender, or country of origin. However, banks, which are FDIC-insured, are also prohibited from redlining, or discriminating against borrowers in low-income neighborhoods.

Imagine if a new neighborhood bank advertised itself as serving no one except thirtysomething doctors and lawyers.

Quelle: <a href="Even Good-Guy Student Loan Startups Still Favor the Rich“>BuzzFeed

This Website Wants To Be The Snopes Of WhatsApp Hoaxes In India

Akash Iyer / Via BuzzFeed India

Shammas Oliyath has spent every lunch break for the last six months telling strangers all over India that a Gujarati woman didn’t really give birth to 11 babies at once, malicious Indian grocers aren’t really selling AIDS-laced fruits, Guinness hasn’t really declared Kannada the world’s oldest language, and the UNESCO certainly hasn’t named Narendra Modi as the world’s best Prime minister.

“It’s a social service,” he said. “I feel really good about clearing people’s misconceptions.”

Oliyath, a software engineer at IBM in Bengaluru, is the co-founder of Check4Spam.com, a website that focuses on fact-checking and busting viral hoaxes, urban myths, and political propaganda that are spreading on WhatsApp and rapidly becoming India’s own fake news crisis.

“We are hoping to become the Snopes of India,” Oliyath told BuzzFeed News, referring to the San Diego-based website that, in its 20-year history, has evolved from busting urban legends (Does a colony of alligators make its home in the New York City sewer system?) to fact-checking America’s 45th President himself. “We want to take the work Snopes has done and apply it in a very Indian context.”

“We want to take the work Snopes has done and apply it in a very Indian context.”

Doing that not only means debunking Indian hoaxes but also doing it on the very platform where they originate: WhatsApp. The Facebook-owned instant messenger is used by more than 160 million Indians and is by far the fastest way that misinformation spreads in the country. Last year, the Indian state turned off internet access in large swaths of the country to prevent WhatsApp rumor-mongering from inciting tensions.

Check4Spam provides a dedicated phone number for people to forward any hoaxes they receive directly over WhatsApp. On a typical day, this hoax-busting hotline gets between 60 and 70 forwards to fact-check.

Oliyath works methodically through each forward he gets, sending back links if the rumor in question has already been busted on his website, and trawling the web to verify new ones.

He usually skips the first few dozen pages of search results and starts searching from the back “because that’s often where the original real post or image on which something fake is based on exists.”

Often, he relies on what India’s mainstream press has already reported, but says that he will frequently double and triple check even traditional sources to prevent any inherent media biases from tainting his debunking.

“Sometimes, we’ll get a lot of a certain piece of fake news or a hoax, so we can actually tell which hoax is trending on WhatsApp on that day,” Oliyath said. “WhatsApp is a good barometer.”

“We can actually tell which hoax is trending on WhatsApp on that day.”

This ability to spot patterns in hoaxes is particularly useful. In the last few months, for instance, Oliyath has noticed a particular kind of hoax gaining popularity: fake promotional messages that promise free cellular data and voice minutes (including this gem where President Trump gives every Indian free mobile minutes) in exchange for clicking on a link or installing an app that inevitably turns out to be malware.

“I’m an English-speaking software engineer and I’m fairly savvy, so I can tell that things like these are fake,” said Oliyath. “But a lot of older people, early smartphone adopters, and people who don&#039;t read or speak English in India are often unable to tell that these promotions are fake and end up installing malware on their phones.”

Worse, Oliyath discovered that a significant number of his non English-speaking users often ended up mistaking his English debunk itself for a genuine promotion and ended up falling for it anyway. So now he writes “fake” in half a dozen Indian languages on these posts to make sure that users who don&039;t understand English know it&039;s a hoax.

Some rumors, like a recent one about buffalo-headed fish found in an Indian river, are fairly easy to bust: Oliyath ran the picture through a reverse image search and instantly found the original one (a regular fish, in case you&039;re wondering). “Most of these guys are pretty bad at Photoshop&#033;” he laughs.

Others are harder. Last year, when a WhatsApp forward about 275 job openings in Indian IT giant Wipro started doing the rounds, Oliyath had a Check4Spam volunteer call and email Wipro’s HR department to check if the news was true (it wasn’t).

“It’s a lot of legwork,” said Oliyath. “It’s tough to do it at scale.”

That’s the reason why Check4Spam recently started accepting debunks from volunteers over WhatsApp. “We allow anyone to volunteer,” said Oliyath. “But I do scrutinize volunteer-submitted debunks before posting them on the website.”

Oliyath said that the site currently receives half a million pageviews a month, driven largely by word of mouth (Snopes can get 2.5 million in a single day). Before Bal Krishan Birla, the site’s other co-founder came on board in July, Oliyath had been struggling to figure out a way to grow it. Birla, a serial entrepreneur and an SEO expert decided that staying topical was the key to growth.

When J Jayalalithaa, a prominent Indian politician, was admitted to a hospital in a critical condition in December, for instance, the duo stayed focused on debunking hoax messages and photographs about her death days before she actually passed away. “Once people receive a WhatsApp forward, they want to know whether it is true or not and they invariably end up looking it up on Google,” Birla told BuzzFeed News. “So SEO is important for us to grow.”

Birla lets Oliyath focus on the actual debunking and calls himself Check4Spam’s tech guy, focusing on keeping the website up and running. But he’s also drawing up a roadmap: he would eventually like to build a browser plugin to detect Indian fake news on the internet. And if WhatsApp ever lets third-party bots hook into it like Facebook Messenger, he thinks that building a fact-checking bot for India&039;s most popular instant messenger would be a terrific use case.

A fact-checking bot for India&039;s most popular instant messenger would be a terrific use case

For now, Check4Spam remains a labor of love. Both Birla and Oliyath said that they’re not looking for funding or revenue yet, mostly because their real jobs keep them busy, but might think about hiring one or two more fact-checkers to ease their load. The real motivation, they say, comes from the feedback they get.

“People are really overwhelmed when they actually send something over WhatsApp to our hotline and promptly receive a response,” said Oliyath. “I’ve had elderly strangers who are obviously new to WhatsApp thank me profusely for our service. Even if the website doesn’t grow or turn into anything significant, I’ll still bust hoaxes on WhatsApp for them.”

Want to verify a WhatsApp forward? Send it over to Check4Spam’s WhatsApp hotline at +919035067726.

Quelle: <a href="This Website Wants To Be The Snopes Of WhatsApp Hoaxes In India“>BuzzFeed