Author William Faulkner said, “In writing, you must kill your darlings.”

Faulkner’s advice to writers wasn’t directed at brevity, but at the discipline of abandoning a phrase or idea that you deeply love if it no longer serves the story.

Hemingway was more direct. “The most essential gift for a good writer is a built-in, shockproof, shit detector.”

These guys would have been great investors. We would love to have them on our team as we continue to pursue the core of our investment thesis—the inevitable and irreversible digitalization of financial services—by killing darlings that don’t really serve it. We would welcome their cynical view that even the ideas we love most, the stories we most believe, are liable to fall out of favor, even if temporarily, and to no longer serve our investors.

For instance, we love what digitalization has done to professionalize and institutionalize segments of residential real estate lending, and we think we are still in early innings for the industry writ large. But, as we suggested in our recent essay, “Hello Universe,” noise – in this case about the fix-and-flip mortgage opportunity – has escaped our digital galaxy and attracted the animal spirits of banks and other “big money” investors.   They have started to bid the asset class to a level where the reward doesn’t justify the risk, so we put that darling on the shelf as it no longer made sense for us to participate. Going to the sidelines does not mean we are killing our thesis; it means that our love is not unconditional.  When conditions improve, and the opportunity remerges, we will welcome our darling back with open arms.

We believe digitalization will continue to burrow deeper into the areas it has already disrupted and spread wider into areas that still seem invulnerable.   One of our newest darlings is one that seemed unavailable to digitalization: loans to students of non-accredited, educational programs, such as nursing schools and technology boot camps. We are confident there will be other new digital opportunities, but we are focused on identifying those that will not only disrupt but also be able to build and sustain a viable business over the long term.

This discipline, this belief that disruption inevitably leads to more disruption, drives our commitment to always keeping our options open, to keeping enough powder dry to participate in new opportunities and to step aside from opportunities that have ripened and turned.  This is the fundamental yin and yang of foresight – equal parts enthusiasm and cynicism – and a willingness to trade a broken heart for an unbroken investment any day.

Even early consumer credit marketplace lending – our very first darling – had its turn in the penalty box. In 2016, corporate leadership challenges, funding pressures, and credit underwriting breaks disappointed many investors, but the industry rebounded and came out stronger as governance, compliance and oversight were strengthened.  Furthermore, companies were rationalized and right-sized, and credit underwriting models upgraded.  Just as digitalization provided the basis for this business, it also provided the tools and resources to help identify problems and address them systematically and quickly.

Inevitable means relentless and unsparing. Digitalization will turn back on its own invention and find new ways to make it even more efficient, to harness and manipulate even greater fields of data. The startups it creates must stay humble enough to be committed to their own disruption and guard against newcomers who discover better processes.

Here’s one darling that yet another author, Charles Dickens, fortunately did not kill. “Change begets change…The mine which Time has slowly dug beneath familiar objects is sprung in an instant; and what was rock before, becomes but sand and dust.”

Industry Update

As the U.S. economy continues to expand, unemployment shrinks, and interest rates tick upward, the digital finance sector continues to thrive with fresh capital flowing to a new wave of startups and with established players turning in impressive performance. Following is an overview of the macro picture and its impact on digital finance, as well as a general update on the sector.

The U.S. economy continues to expand. As the New York Times put it: “To find the last time the jobless rate was lower, you have to go back to when Richard Nixon was in the White House, when men had only recently walked on the moon, and the Beatles were still together.”¹

In September, the unemployment rate fell to 3.7%².  Median household income continued to rise in both real and nominal terms, with a new annual high in 2017.³  Headline average hourly earnings climbed 2.8% year-on-year (“yoy”).  Real median usual weekly earnings for wage and salary workers 16-years and older saw a more muted rise during the third quarter, yet this inflation-adjusted wage indicator is up about 6.3% through 2Q18 from the post-Great Recession low in 2Q14.⁴  With employment full, incomes up, and real wages trending higher, it should not be a surprise that U.S. consumer confidence continues to climb to an 18-year high.^5,6

On the corporate side, S&P 500 earnings are expected to grow 19% yoy in the third quarter, the fifth straight quarter of double-digit growth.⁷  Small businesses also continue to thrive with the NFIB’s Small Business Optimism Index reaching 107.9 in September, the third highest reading in the survey’s 45-year history.⁸  U.S. home prices continue to rise with the S&P/Case-Shiller national index up 6% yoy in July.⁹

Most credit has been performing well. Viewed through a digital loan lens, this healthy economic backdrop has been supportive of credit performance in general as most gauges of consumer, small business, and residential real-estate credit continue to be benign. Government sponsored student loans and auto loans, on the other hand, have seen higher delinquency rates, which we attribute to structural reasons and underwriting practices inherent to those sectors.

While default statistics for U.S. consumers have increased somewhat, a normal occurrence in a growing economy, they continue to remain low by historical standards.  Bloomberg reported in September that the average FICO score reached 704, its highest point ever following eight straight years of increases following the Great Recession.¹⁰

• Revolving consumer credit charge-offs across all U.S. commercial banks were 3.7% as of the second quarter¹¹, in line with the first quarter.
• Paynet’s Small Business Default Index was at 1.9% at the end of August 2018, flat¹²
• Single-family residential mortgage delinquency was 3.3% across all U.S. commercial banks as of the second quarter, the lowest level since the end of 2007.¹³
• Student loan delinquency rates (90+ days) remained elevated at 10.9% at the end of the second quarter, though in line with the 11.1% average levels seen over the past two years.¹⁴
• Auto loan delinquency rates (90+ days) were 4.2% at the end of second quarter 2018, a slight quarter on quarter (“qoq”) improvement but an increase of 25 basis points (“bps”) yoy.¹⁵

Higher short rates continue to be a tailwind for investors in the digital loan space.  In September, Prosper raised borrower rates by an average of 14 bps following the Fed’s rate hike, resulting in a cumulative 147 bps rise in rates since the beginning of the year.¹⁶  Additionally, LendingClub raised borrower rates in August, marking the third increase this year for a total increase across grades between 49 and 109 bps.¹⁷  Greensky, a publicly-listed digital finance provider of point-of-sale solutions, has raised rates three times over the past year in some merchant categories.¹⁸

Higher borrower rates are not deterring platform loan demand.  In the second quarter, LendingClub reported record personal loan originations of $2.8 billion, up 31% yoy and applications rose 50% yoy despite higher borrower rates.¹⁹ Cumulative originations at LendingClub have exceeded $38 billion.²⁰ Cumulative origination volume at Marcus by Goldman Sachs exceeded $4 billion at the end of the second quarter, an increase of $1 billion or 33% qoq.²¹ At SoFi, personal loan volume was approximately $2 billion in the first half of 2018, with cumulative personal loan volume up 20% to $11.9 billion.²²

As we’ve noted previously, our view is that higher platform borrower rates should not be an impediment to loan origination growth.  We expect this trend to continue for two reasons as consumers seek a more cost-effective solution to managing their debt.  First, platform borrower rates tend to be 300-500 bps lower than credit card rates given that in most cases the platform loan is used for refinancing. Because it is fully amortizing, the platform loan targets a zero-loan balance at the end of the term.   Second, other competing loan products are increasing borrower rates at a faster pace. For example, interest-bearing credit card rates rose 137 bps yoy through May 2018²³, with 1-month Libor increasing 94 bps over the same period.

Robust digital finance ABS issuance continues to validate the sector’s acceptance by institutional investors.  Consumer and marketplace loan asset-backed security (“ABS”) issuance totaled approximately $2.5 billion in the third quarter, up 9% yoy according to data provider Finsight. Year-to-date issuance was $8.8 billion. Spreads on senior tranches continue to trend lower averaging about 74 bps over comparable swaps in the third quarter, slightly lower than 81 bps during the second quarter for comparable deals.²⁴ During the quarter, rating agencies upgraded 52 tranches of consumer ABS deals, speaking to improved credit conditions and asset acceptance.²⁵

Profitability and cash flow: Platform financial health continues to trend up. The publicly-traded U.S. digital finance firms continued to report solid growth in loan originations and earnings during the second quarter.

• LendingClub originations exceeded $2.8 billion, adjusted EBITDA was $25.7 million, and 2018 adjusted EBITDA guidance was reaffirmed at $75-$90 million. Cash at quarter end was about $492 million.^26,27
• Square Inc., management reaffirmed 2018 adjusted EBITDA guidance of $240-$250 million and noted that Square Capital’s origination volumes increased 22% yoy with over 60,000 new business loans totaling approximately $390 million in the quarter. Since inception in 2014 and through 2Q18, Square Capital has extended over $3 billion to small businesses.  Cash at quarter end was $1.8 billion.²⁸
• OnDeck originations grew 26% yoy, the net charge-off rate declined to 11.2% from 18.6% yoy, cost-cutting initiatives helped deliver $10.0 million in quarterly adjusted net income, and 2018 adjusted net income guidance rose by 43%²⁹.

Funding Circle goes public in September.   Funding Circle, the UK-based small and medium size business lender, raised £300 million in its IPO at a valuation of £1.6 billion³⁰.  We look forward to following its story as a public company.

LendingClub announces conclusion of investigations by DOJ and SEC.  On October 1, LendingClub announced the resolution and closure of the DOJ and SEC investigations into actions taken by former executives in 2016, marking the end of this saga for LendingClub.  Importantly, the SEC’s Enforcement Division determined not to recommend charges against LendingClub Corporation which promptly self-reported certain executives’ misconduct.³¹

Macro Outlook

Applying foresight to the macro context demands that we look into the crystal ball of leading indicators.  Students of the business cycle know that historically, a set of leading indicators, such as an inverted term structure, the housing market, high yield credit spreads, industrial metals prices, and money supply growth, have been predictors of the end of economic expansion. Policymakers are likely pondering these indicators as they formulate the future path of short-term interest rates.

Is the rate term structure signaling the end of the business cycle?  The yield behavior between the U.S. government’s short-term obligations (i.e., the 1-year Treasury note) and long-term obligations (i.e., the 10-year Treasury note) continues to diverge.  Despite a September sell-off in the U.S. 10-year Treasury, longer-dated yields seem to be telling a different story than short rates:  the Fed has raised the fed funds rate 75 bps this year, yet the spread between 1-year and 10-year Treasuries (i.e., the “1s/10s” spread) has collapsed to about 50 bps at September 30, 2018, half the 100-bps level of one-year ago.  In early September, that spread was less than 40 bps – the lowest level since the fall of 2007.  If the long-end of the curve remains near current levels, future rate hikes could invert the yield curve, possibly signaling an end to one of the longest business expansions on record. We are monitoring the term structure dynamic.

The housing paradox. Another leading indicator that we are observing closely is the housing market.  At a high level, the housing market is still in good health – supply is in check, mortgage delinquencies are low, and home prices are appreciating.  But after years of steady and prudent growth, the housing market may start to become a victim of its own prudence and of the general economy’s health.  Home price appreciation will likely slow as interest rates rise — the average 30-year conventional fixed-rate mortgage ended September at around 4.9%, up from under 4% a year ago. Existing home inventory, which has been low for the past several years, is just starting to trend up as current owners may be looking to express their confidence by trading up – generally by selling an existing home to buy a new home.  Or they may be looking to lock in some capital gains by selling their home and moving to a rental.  Both scenarios contribute to new supply of existing homes, which could start to decelerate home price appreciation.

The policy response: soft landing?  We believe that one of the few policy tools available today to help manage an end of the business cycle is monetary, and so we must consider the Fed’s reaction function to the same leading indicators that we are observing. Based on the Fed’s economic forecasts of slowing GDP growth in 2019 and 2020, we suspect that the next phase of the Fed’s engineered slowdown is already on the agenda. Having seen the early results of its systematic tightening evidenced through higher mortgage rates and tighter real money, the Fed is likely to feather the throttle with further — as needed — bumps and try to land the economy as smoothly as possible over the medium term.