// < ![CDATA[
Khrom Capital 2015 Letter; Long OnDeck Thesis
Also see from Khrom Capital
Khrom Capital With Another Killer Year – Up 23% YTD
Khrom Capital Returns 30% While Holding 43% Cash
Investing Versus Trading: Khrom Capital Q1
Khrom Capital Has Another Strong Year With An 18% Return
Khrom Capital Returns 33% With 40% Cash Portfolio
Khrom Capital Up 1.1% In H1
Khrom Capital – The Hidden Value in Vistaprint
Khrom Capital 2015 letter
March 23, 2016
Dear Limited Partners: In 2015, our Partnership returned 1.4% net of fees and expenses.i On average, we held 7% of our assets in cash throughout this period. For the past five years, the fund has had an average cash balance of 23%.ii
As our Partners know, our philosophy is not to disclose investment ideas in our letters, in order to prevent commitment bias and to protect our intellectual property. On the other hand, it is valuable for Partners to understand the type of investments Khrom Capital makes. To balance these conflicting goals, we find it prudent to occasionally walk through a particular Khrom Capital investment.
A current investment of ours is in a company called OnDeck, which operates an online platform for small business lending. OnDeck’s story reminds us of Moneyball. (We recommend reading this book or seeing the movie.) Moneybal tells the true story of how scouting for baseball players evolved in the 21st century. It describes the transition that talent scouts made from recruiting players using their personal judgments—which were impaired by cognitive biases and lacked regard for scientific data—to instead employing computer-generated statistical analyses. The first team to implement this change was the Oakland Athletics. Having a fraction of its competitors’ budgets, the Oakland A’s were the League’s underdog. Yet, by beginning to use data analytics to more effectively recruit players, the Oakland A’s catapulted to the playoffs and permanently changed the way Major League front offices do business.
OnDeck brings Moneybal to small business lending. The company has taken advantage of advancements in computing and the proliferation of data to help solve a large problem: deserving small businesses have trouble getting capital. The traditional method requires small business owners, who have limited time to step away from their businesses, to dedicate over 20 hours to the loan application process. 1 Worse, those valuable hours are likely wasted because a majority of small firms (under $1 million in annual revenues) are rejected when they apply for credit.1 For those that do get approved, it usually takes weeks for the funds to arrive, 2 which provides little value for owners with time-sensitive capital needs.
This problem exists partly because banks lack the incentive to make small business loans. Transaction costs to process a $100,000 loan are comparable to a $1 million loan, but with less profit. As a result, banks are less likely to engage in lending at the smallest dollar level. To put this problem in perspective, over half of small businesses surveyed were seeking loans for under $100,000. 2 Furthermore, when banks do review a small business loan application, they frequently shortcut the underwriting process by primarily relying on FICO—the personal credit score of a business owner. As a result, a low FICO score is the top reason for a small business credit denial.1 However, the personal credit score of a pizzeria owner, for instance, is not the best way to predict how many pizzas the business will sell. FICO also does not help an underwriter decide on the appropriate loan size to offer and interest rate to charge each unique pizzeria.
The antiquated process that banks still rely on created the opportunity for OnDeck to disrupt the small business loan market. OnDeck has spent almost a decade building out its OnDeck Score—a credit score for the small business, not the business owner. The OnDeck Score aggregates and analyzes thousands of data elements related to a business that are more predictive of its credit performance than FICO alone. 3 This allows OnDeck to more accurately underwrite small businesses’ financial performance. Since a majority of this process is also computer automated, 4 it enables the company to profitably approve loans under $100,000. Small business owners can now complete an online application within an hour from the convenience of their offices and receive funding the same day.
OnDeck’s technology was developed by an impressive team, led by its CEO, Noah Breslow. Noah graduated from MIT with a BS in Computer Science and Engineering and received an MBA with Distinction from Harvard—an ideal educational background for the leader of a financial technology (fin-tech) company. More importantly, among the competitor CEOs whom we spoke with, Noah stands out as the most capable. We took the time to get to know Noah, and his qualities match what we look for in CEOs. He sticks to a clearly defined goal: make OnDeck the best in the world at delivering credit to small businesses. When necessary, he takes short-term pain to achieve this long-term objective. As a recent example, he reduced the company’s reliance on a distribution channel that brought OnDeck significant business, but that did not best serve the company’s customers in the long-run. Noah drives innovation to keep OnDeck ahead of its competitors: his team just pioneered the concept of “OnDeck-as-a-Service” (which we discuss further below). He is highly ambitious and greatly believes in the future of OnDeck, with virtually all of his net worth invested in the company.
In addition to OnDeck’s strong management, the company itself has competitive advantages that should enable it to win a big portion of a very large market. As the largest online small business lender, it benefits from the positive feedback loop of its data gathering. The more loans that OnDeck’s algorithm underwrites, the more data it receives on loan performance, the more it can improve the algorithm’s predictability, the more its underwriting advantage increases versus smaller competitors, the more loans it can underwrite, the more data it receives on loan performance, and so on in a virtuous cycle.
The company’s scale also drives increasing operating leverage. As an online lender, much of the company’s costs are fixed, resulting in a lower cost per loan with each additional loan it makes. Not only does this help build OnDeck’s position as the lowest cost producer, its scale allows it to maintain its lead by being able to invest more into R&D than its smaller competitors.
As a result, OnDeck can now underwrite the widest credit spectrum compared to its competitors, which gives it a cost advantage in acquiring customers. Most lenders to small business are clustered at either end of the credit spectrum—from subprime to prime—and offer only one or two products. OnDeck, on the other hand, has advanced its credit model to a point where it can now underwrite a short duration loan to a relatively new business, a line of credit product to a business with sporadic cash flow needs, or a bank-like multi-year loan to a mature business. This makes OnDeck’s customer acquisition costs relatively lower, since it can convert more advertising dollars versus its competitors.
(For those interested in a simplified example to illustrate the point: say, Competitor A can only underwrite half the size of OnDeck’s credit spectrum. Assume that Competitor A and OnDeck pay for ten sales leads. If Competitor A’s credit model is only able to underwrite three out of those ten applicants while OnDeck can underwrite six, Competitor A converts only 30% of its purchased leads versus 60% for OnDeck. Therefore, OnDeck’s higher conversion rate gives it a lower cost per new loan acquired compared to competitors.)
Moreover, having a wider lending spectrum than its competitors enables OnDeck to generate a relatively higher customer lifetime value. OnDeck can generally underwrite a small business earlier in its life before it qualifies for a bank loan. The company’s breadth of credit solutions are designed to grow with its small business customers as they mature, from their first year in business to their tenth.5 Because OnDeck can provide for a small business over its entire lifecycle by lowering rates and offering a variety of products, the owner is given little reason to look elsewhere for capital.
OnDeck’s economies of scale and scope allow it to invest additional resources in strengthening its lead over competitors. We see evidence of this in multiple areas, from customer service to innovation.
As a result of investing in customer service, OnDeck has achieved a Net Promoter Score (NPS) of 76, putting it within the ranks of the most loved businesses in the U.S. (Apple’s NPS score is 76. Amazon.com’s is 69.) Comparatively, the average NPS score for a national bank is 9. In general, NPS leaders outperform on both winning new business and cross-selling to existing customers.6
With respect to continuously investing in R&D, OnDeck is able to develop innovations such as OnDeck-as-a-Service. The company took the decisioning strength of its OnDeck Score and recently created a platform that allows other institutions to use its underwriting algorithms. Deep integration is required to make these partnerships work, which we think leads to sticky relationships and another barrier to competition. The quality of OnDeck’s first partners makes us optimistic about the potential for this service. These partners include the largest bank in the U.S., JPMorgan Chase— which has 4 million small business checking accounts—and the largest provider of small business accounting software, Intuit QuickBooks—which 4 million small businesses use.
OnDeck-as-a-Service will use technology to seamlessly download business owners’ checking account or accounting software data. Without needing to do anything, small business owners could be preapproved for a loan just by logging into either their Chase checking account or QuickBooks software. OnDeck’s technology allows banks to process applications cheaper and quicker, in hours instead of weeks. According to Chase’s Head of Business Banking, the partnership will allow Chase to “offer almost real-time approvals and same or next-day funding.”7
OnDeck recently replicated this strategic move in Australia, by partnering with that country’s largest bank—Commonwealth Bank—and the leading Australian business accounting software provider—MYOB. OnDeck is bringing its superior value proposition and competitive advantages to a very large market. Small businesses are an extraordinarily large part of the United States economy, accounting for about half its private GDP and workforce.8 In the U.S., there are currently $193 billion in outstanding business loans under $250,000. 9 Estimates show there is a potential $100 billion in unmet demand for small business lines of credit and other credit-related products. 10 OnDeck’s innovations continue to expand its addressable market. By creating new products such as its banklike loans of up to $500,000 as well as its geographic expansion into Canada and Australia, we estimate that OnDeck has just increased its market opportunity by an additional $100 billion. With OnDeck’s current loan book at under $900 million—market share at less than one quarter of one percent—the company still has a long runway to grow.
To invest in OnDeck, an investor must have confidence that its underwriting is prudent and that the business can withstand economic recessions. There are several important reasons that give us that conviction. But first, it is valuable to understand the nature of OnDeck’s borrowers. OnDeck lends to small businesses across more than 700 sectors. The company’s median borrower has annual revenue of $580,000 and has been in business for 7 years. The company lends these customers an average of $52,000. Since OnDeck began lending in 2007, its credit model has funded more than $4 billion of loans.
Because the company’s loans are very short-term—the average term is one year—we already know how the majority of OnDeck’s $4 billion in loans have performed. Except for the most recent quarters in 2015, the billions of dollars that OnDeck lent over the past eight years have completed their lifecycle. And it has been a consistent track record: OnDeck lost less than seven cents for every dollar it lent out since 2007, and only nine cents in 2008. 11
It is crucial to understand the nuances of how OnDeck differs from most lenders. Unlike a typical lender, whose borrowers are required to send in monthly payments, OnDeck gets automated electronic repayments from its borrowers either on a weekly or daily basis. A typical lender gets insight on the health of its borrowers only 12 times a year. OnDeck, on the other hand, receives data 52 or 252 times per year. (OnDeck is the only publicly traded lender that we know to report a 15-day delinquency ratio to its investors, as opposed to the standard 30 days.) By having daily or weekly insights into its borrowers’ cash flows and rarely making long-term loans, OnDeck can be nimble. OnDeck’s management team can quickly pull back on new loans, shorten terms and/or increase pricing when their daily indicators show deteriorations in credit performance.
Most importantly, OnDeck has a very strong balance sheet. It has a large cash balance (almost $200 million) that should enable it to withstand a severe economic recession. We stress tested OnDeck’s portfolio by correlating the severity of charge-offs that various credit assets, such as credit card receivables, incurred during the Great Recession. Granted another economic downturn of this magnitude would be an unpleasant ride—as it would be for almost any business—OnDeck’s cash balance should be sufficient to absorb the losses. At the same time, we think OnDeck’s smaller and less capitalized competitors would struggle to survive. And so ironically, an economic recession may actually benefit OnDeck’s market position in the long run, as a downturn weeds out the weak and bolsters the strong.
We always seek to compare our holdings to other businesses in order to better understand the risks and opportunities that they may face. OnDeck reminds us of Capital One in the 1990s. Shortly after its IPO, Capital One was growing at approximately the same rate as OnDeck today. In its first annual report, Capital One discussed how the company was revolutionizing the credit card market through better use of information analytics, or as management called it, Information Based Strategy (IBS). They wrote,
“Our consistently strong growth is the result of the proprietary information-based strategy we created in 1988… We saw that the technology and information revolution had transformed the credit card business into an information business. One that is extraordinarily data-rich…With this information, we can conduct scientific tests; build actuarially-based models of consumer behavior…Using advanced information technology and sophisticated quantitative analysis, we mine the vast amounts of data we have collected on millions of actual and prospective customers… and tailor products, pricing, credit lines and account management to meet the individual needs and wants of each customer.”
But as with almost any innovation, imitators quickly followed. After its first year as a public company, Capital One experienced significant competitive pressures similar to what OnDeck faces today. According to Capital One’s 1997 annual report,
“In all our businesses, competition is fierce…[c]ompetition in the credit card industry, as measured by the volume of mail solicitations, remains very high…Intense competition in the credit card market has resulted in a decrease in credit card response rates and reduced productivity of marketing dollars invested in certain lines of business.”
In addition to competitive pressures, Capital One entered a turbulent credit environment:
“Although the economy was booming in 1997, with unemployment at a 24-year low, inflation at an 11-year low and consumer confidence at its highest point in 28 years, the consumer credit sector continued to experience increased charge-offs for the third consecutive year before finally stabilizing in the third quarter of 1997.”
But management noted how Capital One’s Information Based Strategy made it nimble: “The vast quantities of data we process daily in servicing our accounts give us an ability to manage credit risk analytically; when conditions change, we can respond quickly.” And so, worsened conditions in the consumer credit market benefited Capital One’s market position; its competitive advantages and prudent underwriting allowed it to prosper while competition suffered. Management wrote,
“Even more gratifying than the results themselves was the fact that they were achieved in one of the most turbulent years in the history of the credit card industry. In the first half of 1997, as charge-offs and consumer bankruptcies continued the ascent that began a few years ago, several of our competitors either exited the credit card business or retrenched. Capital One’s record performance in this challenging climate demonstrates the power of information-based strategy, our innovation and our financial conservatism.”
In November 1994, Capital One IPO’d at $5.33 a share.12 Within less than 5 years, the stock traded at $55 a share—a tenfold increase, equaling a 60% annualized return. Capital One was less than a billion dollar company at the time of its IPO and is a $35 billion company today. Using hindsight, what was the appropriate book value multiple to pay for Capital One in its first year as a public company? Let us just say paying even 10 times book value and holding it for 10 years would have still provided an investor with a satisfying return. Today, OnDeck trades at 1 ½ times book value.
We think OnDeck’s valuation is cheap. Though to the quick observer, OnDeck optically seems like it is barely profitable, a deeper understanding of the company’s accounting erases that notion. Accounting rules require OnDeck to expense advertising and loan loss provisions up front, while its interest income is amortized over the life of its loans. This front-loading of expenses for a rapidly growing company obfuscates its true profitability. Further masking its profitability are OnDeck’s significant investments in R&D today, which we believe will pay off in the future. Our steady-state model shows that OnDeck is trading at a single digit earnings multiple, which we think is too cheap for a company that just grew revenues 62% year-over-year and still has potential for significant growth.
A decade from now, small business owners will continue to need capital, and they will prefer to access it in the most frictionless and cheapest manner possible, coupled with the best customer service. OnDeck fulfills these needs, while continuing to enlarge its competitive advantages. We are likely still in the early days of how technology can transform lending. There are many ways that the underwriting paradigm can evolve in what is now a Big Data world. Given their history of innovation, we are excited by what OnDeck’s management team may still create in the future which we have not thought to value today.
* * *
We remain excited about the investment returns we think we can deliver to our Partners over the next decade. Virtually all of my family’s and my entire net worth remains invested in the Partnership, as we aim to compound our wealth alongside yours. As always, if there is anything you wish to discuss, please feel free to call me. I look forward to writing to you again in the summer.
Sincerely, Eric E. Khrom
Khrom Capital Management, LLC
Khrom Capital footnotes
1 Federal Reserve, Joint Small Business Credit Survey, 2014.
2 Karen Gordon Mills and Brayden McCarthy, The State of Small Business Lending: Credit Access during the Recovery and How Technology May Change the Game, Harvard Business School, July 22, 2014.
3 OnDeck’s model is more accurate than a personal credit score in predicting bad credit risk, and allows it to fund double the number of loans as a personal credit score at the same level of risk [OnDeck, Request for Information on Online Marketplace Lending and Expanding Small Business Access to Capital, 2015].
4 Approximately one-half of OnDeck’s loans are completely underwritten using its proprietary automated underwriting process [OnDeck, SEC 10-K, 2015
5 Based on OnDeck’s 2014 customer cohort, customers who took at least three loans grew their revenue and bank balance, respectively, on average by 29% and 54% [OnDeck, SEC 10-K, 2015].
6 Bain & Company, Customer Loyalty in Retail Banking, 2013.
7 First Annapolis Consulting, Chase & OnDeck Small Business Lending Partnership, 2016.
8 Small Business Administration, Frequently Asked Questions About Small Businesses, 2012.
9 FDIC, Loans to Small Businesses and Farms, FDIC-Insured Institutions 1995-2015, Q4 2015.
10 Oliver Wyman, Financing Small Businesses, 2013.