From Bowie Bonds to Pipe: Financing Recurring Revenue
Plus: Football Index, More Greensill, Tinkoff
|Marc Rubinstein||Mar 12||23||5|
Welcome to another issue of Net Interest. There’s 15,000 of us here now, which is a multiple more than I anticipated when I started this thing, so thank you. If you’re reading this but haven’t yet subscribed, join the party by signing up here:
From Bowie Bonds to Pipe: Financing Recurring Revenue
David Bowie is remembered as one of the most influential musicians of modern times. Less widely appreciated is his side hustle in financial innovation.
In 1997 Bowie turned fifty. By then he had created a massive body of work – over 25 albums including Space Oddity, Ziggy Stardust and Let’s Dance. Together with banker David Pullman, Bowie devised a scheme to raise money against the royalties those albums would generate. Through the process of securitisation, he was able to swap some of his future royalties for $55 million of cash up front. His counterparty was the insurance company, Prudential Financial, which received a 7.9% annual coupon. After ten years, the deal would expire and cash flows linked to the royalties would fully revert to Bowie. Notwithstanding some wobbles along the way – like when peer-to-peer file sharing threatened physical album sales in 2004 – the deal ran its course and liquidated successfully in 2007.
Nevertheless, the idea that a predictable set of cash flows can be traded for cash proved alluring. A special financial structure called a whole business securitisation emerged to channel sticky cash flow streams to investors. Today, it is used by companies which generate steady streams of income without the tangible assets that would provide the backbone for more conventional securitisations. Planet Fitness completed a deal in 2018 to raise $1.275 billion in cash against membership fees across its gyms. And in 2020, drive-in restaurant operator, Sonic, raised $900 million against the franchise fees it collects from its restaurants.
Over the past several years, more and more companies have developed streams of revenue that exhibit the predictability of franchise fees and gym memberships. The trend is wrapped up in the penetration of software as a service but also reflects the success of platform and subscription based business models. The trailblazer was Adobe, which now makes 90% of its revenues from subscriptions, up from 10% in 2010. But it is apparent everywhere – even in financial services, where challenger banks are beginning to offer subscription plans in lieu of fees.
Unfortunately, smaller companies don’t have the opportunity to access the whole business securitisation market so they have to seek alternative mechanisms to leverage the predictability of their revenues. Traditionally they have used merchant cash advances.
Merchant cash advances were developed in the same year Bowie was securitising his royalties. While Bowie was winding down after his birthday concert at Madison Square Garden, a Connecticut businesswoman named Barbara Johnson was probing ways to finance her business expansion.
Johnson operated a group of Gymboree Playgroup and Music franchises. She wanted to launch a summer marketing campaign but didn’t have the funds. Knowing that revenue would resume when kids returned after the summer, she wondered whether she could borrow against those future revenues. Together with her husband, she co-founded a company called AdvanceMe (now CAN Capital) and patented an idea.
Johnson’s idea was that lenders get paid back directly out of credit card receipts, thereby reducing the risk they take on. Unlike whole business securitisations, which are structured with fixed payment terms, merchant cash advances are paid down via a fixed share of revenue and remunerated with a fee. In both cases though, lenders get a very early claim on the company’s revenues.
Johnson’s patent for “automated loan repayment” was ultimately invalidated for being “obvious” – according to the ruling – but not before she’d popularised the idea. (One of the reasons financial services is so competitive is that it’s hard to secure patent protection on product design.)
In the past few years, the market for merchant cash advances has grown enormously. Techniques deployed to assess merchant credit risk in the offline world work even better in the online world. Ecommerce businesses may not generate revenues that are as stable as offline franchise businesses but they generate sufficient data that lenders can underwrite to a better standard. Access to data sources across accounting, marketing, billing and banking, as well as a direct line into payments and view of the online storefront, allow lenders to underwrite credit that wouldn’t have been available in the past. That’s the upside of the API economy.
Shopify launched a merchant cash advance product in 2016 branded Shopify Capital. Originations were up 81% in 2020, to $794 million. As at year end, Shopify had $219 million of cash advances sitting on its balance sheet and $44 million of traditional loans. The business isn’t risk free – the charge-off rate on its cash advances was around 5% – but the margins are attractive.
Some startup companies are taking the concept further and underwriting specific uses for capital. One such company is Clearbanc, which allows customers to draw down funds to invest in marketing. Clearbanc underwrites against the return on investment it models clients will achieve on their customer acquisition costs. It will advance $100,000, say, for a client to invest on customer acquisition and get paid back via a share of customer revenues. Through to the end of 2020, it has originated around $1.6 billion of capital at a loss rate of around 1.5%.
We’ve spoken about Ant Group here plenty as the benchmark for what every fintech evolves into given enough time. What Ant has shown is that being inside customers’ payments and accounting systems – in a way that traditional banks aren’t – is a highly effective way to mitigate credit risk. Via different paths, it’s what’s going on here, too.
As well as whole business securitisations and merchant cash advances, there’s a third model emerging to help companies leverage the predictability of their revenues. It involves selling cash flows linked to individual revenue contracts.
Alex Danco calls the revenue contract “the atomic unit” of the subscription economy. Just as the store is the atomic unit of the offline retail economy – and can be turned to cash through, for example, sale-and-leaseback structures – the revenue contract is the atomic unit of online recurring revenue businesses, with similar potential to be turned to cash.
The problem businesses with recurring revenue face is that their customer contracts aren’t considered assets in the traditional sense. Over the years a whole menu of financing options has emerged to service asset-based companies. Factoring is one of the oldest, predating whole business securitisations and merchant cash advances. It allows sellers to get hold of cash on a faster timeline than customers are able to provide it. But it relies on the role of the invoice – and the corresponding legal obligation to pay – as an asset.
Companies have developed in-house solutions. In particular, they offer customers discounts for annual payments. Zoom Pro costs $14.99 per month per license or $149.90 for a full year, equivalent to a 17% discount for cash up front. Dropbox is the same, as are most subscription-based newsletters that feature on Substack. The market convention is that as a customer, you get twelve months for the price of ten.
But 17% is a steep price for the company to pay! Spotting this, several fintech firms have sprung up to capture what is essentially a pricing arbitrage opportunity. By inserting themselves between the customer and the company, they can offer the company a lower discount for cash up front while the customer continues to pay on their preferred monthly schedule.
One of these companies is Pipe. It was founded in late 2019 by entrepreneurs Harry Hurst, Josh Mangel and Zain Allarakhia. As an indication of how hot an area it is, Pipe announced a $50 million fundraise this week (adding $16 million to the equity it raised last year).
The model is a kind of reverse Buy Now Pay Later (BNPL). Buy Now Pay Later firms like Klarna and Affirm offer customers the opportunity to split an upfront payment into installments; the model we’re discussing here offers the company the opportunity to roll recurring payments into an upfront payment. In both cases the company pays, while the funding partner underwrites customer credit risk. In Buy Now Pay Later, the discount averages around 6% (for a four month duration); in the reverse model, it’s finding its level, but it’s lower than 17% the companies would fund it themselves (for twelve months).
When a recurring revenue business acquires a customer, they are, like in everything, acquiring a bundle of features, many of which can be separately monetised. There’s the data the customer provides, the attention of the customer, the cash flows the customer generates over different time periods. These features can be monetised via referrals, sponsorship, subscription or a lower cost of reactivation if the customer churns. What is elegant about the financing model we’re talking about here is that it strips out a specific series of cash flows (12 months, say) and sells them off, while retaining all the other features.
There’s been a lot of debate around how this form of financing compares to debt and to equity. I see it quite simply. If debt is a claim on the asset, and equity is an option on the asset, then this is a strip of the asset.
Though not a loan, from an accounting perspective, the company can recognise the discount to cash flow as an interest expense below the line, so the product doesn’t affect EBITDA, operating expenses, gross margin, or revenue. If the customer churns, the company may be required to either compensate the buyer of the revenue contract on a prorated basis or replace that customer’s cash flows with those from another customer.
Since launching a year ago, Pipe has had over 3,000 companies sign up, with over a $1 billion of tradable ARR on its platform. Its model is slightly different from others in the market. Rather than underwrite revenue contracts itself, it provides a marketplace for them to be traded. One side of the business originates revenue contracts and rates them; another side presents them to investors to bid on.
On the sell-side, customers range from newsletter writers like Anthony Pompliano (who raised cash from his monthly subscriptions to re-invest in Bitcoin) to publicly listed companies. While Pipe has formerly been focused exclusively on subscription, it is broadening its scope to target all recurring revenue streams.
Liquidity on the buy-side of the marketplace was kickstarted last year via a $50 million injection of funds from a consortium of banks. It is now provided by a range of banks, hedge funds, pension funds and others. Going forward, that range may be broadened out to include platform companies. The company’s recent funding round included investment from one such platform – Shopify. Alex Danco – who works for Shopify – argues that platform companies would be eager investors in revenue contracts as a means to support the growth of their customers. Provided that doesn’t go too far, it could be an interesting development.
Pricing of revenue contracts on the Pipe platform has narrowed from 85-90 cents on the dollar at launch to 90-95 cents currently. Pipe takes a trade fee on each trade it facilitates and takes care of the servicing. Right now, most contracts are for twelve months, but that can vary depending on the nature of the sell-side company’s assets.
Pipe’s co-founder Harry Hurst sees revenue contracts as a new asset class. It’s an ambitious vision because cultivating new asset classes is hard. We’ve discussed before why some gain traction and some don’t. The key is to ensure supply and demand are aligned. When the founders of KKR worked to get leveraged loans off the ground in the 1970s they “proved adept at cultivating trust with debt and equity investors, on the one hand, and target companies and their managers, on the other”. And eurobonds were started in the 1960s as a mechanism to bring together (with a tax kicker) sleepy assets sitting in Switzerland with companies looking for ways to raise finance more cheaply than in tightly regulated New York.
In Pipe’s case, supply on the sell-side stems from the growth of recurring revenue streams and their operators’ hunger for efficient sources of capital. Low rates provide demand on the buy-side, particularly if these assets can be leveraged, and if Danco is right about platform companies taking an interest, that opens up an entirely new pool of demand.
As a marketplace, Pipe has taken a leaf from the ICE and MarketAxess playbooks, bringing potential participants of its platform in as strategic investors. As well as Shopify, it now has Siemens’ Next47, Raptor Group, Slack, HubSpot, Social Capital’s Chamath Palihapitiya, Marc Benioff and MSD Capital as shareholders.
However, exchanges work best for standardised contracts. The revenue contract is a short term contract averaging twelve months and paying down every month, so trading is assured if buy-side investors want to stay fully invested. But for secondary trading to take off and revenue contracts truly to be regarded as a new asset class, the next phase would be to introduce more standardisation among contracts.
Sound and Vision
David Bowie famously grasped the potential of the internet before many (“It's just a tool though isn’t it?” asked the interviewer in 1999. “No it's not!”)
The internet allowed different business models to emerge with more recurring revenues and for finance companies to plug straight into them. Fintech works best when it harnesses the common feature of both finance and technology - to improve efficiency – and it feels like this crop of companies may be doing that. It’s an area of fintech I will be watching with interest.
More Net Interest
One of the earliest Net Interest pieces I wrote was about the convergence of stock market investing and betting. I didn’t touch on it but back then a company called Football Index was showing rapid growth. The company operated a stock exchange-like platform on which punters could buy shares in players that rise and fall in value based on their performance on the pitch, and demand on the index. In addition, the platform would pay out cash “dividends” based on players’ performance – modelled on Fantasy Football points but with real cash.
Football Index invested heavily in growth. It sponsored two football teams, advertised widely, communicated its expansion plans and attracted customers based on tales of people making lots of money on the platform. I even backed my son on the platform to the tune of £100!
This week, Football Index collapsed. Faced with liquidity issues, it slashed dividend rates across the platform, leading to a crash in prices as customers looked to get their money out. As they hit the exits, bid-offer spreads widened dramatically.
The company behind Football Index has now filed for administration. It looks like a classic Ponzi scheme.
The revelations around the Greensill collapse keep coming. Trouble is, we still don’t know who’s left holding the bag.
Softbank indicated as soon as the news broke that it would have to write off its $1.5 billion equity investment in the company, but then it emerged that they invested another $400 million just a few months ago.
General Atlantic is another shareholder that will be taking a loss, in their case $250 million.
Retail depositors in Greensill’s German banking subsidiary are fully insured up to a limit, but municipalities aren’t – German towns are braced for €500 million in lost deposits, according to the FT.
Credit Suisse has a $140 million loan outstanding to Greensill which has “no conceivable way” of being repaid.
Credit Suisse clients have $10 billion of funds frozen in Greensill-sponsored funds; the bank confirmed that $3 billion of cash has been paid out, but that still leaves $7 billion.
John Hempton wonders what exposure insurance companies Insurance Australia Group and Tokio Marine have. IAG put out a press release saying that they have no net exposure but that’s an issue that may go to court. Particularly as Credit Suisse is “working under the hypothesis” that the insurance policies will pay out.
Ultimately there are two outstanding questions. How much Gupta exposure there is across the Greensill estate. Gupta hasn’t gone bust yet, but he’s likely to be the source of further losses. And, second, how tight the insurance contracts are. The first question will be answered more quickly than the second.
We explored Russian digital bank Tinkoff here in Net Interest a few weeks ago. This week, they produced their full year results. Total customer numbers grew 30% to 13.3 million over the year. But the number of monthly active users grew more quickly, up 55% to 9.3 million as engagement increased. This increase in engagement is a feature we’ve seen across digital banks over the course of the pandemic year. In Tinkoff’s case that higher engagement translates into revenue, with 9.1 million of the users having generated revenue in the last month. Products per customer increased from 1.3 to 1.4 even on top of high customer growth.
Tinkoff’s original target - to hit 20 million customers by the end of 2023 - now looks a little conservative. The company will update its plans at an investor day scheduled for 7 April. But if you can’t wait, you can always talk to Oleg, the company’s chatbot (named after its founder), who has taken up a residency in Clubhouse.