Thursday 6 February 2020

Afterpay Touch: a more expensive solution to an existing, solved problem

Afterpay Touch (APT AU) is a stock that has acquired a cult following in recent years. The shares have risen spectacularly, and catapulted the company's market capitalisation to an astonishing A$10bn - almost 40 times trailing revenues (there are no earnings - the company lost about A$40m last year on revenue of $260m).*

To the uninitiated, APT is an ASX-listed fintech 'buy now pay later' (BNPL) payment system/platform, that styles itself as a replacement to traditional credit cards. The value proposition from the customer side is that consumers can use APT to purchase goods/services and pay back the amount in subsequent fortnightly installments, and unlike credit cards, do so interest free and without any pesky credit card fees. They have been adding a lot of merchants to their network, as well as consumers with APT accounts, and this has created a putative two-sided network flywheel effect that has caused investors to salivate about the possibilities - particularly as they begin to pursue international expansion.

The company's presentation materials are replete with stories of multi-trillion dollar addressable markets (indeed, in the company's FY19 results presentation, you have to get about 30 slides in before you get to any meaningful financial information, whereupon you learn that they lost money in FY19). The company seems to envisage itself disrupting the traditional credit card industry, and becoming a conduit through which trillions of dollars of consumer payments are made the world over.

Going by the company's share price, a lot of investors evidently believe them, and have bought into the story hook line and sinker. There is only one problem with the bullish narrative, though: the company's core product is actually a more expensive solution to an existing, solved problem, that actually makes the world a worse rather than better place. Long time readers of this blog will know that I'm a skeptic of many putatively disruptive 'Valley' businesses, but I will acknowledge that at least most of them are attempting to make the world a better place, and offer better products/services at lower prices, even if most of them are losing money trying to do it. APT is different, and I find it very difficult to think of any 'disruptive' business models that have succeeded where their product is a more costly and inferior solution to existing alternatives, for obvious reasons.

Furthermore, the very factors which caused the company to enjoy some initial success in building out its network do not scale, and are also exposed (as I will discuss) to considerable regulatory risk. Indeed, as the company gets larger, rather than becoming more valuable and entrenched as network-effect businesses typically do (and APT's share price anticipates), it will actually increasingly undermine its merchant value proposition, and make the fundamental structural weaknesses of its business model impossible to ignore - potentially bringing the whole house of cards tumbling down.

I will acknowledge one thing, however: from the consumer's point of view, APT is clearly a great product and in many respects better than a traditional credit card, for a very simple reason: there are no fees, and it allows you to borrow to pay for purchases without incurring any interest (although the rewards programs associated with many credit cards need to be forgone). APT and the bulls belabour this point ad nauseam. The problem though is that this is not the end of the story, as the issue is not the consumer-side experience, but the fact that APT charges merchants 4%+ to accept APT payment. Indeed, the AFR has reported that APT charges merchants a fixed 30c transaction fee, plus a 3-7% sales commission (while APT's accounts indicate it averages to a bit over 4%), which is significantly more than merchants typically pay to accept credit cards (which ranges from about 0.6% for large merchants to 1.5% for small merchants, in Australia according to the AFR), or cash (0%).

APT might at first appear to be saving consumers money (and be perceived that way by consumers), but somebody has to pay for this 4%, and that somebody will ultimately be the end consumer, because merchants will have no choice but to raise prices to recoup this 'APT tax'. The bulls attempt to deny this fact, but denials make no economic sense (at scale). Most retailers earn in the vicinity of 5-10% operating margins, and the industry is not going to settle for seeing more than half of its profitability stripped away and passed on to APT. And it makes no sense that a product as simple as APT's would, in competitive free markets, allow it to capture more than half of the entire retail industry profit pool.

In the long term, equilibrium industry margins are determined by the complex interaction of a number of forces, including demand and supply, competition, industry cost structures, and the cost of capital. If 5-10% is the economic level of margins for the industry, that is where it will stay, and just like if labour and rental costs were to rise and need to get passed on to end prices to preserve industry margins, so it will be if merchants have to pay a 4% APT tax on turnover. The mechanism might be implicit rather than explicit ("our margins are under pressure - let's bump prices up slightly and see if they stick"; rather than "let's put prices up specifically to compensate for the APT tax"), but the outcome will be the same either way: consumers will ultimately end up paying.

It is staggering that the APT bulls actually believe it is feasible the world will move to a situation where merchants the world over will be paying a 4% transaction tax to the likes of APT; retailers won't put prices up to compensate; and that this is somehow innovative or disruptive, and will represent an improvement to the world as compared to the status quot ex ante. The reality is that this is a much worse world - a world where everyone pays much more for goods and services than they otherwise would, so companies like APT can pocket a massive, parasitical middleman fee.

Why then, you might ask, has the company achieved a degree of success/scale to date? For two reasons: (1) at a small scale, its platform does offer early-adopting merchants some meaningful benefits that offset the high transaction costs; and (2) APT's contracts with merchants have restricted their ability to impose a transactional surcharge when accepting APT payment. The former factor above contributed to merchant's willingness to acquiesce to the latter, but it is very likely that both of these two factors are fundamentally unsustainable.

It doesn't take a particularly deep level of thinking to see that the putative benefits for merchants are a fallacy of composition that don't scale. APT and the bulls are always quick to emphasise that APT allows merchants to boost their sales, but there are only two mechanisms by which this can occur. Firstly, an APT merchant can cannibalise the sales of competitors who do not accept Afterpay. From the consumer point of view, APT is a superior payment option than using a credit card/cash, so if Merchant A accepts APT and competing Merchant B does not, consumers might decide to shop more frequently with Merchant A, allowing Merchant A to pick up market share at Merchant B's expense. However, the benefits of this don't scale. It's only a matter of time before Merchant B decides to accept APT as well, and then the cannibalistic effect disappears, but the 4% transaction fees do not.

The second way APT can boost a merchant's sales is by providing consumers with yet another means by which to outspend their income. If a customer lacks the cash or credit card balance to purchase item X, without APT they will have to go without. With APT, they can now 'afford' to buy it. While in the short term this will indeed boost merchants' sales, the ultimate effect is merely to pull forward consumers' purchases, and so in the long term drives no sustained benefit.

Consequently, what we are left with is simply a more expensive solution to an existing solved problem. What is that solved problem? Fast, convenient, secure, and inexpensive electronic payments, and ready access to unsecured credit lines for customers (credit cards). If you think about it at a high level, the first payment system in modern times was cash. Cash has (and still has) its advantages - for a start, it's free (no transaction fees), and relatively fast and simple. However, it has some disadvantages - particularly in the modern era: you can't use it to buy things online; and you also can't use it to buy big-ticket items that require consumer credit, which you are able to afford by paying back over time, but don't have the cash to pay upfront for today.

Credit cards were the solution to this, as well as other use cases (e.g. for people travelling). In the past, credit cards had several attendant disadvantages, however. They were not only more expensive (requiring the acquiring of merchants; payment terminals; etc), but also importantly, they were slower to process than cash. However, with continued innovation, most of these disadvantages are now disappearing. The emergence of 'tap-pay' credit and debit cards which can process transactions virtually instantaneously has been particularly material, and is both convenient for customers and important for merchants, as it reduces queuing times at checkouts. Lower cost acquiring solutions (e.g. Stripe) are also emerging, and not ending up with a pocket full of coins is also a significant convenience benefit for customers. Credit and debit cards still cost more than cash, but the many advantages and convenience have increasingly outweighed them.

Not surprisingly, this has resulted in credit and debit cards increasingly becoming a superior solution to cash, and is why Visa and Mastercard have thrived in recent years, as the market share of credit/debit card payments have risen, while cash's market share has fallen. Furthermore, V/MA's global 'payment rails' represents infrastructure that other fintech companies can build upon to offer new innovative financial products, and at relatively low cost (V/MA transaction fees are only in the ballpark of 10bp; most of the cost of traditional credit card fees are split between banks and merchant acquirers, as well as funding the cost of rewards programs).

In other words, fast, secure, and affordable digital payments, as well as on-the-spot consumer credit, is already a solved problem. And now along comes APT, and wants to provide a similar solution at triple the cost. It doesn't make any economic sense. The reason there is so much confusion is that (1) from the consumer perspective, they falsely believe they are saving money (as the APT costs are hidden and incorporated into retailer prices); and (2) many merchants testify that they have benefited from APT, but as discussed, this was due to the early-adopter cannibalisation, and consumption pull-forward effects. What this means is there is an incredible lack of clear thinking and prevalent misinformation with respect to APT's business model and its likely ultimate effects.

There are signs, however, that things might be set to change. The AFR has reported that the RBA is looking into measures that would prohibit BNPL providers such as APT from denying merchants the right to impose surcharges. The policy discussion has arisen in response to advocacy from consumer protection groups, as well as merchant associations, who unlike the stock market, understand the situation for what it is, and argue correctly that it results in a lack of transparency and market failure (consumers perceive the product to be free, when in actual fact it is extremely costly, and this is resulting in consumers making suboptimal decisions, and stifling the emergence of lower-cost alternatives). While the bulls have recently been celebrating and back-slapping as APT has surged above $40 - a valuation that requires the company earn and distribute $2bn a year sustainably by 2026 to generate a 10% return - about 10x the company's current revenue - the emergence of an existential risk to the viability of the company's business model has been ignored.

The thing that irked me most - and prompted me to write this article - was the fact that APT is lobbying hard to oppose this RBA policy measure, and the shameless, self-serving misinformation the company has served up to oppose the policy absolutely needs to be countered (contained in the article previously referenced). It is important to emphasise here that no one is proposing that merchants must impose a surcharge - merely that they have the freedom to do so. Why is APT so rabidly opposed to their merchant customers having such choice? If the product is so great and it delivers so many benefits to merchants, as APT claims, then surely the merchants would choose of their own volition to not impose one for commercial reasons, right? In that respect, APT's protestations speak volumes about the extent to which APT's management recognises that the success of its business model hinges critically on its ability to continue to impose hidden transaction cost on customers.

In the article they make many spurious arguments, including that the merchant costs should be compared more to what Facebook or Google charges for click-through sales leads, rather than alternative payment platforms (a laughable claim); and that the RBA's mooted restrictions would "stifle innovation, compromise consumer choice, and reduce competition". This latter argument is so bad, and so shamefully self-interested, that I think it has crossed the line into outright unethical conduct by APT management, because the the truth is the exact opposite to what APT claims.

It is absolutely vital to the emergence of increased competition (which is a key driver of innovation), and hence enhanced customer choice (and value), that there is price transparency, and that there are avenues for competing payment platforms to proliferate and offer more attractively-price alternatives. Let's be very clear again: it is customers who are ultimately paying the 'APT tax', and customers paying in cash (or with a credit card) are also being forced to pay it. From a consumer protection perspective, it is vitally important that this tax does not become systemic and entrenched, and leave customers with no choice but to pay it - even if they are not an APT customer. By making the cost transparent and explicit, you allow consumers to make a more informed and efficient choice about which payment methods are the most optimal/cost effective. Consumers can ask themselves, ok I like APT, but do I like it enough to pay 4% more? That is the right question to have customers ask.

Indeed, I might even be prepared to go as far as arguing that regulators should insist on merchants imposing a surcharge (although importantly, restricting the magnitude of that surcharge to the actual cost borne from external payment providers). That would allow room for innovative and lower cost payment systems to emerge, because you would give consumers a reason/price incentive to change. Currently, that price signal is lacking, which is one reason why we have emerged with a global duopoly with V/MA. The way things currently work, where fees are hidden/built in to prices that all customers pay - even those paying cash - the system tends towards the creation of monopolies and excessively high fees. I for one really hope regulators address this situation.

But if regulations change, and merchants have the freedom (but not the obligation) to impose a surcharge, will they choose to? The answer, I believe, is quite likely yes; it will likely start slow, but a trickle will eventually turn to a torrent. Why? Because those that impose a surcharge for APT will have an important cost advantage over those that do not, and will therefore be able to offer all customers lower prices. And if surcharges start to appear here and there, it could quickly become an industry-wide trend. This will particularly be the case as the touted 'merchant benefits' start to fade, for the reasons previously discussed.

If surcharges indeed do become a trend, the consequences for APT could be quite catastrophic. Their product will now be revealed to be what it has always been - a more expensive solution to an existing, solved problem. Furthermore, 'adverse selection' could also become a serious issue for the company, as the customers that can afford to pay you back will probably opt to pay cash or via credit card, as they will enjoy lower prices by paying that way (avoiding the surcharge), while only those heavily strapped for cash would agree to pay the surcharge. This could result in APT ending up with a portfolio of low-quality payday loans that suffer very significant defaults.

Time will tell. I am a generalist covering all industries and countries, and have only spent a couple of hours researching APT, so I might be mistaken or have missed something. I am not short the stock, because for reasons discussed in past blog posts, I believe shorting is dangerous in a world of ample liquidity and irrational exuberance, as many Tesla shorts have recently had to once again relearn the hard way. Many sensible people were wiped out shorting worthless dot.com stocks too early in 1999.

I would not be at all surprised to see APT eventually fall >90%, however, and it's also not impossible it goes to zero (although much less likely than merely a >90% loss). One suspects that if APT were come to be seen as a payday lender with weak underwriting discipline, rather than a hyped-up conquer the world fintech payments platform/network, which a shift towards APT transactional surcharges could quite easily lead to, it wouldn't trade at 40x sales.


LT3000


*It is also notable is that about one third of the latter were late fees charged to past due borrowers, which APT book as revenue prior to collection - even though the collectability of late fees from delinquent borrowers ought to be considered questionable. This is very aggressive accounting.