Hey all, Jason here.
After three weeks in Mexico, I’m back home in the Netherlands. This week’s issue is informed by those travels and conversations I had with banking and fintech practitioners in the country. I’m by no means an expert on the financial services system and barriers to inclusion in Mexico, so if I’ve missed something or you have additional context to add, feel free to let me know.
New here? Subscribe to get Fintech Business Weekly each Sunday:
Fintech Meetup Returns in 2022
Sponsored content: I had the pleasure of attending the inaugural Fintech Meetup event earlier this year, and it's one of the most valuable events I've ever attended.
The event received such a positive reception, its organizers are already planning for 2022. The next Fintech Meetup will take place online March 8th - 10th, and will keep the same online meetings event format as the inaugural event, but with a number of exciting platform enhancements to make the experience even better. 2022 participants are capped at 4,000 and the event will surely sell out.
Learn more about the event, a limited number of sponsorship opportunities, and get your tickets here:
Why Credit Card Rates Can Hit 151% in Mexico
As frequent readers of this newsletter will know, I was recently traveling in Mexico. While in Oaxaca, I found myself in a Coppel department store, and I immediately noticed that every single item in the store, even a MXN $150 (USD $7.5) SIM card, could be financed in bi-weekly installments.
I was seeing one of the hottest trends in fintech, in the flesh… in a remote corner of Mexico? (Yes, I guess this is what I do on vacation, but that’s a different discussion.)
Of course, even if the UX and name have evolved, buy now pay later isn’t new. Retail financing has long been a tool deployed by merchants, formally or informally, to boost sales by enabling consumers to defer or spread payments over time. This becomes even more important where consumers have few mainstream credit options or may have difficulty saving for a purchase, which is often the case for consumers without access to a bank account.
So it shouldn’t be too surprising to see this type of point-of-sale financing in developing economies; in fact, quite the opposite. At retailers like Coppel, the requirements to qualify can be minimal; even a bank account is not required, as cash repayments can be made in store or at an Oxxo (popular convenience store). About 70% of Coppel’s sales are financed through its private-label credit offering.
And the rate? This varies by product being purchased, but, based on my back-of-envelope calculation, was around 30 to 40% APR. High by American standards, but favorable compared to other consumer credit products in Mexico.
Coppel has leveraged this financing business — and the rich data set it has built on its customers — into its own bank, BanCoppel. It began operating about 14 years ago and has roughly 1,000 branches.
What About Credit Cards?
General purpose credit cards are merchant-agnostic and thus arguably offer greater utility (and greater profit potential for issuers). But only about 10% of Mexicans hold a general purpose credit card vs. 79% of Americans.
While there are a variety of intertwined barriers to increased credit adoption, no doubt the high cost of credit is one of them.
According to a December 2020 market analysis from Banxico (Mexico’s central bank), rates on credit cards range from 24.5% CAT (CAT is the total annual cost of credit, computed similarly* to APR), for a World Elite Platinum Card from Santander, to 151.1% for the Volaris Clásica card offered by Invex. For cards with the lowest credit limit, those below MXN $4,500 (about $225), rates ranged from 43.7% to 129.6%.
The average across all credit card accounts in the US? A comparatively cheap 14.61%. When looking at accounts of customers that carried a balance, the rate rises to 16.30%.
[*note CAT includes credit card annual fees, which are not included in APR calculations]
What Factors Help Account For Low Credit Penetration & High Costs?
High Rates of Informal Employment
Recent data show that 58% of workers in Mexico are “informal,” including wage earners that do not have access to social security and self-employed workers not following a formal accounting system.
Informal employment complicates access to financial services and especially credit as this income is typically paid in cash, may be unverifiable, and generally is less stable than formal employment. BBVA Research has historically found a positive correlation between the rate of formal employment, the real wages of those in formal employment, and consumer credit growth.
Low Bank Account Penetration
Despite relatively high acceptance of cards in more developed parts of Mexico, cash is still overwhelmingly the preferred payment method, accounting for as many as 86% of transactions.
This is partially explained by the high rate of informal employment. High rates of cash usage also have a symbiotic relationship with low bank account ownership — if you’re paid in cash and are spending in cash, there is less reason to bother with a bank account, particularly with the high fees many Mexican banks charge.
Fees associated with bank accounts can be significantly higher in Mexico as a share of income or account balance. This is evident in the average proportion of banks’ revenue derived from fee income vs. non-fee income — around 30% of Mexican banks’ income comes from fees vs. about 4% for US banks.
Inflation & Central Bank Mandate
Inflation risk is a key ingredient in pricing the nominal interest rate of credit products. This is another area where history and the present in Mexico diverge significantly from the US.
Banxico managed to bring high inflation stemming from the 1994 “peso crisis” to below double digits by 2000, but annual inflation in Mexico remains about double that of the United States. The scale here makes it difficult to see, but in 2019, Mexico’s inflation rate was 3.64% vs. 1.81% in the US:
The Mexican central bank also has a different mandate than the Federal Reserve. While the Fed has a dual mandate to promote price stability and full employment, Banxico has only one mandate: to maintain low and stable inflation.
An additional complicating factor in controlling inflation is the Mexican economy’s deep links to the US — and to the US interest rate and US dollar. In order to maintain the value of the peso and mitigate inflation risk, Mexico’s interest rate policy changes often closely mirror those of the United States (note dual axis):
Despite relatively tight access to credit, default rates on consumer credit products in Mexico can be orders of magnitude higher than those in the US. Higher default rates mean higher interest rates, even as those steeper finance charges may make borrowers more likely to default.
The delinquency rate (above, left) does not include charge-offs (does not include defaulted loans cleared by applying loss reserves); the adjusted delinquency rate (above, right) does include amounts that were charged off.
For comparison, delinquency rates on US credit card loans hit a high of 6.77% in Q2 2009 amidst the fallout from the global financial crisis. In Q1 2020, before the onset of the pandemic, delinquencies stood at 2.65% and actually dropped since then to 1.58%, owing to government stimulus measures.
Assuming Banxico’s adjusted delinquency rate is comparable, that would make defaults on credit cards in Mexico about 5.5x higher than those in the US. What factors can help explain this radically higher rate?
Less Data for Underwriting Ability & Willingness to Repay
While Mexico has credit bureaus that operate similarly to their US counterparts, one Mexican fintech founder I spoke to estimated they had data on perhaps 10% of the population vs. over 80% of the US population that has a credit file sufficient to generate a traditional credit score. A consumer’s payment history is key to understanding their willingness to repay.
Determining someone’s ability to repay is an exercise in comparing their income vs. existing debt obligations and monthly expenses to understand if they can comfortably service the debt they’re seeking to take on. With credit bureau data covering few Mexicans, it can be difficult to determine whether or not they have other debts. For the 58% informally employed, verifying income may be difficult or impossible.
Ability and Cost to Collect When There Are Limited Downsides to Not Repaying
While collectors in Mexico leverage similar channels to reach borrowers as in the US, efficacy depends on the enforcement mechanisms available. For Mexican consumers who already had scant access to credit, the risk of losing access to credit by defaulting may feel less acute.
And as credit scores are used in far fewer ways in Mexico than in the US (apartment leases, car rentals, insurance, employment screening, etc.), not repaying implicitly carries a lesser penalty in Mexico than in the US.
While collection can be pursued through the courts in Mexico, it tends to be a lengthy, complicated, and expensive process. The ability to reach delinquent borrowers, especially in rural and poorer areas of the country, can be limited.
The result is relatively low credit penetration as a percent of GDP, with Mexico reaching a credit-to-GDP ratio of just 64% in 2020.
With low bank account penetration (37%), low credit usage, few bank branches (13.7 per 100,000 population vs. 30.46 per 100,000 in the US), and high smartphone usage (73%), Mexico is well-positioned to benefit from fintech.
Unlike the crowded US market, where establishment banks and challengers are competing for mostly well-served consumers, the opportunity in Mexico (and throughout LatAm), is to bring new users into the market by serving those ignored by, priced out, or physically unable to access traditional banks.
How Bitcoin Stacks Up For Remittances to El Salvador
In case you missed it, last week, El Salvador’s Bitcoin Law took effect, with Bitcoin joining the US dollar as legal tender in the country. El Salvador’s economy has been “dollarized” since 2001, in an attempt to stabilize and grow the economy, with mixed results.
Much of the official narrative and media coverage around the Bitcoin Law has focused on remittances — which makes sense, given that about $6 billion was sent to the country in 2020, accounting for about 23% of its GDP. This CNBC story loudly proclaimed the move to Bitcoin could cost Western Union and other remittance companies $400 million a year in lost revenue.
So I decided to compare how a legacy company, newer fintechs, and Bitcoin would work in practice when sending USD $200 from the United States to El Salvador.
The following is based on publicly available information as published on company websites; all product names, logos, and brands are property of their respective owners in the United States and/or other countries.
Legacy Transfer Service: Western Union
Based on analysis of Western Union’s online money transfer service, the costs and times to fund would vary as follows based on how the sender is transferring money to Western Union and how the receiver is accessing the funds:
Note this doesn’t include the sender using cash in person, which would incur additional costs.
If WU’s site is accurate, it’s possible to send a $200 remittance via debit or credit card to a mobile wallet (Tigo), nearly instantly, for no cost. This admittedly doesn’t make a ton of sense, as WU would incur interchange fees on the debit/credit transaction on the sender side, but perhaps consumers rarely use these methods vs. higher-fee and more profitable in-person transactions?
Fintech Transfer Servies: Wise, Xoom
Wise (formerly TransferWise), is arguably the most well-know fintech in the international transfer space. Like most fintechs, it is online-only, which makes it a less likely option for remittances, as one or both parties may be operating in cash (eg, a physical agent presence is necessary).
I couldn’t find any public documentation on Wise’s site explaining how or if it was possible to send USD from the US to El Salvador (I could only find this SEO landing page targeting the UK.)
If you step through Wise’s flow, indicating that you want to send $200 USD -> USD, but not to a local bank account, you’re greeted with the following warning:
“To send USD outside the US, we need to use SWIFT. This means your transfer will cost $2.90 more and it may take 2-5 days longer. Other banks involved may also charge fees.”
The transaction would be slow, taking 2-5 days (or longer), and you’re limited to sending it to recipients with a bank account — not a solution for most El Salvadorians. Wise’s fee would remain pretty low (~$3.94), though it warns “other banks involved” could charge fees.
Xoom, a peer to Wise that was acquired by PayPal in 2015, offers more robust functionality for remittances to El Salvador. Based on its site, the cost and times to send $200 from the US to El Salvador would be as follows:
The pricing here also doesn’t make a ton of sense — specifically, that debit/credit are priced the same (different interchange rates) and that debit/credit -> mobile wallet is less expensive than ACH or PayPal.
Using Bitcoin As Transfer Service
Many advocates of the advantages of crypto for international payments/remittances seem to make the assumption that the sender may already hold funds in that cryptocurrency; in this case, Bitcoin.
That doesn’t strike me as a fair assumption. A typical user remitting money to a friend or relative in El Salvador is probably starting in USD, not Bitcoin — which can impact the speed and cost of the transaction.
A hypothetical transaction leveraging Coinbase, a popular US crypto platform, to move USD from a US user’s bank account to a Salvadorian user’s Chivo wallet via Bitcoin could look something like this:
At the time of writing, the US user would incur a fee from Coinbase of $2.99 - $7.67 to convert $200 into Bitcoin, funded via ACH, debit or PayPal. Coinbase steers users to fund purchases via ACH (because it’s cheaper), but this transaction between the user’s US bank account and Coinbase can take up to 4 days to clear; Coinbase will front the user the $200 + the fee, which entails some credit risk to Coinbase.
Coinbase’s FAQ also states it may take a spread on the exchange rate when a user buys or sells crypto currencies (emphasis added):
“Coinbase includes a spread in the price when you buy or sell cryptocurrencies or in the exchange rate when you convert cryptocurrencies. This allows us to temporarily lock in a price for trade execution while you review the transaction details prior to submitting your transaction.”
The blockchain transaction to convert the USD to Bitcoin would currently take about 10 minutes on the open network (if Coinbase doesn’t fulfill internally), and then presumably another 10 minutes + Bitcoin network fee to send to an El Salvadorian’s Chivo wallet address. Chivo wallet users have the option to automatically convert received Bitcoin into USD at no additional cost.
The transaction times and costs for sending Bitcoin can vary wildly, depending on network traffic. Transaction costs spiked to $62 this April (regardless of amount of Bitcoin being moved), and average confirmation time stretched to over two days in May 2021, though both have been more stable since early July.
Such volatility in time and cost can create uncertainty, undermining the utility as a mechanism for moving money, especially for low-dollar value remittance transactions.
And the Winner Is….?
It’s not particularly clear there are any speed OR cost advantages to using Bitcoin vs. existing remittance channels. In the US-El Salvador case, most parties will still be moving money between two traditional banking systems. It’s unclear what value, if any, substituting a blockchain for current rails has; this could change if both sides of the transaction are “crypto-native”, but, at present, this seems unlikely to be the case.
Still, the Chivo wallet launched by El Salvador is an interesting public policy experiment. It can hold both Bitcoin and USD, and allows users to convert between them and to transfer to other Chivo wallets, including merchants’, at no cost.
Even if most Salvadorians are skeptical of using Bitcoin as a currency, Chivo could succeed in improving bank account/mobile wallet penetration in the country, which would increase the efficiency of traditional USD remittances and bring down their cost.
Other Good Reads This Week
Fintech in LatAm: Banking the Unbanked (Aika’s Newsletter)
Lending Bitcoins is Tricky (Bloomberg Opinion)
Inside the Cult of Crypto (FT)