Everything You Wanted to Know About Delinquency (But Were Afraid to Ask)

*I downloaded the entire database of debtors in Argentina and spent 48 hours looking at it. This is what I found, for those interested in knowing more. It reads better at https://todosobrelamora.vercel.app/

It had been a while since a public debate pulled me out of the lethargy of knowing that most things said in the media or on social networks are false, partial, biased, or stem from ignorance. But in the case of delinquency, I wanted to understand exactly what was happening. And I couldn't. Why did delinquency increase? How? Who are the defaulters? How did they get into debt? Are there patterns we aren't seeing? All of that, with a few honorable exceptions, was very vague, imprecise, or directly a media operation.

That's why, like any ordinary citizen, I found out what the best primary source was—the Central Bank's Debtors' Registry—figured out how to access the microdata, registered with Fiscal Key 3 in ARCA, signed a sworn statement saying I won't reveal personal information of the people in the database, and downloaded the entire database of debtors in Argentina. Millions of data points with names, CUITs, delinquency amounts, entities owed, and much more. With all that, I spent 48 hours obsessively looking at the data, and this is what I found.

**Two Disclaimers One**

. I am not a specialist in the subject, so there may be errors. There may even have been an error in crossing databases, because they are many and very complex. Therefore, the idea is not for you to take this document as revealed truth, but as an example that we can be more curious and, as the scientific method dictates, use information to test hypotheses and get as close as possible to reality until someone proves us wrong. If I made a mistake, I will correct it. It is an open conversation.

Two

. There was much debate about "sensitivity" on this topic. When I was 10, I saw my dad's car get repossessed because of his debts. I still remember the license plate of that car, and until I was 15, I swore I would get it back (later a friend pointed out that by the time I could drive, the car would be junk—and I took my time until I was 32 to learn to drive: what you'd call a trauma). So I understand the sensitivity of the issue. But these days I saw a presidential candidate's media outlet trying to connect the suicide of a retired couple with the cruelty of the Minister of Economy. It seems very obvious.

Let's start with some basics

Credit is good

It seems crazy to have to say it, but credit is a crucial element in economic growth and the prosperity of countries and people. It allows companies to get capital to produce, expand plants, provide more jobs, and innovate in their products. It allows families to bring forward consumption—a house, a washing machine, a car—that would otherwise be impossible to obtain.

In Argentina, there is very little credit. Most people in Argentina do not have access to credit. Not for mass consumption, nor to buy a house or a car. In the United States, mortgage credit is 45% of GDP, in Chile 28%, in Brazil 10%, and in Argentina less than 1%. In terms of people, Chile has 1.4 million people with mortgage loans. Argentina has less than 200,000 with more than double the population. If we talk about credit cards, the BCRA says Argentina has 14 million people with a credit card (37% of the population). In Chile, it is estimated that every adult has a card. This lack of mortgage, pledge, and consumer credit is in fact one of the reasons why it is difficult for us to grow as a country and one of the reasons why there has been no upward social mobility for many years.

Yes. Delinquency is high in recent historical terms and compared to the region. But the last 2 years are also the time with the most families and companies with loans without delinquency in recent decades: In terms of money, in Argentina, family delinquency is at 15.5% of the total owed if we count all categories. In Brazil it is 5.2% and in Chile 3% (in both cases without counting fintechs). Is it a problem? Yes. It is a sign that more people have taken a loan they cannot pay. However, today there are 15 million people who have loans of some kind that they pay month by month. In the last two years, we are at a peak of credit to companies and families, and that is good, not bad. Anything you want to do to "solve the problem" must take into account those who DO pay. And the problem of delinquency has more to do with inflation, financial education, wages, the political electoral risk of 2025, and the government's movement with LELIQs and rates, than with the design of the financial system.

This series counts people: how many of those listed in the Registry have at least one unpaid debt. The rate was flat at 13.6% until October 2024 and reached 27.8% in June 2026. The figures circulating in the media—12.8% from the BCRA for banks, 15.5% including wallets—count money: what proportion of the loaned balance is in an irregular situation.

Most people owe "little" and 15% owe a lot

17% of defaulters owe less than $100,000. If you go up to $150,000, you already reach almost 25% of the entire base. And almost 40% owe $400,000. On the other end, the 14.5% of defaulters who owe the most concentrate 73% of the money in delinquency.

The two curves are shifted by more than an order of magnitude. The peak of people is at $513,000; the peak of money is at $13.4 million. They cross around $3.5 million: below that line, there are many people and little money; above it, there are few people and almost all the money.

Where is that delinquency?

Simplifying: delinquency in volume of money is in traditional banks, public and private—65% of the total amount. Delinquency in the number of people is in "serious" fintechs (Mercado Pago, Naranja X, Ualá), but in low amounts. As an example, Mercado Libre concentrates 19.0% of defaulters and only 2.9% of the system's delinquent amount.

In Argentina, there are 518 financial entities that appear in the Debtors' Registry. To understand a bit better, I built my own taxonomy of entities: for example, I grouped Cetrogar, Megatone, and Frávega, which are all electronics stores; or Cencosud, Carrefour, Coto, and La Anónima (Tarjetas del Mar), which are supermarket cards; or quick cash access entities like Credicuotas, Waynicoin, or Mins, which as a category have a delinquency rate of 52.1%, with peaks of 65% and 86% in those last two.

There is a semantic issue here: the term fintech or virtual wallet has been used a lot in the media to group very diverse things. One thing is Mercado Pago, Naranja X, and Ualá, which are companies with thousands of employees, with world-class management and systems, and another thing are Moni, Mins, or Waynicoin, which are the digitalization of the traditional "Efectivo Sí" system: it asks for your ID and gives you money instantly, but now with an app. With this in mind, I took care to conscientiously categorize 90% of the delinquency. The rest may be more imprecise. The result was this.

This table looks much better at the external link: https://todosobrelamora.vercel.app/

If we look at that in detail, we can find very interesting things.

The four small credit sectors—fintech, quick credit, non-bank card, and recovery portfolio—account for 90.6% of defaulters against 26.2% of the money. The two large banking blocks are the exact mirror: 40.9% of people and 64.5% of the money.

The dangerous quadrant is the top right: high rate and high money per head. No one is there. Quick credit is at the bottom right—half of its clients in delinquency, for medium amounts—and banks at the top left: few people in delinquency, a lot of money each.

Mercado Pago is the best student, not the worst

(The group called Fintech in this report accounts for 6% of the total amount of money in delinquency and 29% of the people)

If one were to see many of the notes, streaming programs, radios, editorials, and social networks of the last few days, it would seem that the problem of delinquency is almost 1 to 1 the problem of Mercado Pago. Many added their personal criticisms of Marcos Galperín to that. But, as often happens, if one looks at the information, it is difficult to maintain that Mercado Pago is the problem. In fact, it is quite easy to show that it has been a very good student. Let's see: Mercado Libre currently has the largest number of people with debt in the financial system (+6.5 million. More than 2 million more than the next one and more than 4 million difference with the main banks). This is because for years its strategy has been to include people in the financial system. If it weren't for MP, today there would be fewer people with access to the financial system, fewer businesses with QR or credit cards, fewer people receiving interest on their account money, fewer brokerage accounts, etc. It was a pioneer in expanding credit to people and companies throughout the country and pushed the entire system to be better. And that is good.

To include someone, Mercado Libre thoroughly analyzes the risk each person has with its own scoring. It is probably the best risk system in Latin America and one of the best in the world. But no matter how much information one has, there is no way to know if a person who has never taken a loan will pay it when they get it. The only way is by testing. And the best way to test is by lending a little to each person to see how they react. As a consequence, Mercado Pago has many people in delinquency. Yes. But all with some of the lowest amounts in the system. The median delinquency for Mercado Libre is $116,000. In its same category, the median for Naranja is $260,000 or for Brubank (which has very few debtors) it is more than half a million pesos. If we compare with other categories, the median for Banco Provincia is $1.75 million, and for Credicuotas it is $840,000. In other words: Mercado Libre is doing an ambitious exercise in financial inclusion in a context of credit expansion, but with excessive care regarding the risk they and the person taking the credit assume. That is the correct way to do it. I'm not saying they should be congratulated in the media, but from there to wanting to turn it into the culprit of the system's delinquency and a "usurer" is a delusion. The problem of current delinquency is precisely that many other entities did not behave like Mercado Libre.

Non-bank quick credit

(The group called Non-bank quick credit in this report accounts for 7.7% of the total amount of money in delinquency and 22.6% of the people)

Under this taxonomy, I put all those companies whose business is to give credit "on signature alone" or by presenting an ID, although many of them now do it digitally and call themselves Fintechs or Wallets. Here I grouped 68 companies such as Crédito Argentino, Credil, Credicuotas, Moni Online, Argencred, Waynicoin, or Mins. They represent 7.7% of national delinquency, but due to their characteristics, many tend to be emergency or consumer loans for the lowest social strata (due to their design and easy access). Today they represent 22% of the system's defaulters. More than a fifth. Unlike Mercado Pago, the median for this group is almost half a million pesos of delinquency per person. Crédito Argentino (CFN S.R.L) has the largest amount of delinquent debt in this sector, with half of its entire portfolio in delinquency and a median of one million pesos. The average delinquency for this group is 52%. These entities clearly have a laxer risk management by design, but they have been very inefficient in managing that risk in this context of credit expansion and restriction of purchasing power. It is also a sector that, in the opinion of the writer, should be more controlled by the BCRA authority. The audit of their risk systems, their accounting, and, in particular, alarms or restrictions should be triggered when delinquency grows above a certain threshold. Otherwise, things like Waynicoin or Mins happen, which have 65% and 86% delinquency respectively.

Private banks

(The group called Private Banks in this report accounts for 45% of the total amount of money in delinquency and 27% of the people)

Private banks today have 1.59 million people in delinquency. They have 27% of the system's defaulters, a median delinquency of $1,458,000, and 45% of the total delinquency. These are middle-class families who are probably in debt with cards or some personal loan.

When the government says "it's a problem between private parties," it likely has this segment in mind much more than any of the others we are talking about here. And the reason is that indeed here the problem has much more to do with a solution between private parties than anything else: banks are already refinancing their clients' delinquencies because they have the incentives to solve that problem from within their own entities. If we look at the last available month, a quarter of the recoveries of private banks (that is, people who stop being in a delinquency situation) are marked in the Debtors' Registry as refinancings. At the extreme, for Galicia, 50% of recoveries are refinancings. It is still not enough to massively pull people out of delinquency, but it is starting to move. Furthermore, families in this segment are usually the ones with more tools to recover purchasing power, reduce expenses, and face delinquency. I imagine that in the risk committees of banks, as in some fintechs and non-bank entities, there must be many slaps on the wrist for those who got the risk strategy wrong.

Banks are also "writing off" debts they consider uncollectible. For the person experiencing debt, that means—in addition to stress and anguish—the closure of their access to credit for many years. But for the system as a whole, it is the way to resolve the conflict: the bank assumes the loss and the person will receive calls for years from some recovery entity to pay their debt until they can re-enter the system.

Non-bank cards, electronics stores, and supermarkets

(The group called Non-bank cards, electronics stores, and supermarkets in this report accounts for 14% of the total amount of money in delinquency and 35% of the people)

Much of the current debate is about banks and fintechs. But if we add non-bank credit cards, electronics stores that have their own cards or loans, and supermarket cards, we add 35% of the defaulters and 14% of the system's delinquency.

Regarding non-bank cards, the main one is Tarjeta Naranja, with 4.23 million debtors, of which 1 million are in delinquency. It is 7% of the total delinquent money in the system. Here there is a taxonomy decision of mine that is worth making clear: the Naranja ecosystem was split in two. I counted Naranja Digital (Naranja X) as fintech—it is a mass-use wallet—and Tarjeta Naranja as a non-bank credit card, because its origin and business are that. It would be just as valid to group both as Fintech.

Electronics stores, for their part, include Megatone, Naldo, Cetrogar, Frávega, among others. In total, they involve 8.3% of defaulters and 3.2% of the system's amount. 41% of those who took credit in these entities are in delinquency and the median delinquent credit is $672,000. Everything seems to indicate that it is another of the sectors where the risk strategy went haywire. And besides, one could assume, but not affirm, that a family in debt is less afraid of not paying Frávega than the bank.

Finally, in this segment are the supermarket cards: Coto, Carrefour, Cencosud (Jumbo, Easy, Vea), and La Anónima. They represent 2.5% of the system's amount, with a median delinquency of $807,000.

I would take what follows with a grain of salt, because the initial debt is not known, nor if people refinance, pay the minimum, or what they do month to month. But looking at what happens to the balance of an overdue account from one month to the next, supermarkets and electronics stores are making different decisions. Electronics retail continues to accumulate interest: in Megatone, the balance grows in 84.5% of overdue accounts, at a clear rate of 118% annually. Supermarkets freeze the problem: in Cencosud, 80.9% of accounts remain identical to the peso from one month to the next, and since if we look at the amounts that go down only 10.8% of their debtors pay something, the explanation cannot be that they are paying all the interest. Clearly, they are using a strategy of diluting debts with negative rates (not even increasing for inflation) until things settle down.

Public banks

(The group called Public Banks in this report accounts for 20% of the total amount of money in delinquency and 14% of the people)

Between Banco Nación and Banco Provincia, they have 15% of the delinquency amount in all of Argentina. Like Private Banks, their delinquency is around 15% of the total debtor portfolio and has a median delinquency of 1.4 million (because like Private Banks, they have a client portfolio more geared towards the middle class). When you break it down between banks, Nación has a median delinquency of 1.12 million, Provincia 1.75 million, and Ciudad de Buenos Aires 2.1 million, probably related to the purchasing power of their clients. It should be noted, as some have, that these weeks there was much talk about Mercado Pago, which is 2.8% of the system's delinquency, and not about Banco Provincia, which is 6.6%.

On the action side, Banco Nación launched its Regularization Program for situations 3-5—up to 120 months, 12% TNA with salaries in the bank, 14% for the rest, consolidation up to $100 M, 100% virtual process since August. Provincia did "Ponete al Día": up to 72 months, 31% annually for advanced delinquency with over-indebtedness, possible capital write-offs, and it already has more than 100,000 operations for ~$390,000 M.

As a curious fact, those who seem to be doing a particular job are some privatized public banks (Santiago del Estero, Santa Fe, Corrientes, Entre Ríos, etc.). They are the fifth largest actor in debt size in the system with 3.31 trillion pesos lent, but they have only 1.9% of the country's defaulters and are only 1.8% of the system's delinquency. We would have to understand why, which is something that exceeds this analysis.

Recovery portfolios

(17.8% of people in delinquency)

There is a category that covers 17% of defaulters, which is recovery. This happens when a bank or non-financial entity writes off the credit and "sells" it to a company in charge of litigating or pressuring the debtor to fulfill their payments. When we see the general base of defaulters, there is 17% that the system already considers lost but who still appear.

Deceased

A curious fact that emerged while looking at the databases is that 3.22% of the country's defaulters are deceased. It is 1.34% of the national delinquency in money amount and is included in each of the banking entities or in the recovery category. Most appear as delinquent because upon death the person stopped paying that debt and after 90 days enters the delinquent category, not necessarily because they were debtors in life.

More young people entered the financial system, but they have high delinquency

If there is a topic we should all pay special attention to, it is the 18 to 25 age group. It went from 16.8% to 23.3% banking penetration in two years: 338,624 new young people with registered debt, a growth of 38% when the population of that age did not grow. It is the largest expansion of access of all ages in relative terms. That can be positive or negative depending on the financial education we give them. For now, it seems to be complex: it is the group with the worst delinquency rate in the system, and the one that worsened the most in level: from 24.7% to 40.8% of young debtors in delinquency.

The median of that delinquency is $282,000, but it differs greatly by province. There are provinces like Catamarca, San Luis, San Juan, or La Rioja where delinquency in young people reaches between 12 and 13.5% of the population between 18 and 25 years old. There are provinces that are much more complicated than that:

Catamarca, San Luis, San Juan, and La Rioja have between 12.5 and 13.5 young people in delinquency per 100. In the City of Buenos Aires, it is 5.4. It is likely that the CABA calculation is distorted downwards by the configuration of the ARCA registry.

The other day in a WhatsApp group they were telling me about the problem of gambling and debts among the youngest. It is time to look at this problem more seriously. Fintechs have a special job to do with this, because due to their digital and simple nature, young people are overrepresented there.

Poverty, Unemployment, and Geography (what we can know and what we can't)

Much was said these days about whether delinquency grows because there is less activity or more poverty. The available information does not allow confirming or refuting any of those assessments. For something obvious to start with, which is that accessing the financial market has a barrier: if you don't have a registered income, it is harder to enter the formal system. Adults with activity declared before ARCA have credit in 66% of cases; those who don't, in 44%.

But beyond that, some things can be seen. If we cross the EPH with debtors by geography, we see that in Concordia, the poorest urban area in the country, for every 100 adults there are 13.7 defaulters against 13.5 for the national average: exactly the same. The interesting thing is how that tie is reached. Concordia is the place where the fewest people enter credit in the entire country—39.9% of adults—and among those who enter, delinquency is among the highest. Poverty explains about a quarter of the differences between cities.

Regarding unemployment, the four urban areas with the most delinquency—La Rioja, San Luis, Catamarca, Tucumán—have unemployment of 3.4%, 1.5%, 3.8%, and 5.6%. Santiago del Estero has the lowest unemployment in the country, 0.6%, and is seventh in delinquency. And those with the highest unemployment, all near 9.5%, are from the middle of the table down: Mar del Plata and La Plata at the bottom, Río Gallegos and San Nicolás in the middle.

What seems to be more indicative is where people borrow. The higher the proportion of debtors in a city who owe someone who is not a bank, the higher the delinquency. It surely has to do with the proximity and use of these laxer non-bank entities, with fewer controls and less capacity to measure risk.

These are not definitive statements. Because, for example, the dynamics of unemployment is something that happens live and delinquency is something that is announced ex post, 90 days after the debt is due.

Therefore, the point is this: it cannot be said with these data that unemployment is pushing delinquency nor that it has to do with poverty. It is more likely that what pushes delinquency is something more obvious, the fall in purchasing power, than either of the other two variables.

Debtors who owe more than one entity

To try to understand if there is a chain of delinquency like the one they report on TV (I lose my job, I get into debt with one, I get into debt with the other, etc.), I tried to do different analyses. To start with, 50% owe a single entity. And 90% owe a maximum of 3 entities. Let's remember that owing an entity is not necessarily a problem as long as you don't enter delinquency (you can owe a bank for a credit card and a fintech for a personal loan, for example, or vice versa, and be up to date with both). Here you can see how those connections are.

The three banking blocks occupy almost a third of the ring but hardly connect with each other: only 9.2% of the volume is bank with bank, 57.5% is bank with non-bank, and 33.3% between two non-banks. The thickest ribbon is Private Bank with Fintech: 3,028,307 people.

When staying only with defaulters, the ring hardly changes shape: 5.9% / 52.1% / 42.0%. Defaulters have 2.0 categories on average, barely more than the typical debtor, and their links with banks are still there because many are up to date. The one that grows is Quick Credit, which goes to 2.6 million crosses with 1.5 million defaulters.

Does all this analysis mean there are no problems?

There is a data point that helps put the rest in perspective: in the last 23 months, almost 12 million people left the system, and 86.6% left while being up to date. Clean exits have been flat at about 450,000 per month for two years. What changed is not that the system is expelling people by mass default: it is that entries fell from 840,000 to 542,000 monthly. New credit is no longer being given and that is also a problem for economic activity.

What do governments have to do? (not just the national one)

The national government is doing something that seems "insensitive" to many (and that seems right to me), which is forcing banks and non-bank entities to take responsibility for having taken more risk than necessary. That is already having real effects, as we said previously, both in public and private banks as well as in fintechs or non-bank cards.

Some suggest that the government could lower reserve requirements for banks, or that it could do some other additional things that I cannot judge because they exceed my capacity. What I see looking at the information is that it could also more actively control the more satellite financial entities, such as non-bank credit cards and quick credit. Clearly, those institutions are managing credit in a way that is not very consistent with the risks involved in operating in a disinflation program and volatile rates.

For poverty, the government doubled the AUH from what it received. For unemployment, the solution lies in activity. The problems of the financial system must be solved with financial system mechanisms. In the coming weeks, many measures will be proposed that will only (like grouped tenants or UVA mortgages) break the system even more. That doesn't mean nothing should be done, but siren songs should certainly be avoided. But from the data, two additional things emerge that could be done (involving the national government, but also the provinces and citizens).

The weight of taxes on credits

The other thing that should be done is to eliminate the amount of taxes there are on interest and credits. Argentina taxes credit with an unusual intensity: it superimposes VAT (21% on interest to consumers, 10.5% to registered companies), check tax (0.6% on debits and 0.6% on credits), provincial Gross Income (up to 9% on interest), Stamps (~1.2% on capital), and municipal Health and Safety rates (up to ~7.5% on interest, with an average of 4.7%). That doesn't happen in any of the countries you want to compare yourself with.

Regarding the weight on the loan, the most cited figure is from the IERAL of the Fundación Mediterránea: "taxes represent on average 44% of what is paid when taking a loan, while the interest rate without taxes represents the remaining 56%." It is worth clarifying that this study is from 2022; in 2026, IERAL itself places the impact between a quarter and a third of the financing cost, and Econométrica calculates that taxes add more than 43 percentage points to the total financial cost of a personal loan. The order of magnitude, in any of the three versions, is enormous. Most countries exempt credit from consumption tax and almost never put taxes on interest.

Provinces are the ones that receive, after revenue sharing, the largest part of this flow. So they are as much a part (or more) of the problem.