I knew that Georgian banks make a lot of money. But I didn’t understand just how much until I looked at some numbers.

In 2021 (the latest year for which I can find free global data), our banks were 8th most profitable ones in the world. 2021 was a bit of a standout year for us compared to what came before (2012-2021 average return on equity was 18%), but for 2022-2025, the average was the same 24%. For Bank of Georgia, much the leader, ROE at times hit world-leading 30%.

You’ll see on the chart that it’s mostly low and lower middle income countries in this top 15 together with us and two other upper middle income ones (Kazakhstan and Dominican Republic, and for Kazakhstan it was a genuinely one-off year). It’s curious company to be in. Are all these countries’ banks so profitable because those countries are the sorts of places that don’t regulate banks properly and allow them to get a lot of money out of businesses and populations?

For Georgia at least, the answer is, “most probably yes.” Our two main banks control around 80% of the market, and they use that control to squeeze businesses and consumers by charging them high interest rates on loans and offering low interest rates on deposits. That’s how they earned, in 2024, a large share of their GEL 2.7bn in post-tax profits. Of those profits, 50-70% they invest in more banking than we need; the rest they send to their 80-90%+ foreign shareholders. This is good for them and their shareholders and bad for the country, and this should change.

There is some nuance to this, but not much.

Now for the nuance.

ROE, what ROE?

Return on equity is one of the standard measures of how profitable a bank (or any other company) is.

ROE = net income / shareholders’ equity

Net income is just accounting-speak for profit. And shareholders’ equity is the money you have to put up so you can get a banking license (in Georgia, around GEL 50 million). You can later grow it, too, if you get more people to invest or if you put your own profits back into your bank. Regulators want your equity to be a certain percent of your loans—say, 5-8%—so you can cover any losses on those. Which means that the more loans you want to make, the more equity you need to have (exactly how much depends on how risky your loans are and on how strict your regulator is).12

When you take net income and divide it by shareholders’ equity to get ROE, you see how good a bank is at making money with the money that it has. It’s one thing to earn 100 million on equity of 1 billion, and another thing to earn that with equity of 10 billion. For a bank, very good ROE is something like 15%. 5-10% is not unusual.

Cash factories

Georgian banks’ overall ROE, as you already saw, was 23.5% in 2021. It stayed at about the same level since: the average for 2022-2025 (through October) was 23.8%. Of the two big banks, TBC was most recently at 24.4% in Q3 2025 and BoG was higher yet, at 27.8% in that same Q3 2025. That is very, very good (for the banks). And even before 2022, when the war in Ukraine—or rather, Georgia’s role in helping Russia and Russians deal with the consequences of it—boosted bank profits just like it boosted everything else in Georgia, the 2012-2021 average was almost 18%, putting Georgia at #25 globally.

Why is the ROE of Georgian banks so high? And why is Georgia in the company of countries that have much, much lower levels of GDP per capita than it does?

To get a high ROE, you need high profits (let’s forget about shareholders’ equity; you try to size that to how many loans you can profitably make). And high profits, if you are a bank, can come about in one of four main ways:

1. Make more from loans by charging higher interest on them;

2. Pay less to depositors;

3. Make more from non-interest sources like fees or trading;

4. Spend less on your buildings, salaries, etc.

Now, it’s probably not buildings/salaries because we have some very nice banks that pay nice salaries, and our number of bank branches per 100,000 population is pretty high by global standards, in the top 20.3 And a World Bank study that compared Georgian banks’ cost efficiency to peers showed that peer countries are more efficient (on the chart, lower is better):4

It’s also not non-interest income, at least not by itself: that income is about a third of the banks’ profits.5 Important, but not dominant.

I think that it is 1 and 2 together: banks charge a lot for loans and pay a lot less than that to depositors. For Georgian banks, the absolute difference between the two is on the order of 6-7%—so, say, the average loans are offered at 12% and the average deposits at 5-6% (0% on current accounts and up to 10% on term deposits). This difference—another way to think of how profitable a bank is—is called Net Interest Margin, or NIM, and banks in Europe and the US would be happy with NIM at 3%. They wish they were in Georgia.

My interim thesis, then, is: banks in Georgia are so profitable because they can extract a lot of money from borrowers and not pass it to savers. The question is, why?

The main (obvious) suspect is

The BoG/TBC duopoly

Georgia’s two biggest banks together control 76% of the banking system’s assets (of which ~70% are loans they issued and the rest is cash and investments; the BoG/TBC shares are 39/37) and 81% of physical persons’ deposits (46/35). This, if you compare to it to other countries, is pretty bad.

When people study competition (or the lack of it), they often use something called the Herfindahl-Hirschman Index, or HHI. They look at the percentage shares of all the players in a market, square each value, and then sum the squares up. If you have a market with just one player, your HHI is 10,000 (or 1, if you’re squaring 1 instead of 100%). If you have 100 players that each control 1%, your HHI is 100 (or 0.01, for 100 * 0.01²). In the US, markets with HHI between 1,000 and 1,800 are considered concentrated, and anything >1,800 is highly concentrated.6 I don’t have HHI data for the whole world (if any of you do, send a shout), but I do have it for the EU and (partially) for Georgia. And this is what the data shows:7

You see that orange line pointing towards infinity and beyond? That’s us, with 2024 HHI of 0.31/3,110 (though falling to just under 3,000 in the latest data for November 2025).8

You may say, OK, but you shouldn’t really be comparing us to the EU. Well, let me give you some charts from the World Bank study I already mentioned that looked at the lack of competition in Georgia’s financial sector (the study is well worth a read, by the way). It is from 2021, and the data is only up to 2019, but things have only gotten worse since then.

You’ll see that:

• HHI has been growing and growing and growing;

• Market share concentration in everything, and SME loans in particular, has been growing, too;

• Mortgage loans in particular were super-concentrated, while SME loans were the least so (even after all the growing concentration);

• And we were way, way ahead of peers as far as concentration is concerned.

(the study, by the way, mentions “top 3”—meaning also Liberty—because, although Liberty only has 6% of the system’s assets or so, NBG considers it systemically important).

And there is the regional angle, too:

HHI reaches absurd heights for, say, firm loans in Mtskheta-Mtianeti and SME loans in Samtskhe-Javakheti. Guria and Racha are pretty concentrated, too.

Does all this concentration necessarily explain the banks’ profits? Well, no.

There could be other reasons, too. But I don’t think they are as convincing.

The Russians, maybe? Perhaps a bit. You saw above that ROE did jump post-2022. But it was very high before 2022 already. The Russians certainly brought more volume to the banking sector: more absolute profits. Maybe they brought some more relative ones, too, because some of them accept worse terms from the banks because not all banks work with them. But, for once, I don’t think this is the main story: loans are how banks make 70% of their profits, and non-resident loans in the banking system are only GEL 2.5bn out of 69bn total (and Russians would be considered non-residents for at least their first year here and possibly beyond).

Non-performing loans? Don’t think so. First of all, these are pretty low in Georgia: on the order of 2-3%. Some of our neighbors on the ROE rankings—Ghana, say—have NPLs of around 20%. Our NPLs are closer to the US, Britain, and Germany, where the ratio is 1-1.5%. And ROE, anyway, takes NPL losses into account (NIM does not, but again, with NPLs this low, who cares).9

Country risk? Sure, there is some. And you could say that foreign investors demand a higher return on their equity when they are investing in a country where Russian tanks stand 500 meters from the main highway. But investors can demand anything they want and not give you the money if they can’t have it—banks have to be able to actually get it. Also, I hope you would agree that country risk seems lower for us than for Burkina Faso, Burundi, or Niger, all of whom are our close neighbors on the ROE rankings. So, no.

NBG monetary policy? This is a better one. It doesn’t affect loans in foreign currency, but about half of the banking system’s loans are in GEL, and many of those (on the order of 100k mortgage ones, I hear) are written as “NBG key rate + X%”. If NBG raises the key rate because it wants to fight inflation, it a) makes people less likely to take loans, which reduces future bank profits; b) makes people who already have loans pay more to the banks, which boosts current bank profits; c) makes people likely to ask for more money for their deposits, but a few months later; and d) causes banks to have to pay more to the NBG when they have to borrow from it (which, though, they don’t do too much). (b), I think, trumps all the rest, and the banks profit.

But even with monetary policy contributing some to this high profitability situation, I do still think that, if I were Georgia’s competition agency—I heard we have one—I’d look at concentration first and foremost. I’m hearing, for example, that it is more expensive to get a big loan in Georgia—say, something above GEL 30 million—than it is to get a small one. You’d think a big client would get better terms, but hey, when there’s a duopoly, where can the big client go? That’s a pretty good indication for you of how the two banks use their market power.

Could I guarantee that if we had five big banks instead of two we’d see lower loan rates, higher deposit rates, and lower banking profits? No. But it’s interesting to note that Armenia, with five big banks instead of two, had an average 2012-2021 ROE of 7% (we had 18%).10 It’s also interesting to see how banking profitability correlates with a country’s income level (except for us):

Low income countries have GNI per capita of $1,135 or less. These are places like Burundi, Somalia, and Afghanistan. Georgia, let me remind you, is an upper middle income country with a GNI per capita of $8,500 in 2024. We are, it seems, in the curious situation of our anti-monopoly regulation being, shall we say, somewhat behind our income level. And this is 2012-2021 data. As I already said above, our 2022-2025 average bank ROE was 24%.

All of this, though—the why of very high profitability—matters less than

The consequences

In 2024, Georgian banks had after-tax profits of GEL 3.1bn (and pre-tax ones of 3.7bn, for an effective tax rate of 15%). That’s a lot of GEL bn. Georgia’s GDP, may I remind you, was GEL 93 bn in 2024. So this is 3.3% of that. And you’ve got up to another 1.6bn in salaries (which, by the way, is a low share of salaries, just 30% of total profits+salaries+taxes, and the 1.6bn also includes insurance). Banks be like,

It’s not as exciting, though, if you are a regular Georgian who has to get a GEL mortgage at 15% annual interest. Or a Georgian business paying the same for a business loan. A 15% loan doubles in five years. A 3% loan—which is what you’d pay in many European countries—doubles in 24. Our banking system takes a lot of money out of the economy and then a) spends it on a lot more banking than we need at our level of development—cue all those bank branches everywhere, the award-winning banking apps, and two or three payment terminals in every shop11—and b) passes what remains—usually 30 to 50% of profits—to its foreign shareholders through dividends and stock buybacks (TBC, by now, is 80%+ foreign-owned and BoG is 90%+).12 It also makes it very difficult for any real business to emerge: if you’ve got loans at 15%, you need 20% margins for the loan to make sense, and few businesses other than real estate and gambling have those kinds of margins. Which is why gambling and real estate is all we have. All this, I would argue, is not a good thing.

We are a very poor country with 31% effective joblessness, massive emigration, failing education and healthcare, and a 27% share of population experiencing “moderate or severe food insecurity”—i.e. living on bread and ramen and/or often going hungry. And we have built ourselves a banking system that funnels money towards real estate speculation and passes the profits to London hedge funds. This, sadly, is pretty typical of countries at our level of (under)development. And if we don’t want to become even more underdeveloped than we already are, we should probably do something to fix it.

How? Oh, the list is endless. You would start by having a proper investigation into what’s going on—and no, I don’t mean a PR stunt led by the State Security Service. If the investigation finds evidence of anti-competitive behavior, as it probably will, you could do a few things. For example, you could limit how big of a share of the market—say, 20%—any one bank can have. Then—if the banks don’t collude again, and if you look thoroughly at all the other ways they are abusing their market dominance—you’d be passing the savings to businesses and consumers. Or you could leave things as they are, but tax banks more than you tax other companies. Then the banks would effectively be acting as tax collectors for the government, and hopefully we’d have a government that could spend the money on something smart (like creating some jobs in things other than construction and retail).13 Or you could try to give banking licenses to fintechs that could challenge the duopoly (though these fintechs would probably have to be foreign, given the amount of money needed). Or you could even—gasp—try to build an economy that does non-third-world things like produce something and have a monetary policy that would go with it (with much lower interest rates, that is, and possibly higher inflation at least at first). And a lot more than this besides.

It’s nice to have pretty bank branches and some very smooth banking apps. But not at the cost of 1,000 lari in banking profits per year for every working-age person in this country (and probably another 500 lari on top of that spent on the apps and the branches). Sure, we all need banking. And banks and their investors are entitled to reasonable profits on the services they provide. But nobody is entitled to obscene ones. Banking is a regulated utility: same as electricity or phone service. Imagine if you went around saying, “Hey, our electricity distribution companies are the most profitable ones in the world.” That wouldn’t be a good thing, would it now? But that’s the situation we are in with our banks.

I don’t blame the individual bankers too much: they do what they can get away with, and if they didn’t do that, their shareholders would fire them and would find someone who does. But maybe it’s time to let them get away with a lot less. If we are going to be an upper middle income country in reality and not just on paper, something like an 11% ROE—our income group average—seems to me to be a lot more appropriate than 24%. And if we ever get to that level, I’m sure that this country’s people and businesses would find something to do with GEL 1.5bn a year that they’d have back in their pockets as a result.