The Opinion Of Sheep

Sorry kitten, the walls are closing in on daddy

3d model of javier milei on a tank during independence day. i have no idea what is even going on in this fucking country anymore

A few weeks ago, it had become trendy to admit defeat on the anti-MAGA takes: the US economy seemed impervious from a major slowdown or from significant price adjustments. Many liberals admitted defeat: there would be no Trumpcession in 2025. Except, in the span of last week, the US saw a significant slowdown in domestic demand (particularly consumer spending) and job market revisions that made Q2-25 the worst three-month span in 15 years (outside of COVID). Plus, July PMIs were atrocious on both output, employment, and price indicators.

The same has been going on in Argentina since April and May: the exchange rate was stable and inflation kept trending down, to less than 2% a month and the lowest since 2017 (outside of COVID). The economy has, to be fair, been sluggish, with not-so-good prints in output, industrial activity, and construction in May and June, as well as consumer spending indicators starting to decline versus 2024. The job market, aditionally, was starting to take a darker turn, with whatever the opposite of rosy data is in the last couple of months. So I felt like I had to throw in the towel and start to think about what, exactly, went right. Except, in late June, this happened:

peso/dollar exchange rate

The exchange rate, basically out of nowhere, was devalued by around 11%; considering inflation, it’s nearly 20%. In fact, at the end of July, the exchange rate was just over 5% away from the top of the currency band, the point at which the central bank should have started to sell reserves to keep it down. Why?

Anatomy of a Jump

Let’s take a cue from Sherlock Holmes (or, since it’s about money, his cousin Shylock Holmes, to take a cue from El Peluca’s favorite president): “eliminate all other factors, and the one which remains must be the truth”. What could explain a spike in the exchange rate and a big spike in interest rates (more on this a bit later)? Well, it usually responds to either the domestic picture getting way worse, or the international picture getting way worse.

Well, the international economy wasn’t that bad until really recently - back in late June, the tariffs were going away, the war between Israel and Iran was over, and TACO was the law of the land. With strong markets and a seemingly strong economy, the US was pretty chill up until last week. So it’s not really the outside world, yet.

The second likeliest culprit is politics, but that’s not that relevant. Milei’s approval has trended down since May (from +1 to -13 last month), and the main political development has been kind of a wash: in mid-June, the Supreme Court declared that the former President, and de facto leader of the Peronist Party, Cristina Kirchner was guilty of corruption and couldn’t run for public office for 6 years. This isn’t really good news for anyone - Kirchner was profoundly unpopular, and her conviction is relatively widely supported, but she was also undisputedly the main figure in contemporary Peronism, so her removal saves the party from a very nasty fight that certainly would have benefitted Milei - all of their remaining major figures gathered (alongside the leadership of the left wing parties?) and declared the typical lawfare prosecution blah blah blah. The moribund Peronist party waking up from 2 years of whatever it’s been doing has come at a bad time, since Congress passed multiple bills expanding payments to retirees, disabled people, universities, hospitals, and reverting many of Milei’s privatizations, with a total cost of over 1% of GDP - more than the expected financial surplus for the year. Milei managed to have his veto sustained in the first two, but got trounced in Congress on the other two, after some pretty extensive conflict with his Vice President over the voting schedule. So political risk isn’t that big a deal, I would say, or at least not big enough to keep the markets panicking for a month and a half.

The third option is from the currency market itself: the government has been stocking up reserves at an extremely paltry rate. The IMF asked for net reserves of USD 0.5 billion in June, and instead got… -4.2 billion, and currently at six and change. The authorities did overperform on the fiscal target (with 0.8% of GDP, versus 0.7% target), but surprisingly also surpassed the maximum increase in central bank assets (9 trillion pesos versus 8.2 trillion allowed) - put a different way, Milei printed too much money this year. But the IMF gave Milei a pass, handed out a contingent USD 2 billion disbursement anyways, adjusted the reserve targets downwards (from USD 4 billion, to a much more lenient target of USD -3.4 billion), eliminated the September review, and praised a variety of efforts the government is undertaking, particularly for fiscal and labor reforms. So if the IMF spooked the markets, they should be unspooked by now, right?

The last factor to consider is basically that the currency market may have been under some kind of supply/demand imbalance. Mid-to-late June has usually higher demand than average because of mid-year bonus payments, then the winter holidays, plus it’s a bad part of the year due to agricultural sales. If you throw in a depreciating dollar with increasingly tighter Fed policy, some international tension, and some political uncertainty, maybe that spikes exchange rates? Well, no. The main issue is that agricultural exports were unusually high in June, wrapping up at USD 4 billion or so, and crashed in July (because the government cut export taxes in March, and the cut expired last month). So having a big spike in supply and a small but expected spike in demand can’t really explain why we’re still basically at the top of the inflation bands with relatively high positive interest rates.

Before moving on with diagnostics, the part that’s worth noting is that inflation expectations are relatively anchored in the sub 30% range, and that the recent movements in the dollar haven’t translated into prices too much, which is both a positive signal of where the economy is heading and a consequence of fiscal and (relative) monetary prudence. It doesn’t help anyone to deny reality: inflation is down, and society is getting used to a floating exchange rate after six years of pegs. The issue isn’t really whether these accomplishments have happened, but rather, whether they’re sustainable - and I have my doubts, as current events go. My ultimate take here is as follows: back in June the government shook up its monetary policy strategy, and I think the markets realized the whole thing made basically no sense and was a problem not just for the currency market, but also for fiscal policy.

Oopsie, made a whoopsie

To understand the changes, we first need to understand the status quo: last year, the government invented LEFIs (Fiscal Liquidity Bills in Spanish), a bond that let the central bank set interest rates but where those interests would be paid by the Treasury (using the accounting trick of not including it in regular interest payments¹). I warned, more than once, that LEFI debt was still Treasury debt, and thus would start weighing on payments - in fact, the IMF noted that after accounting for below-the-line interests (which includes LEFI and fiscal policy instruments like LECAP and BONCAP, which are designed to also handle liquidity but without monetary policy guidance implications), the deficit would have been 1.2% of GDP, meaning that, on the net, these hidden interests were of 2% of GDP - or five times the size of regular interest payments.

The new monetary policy is pretty straightforward, beyond some stuff about reserve requirements and e-checks: they got rid of LEFIs, and the central bank doesn’t set interest rates anymore - the market does. Instead, the central bank targets monetary aggregate levels, particularly M2. M0 is cash, M1 is cash plus some cash in the power of banks, M2 is M1 plus short term deposits, M3 is M2 plus fixed-term deposits, and so on and so forth until you get to stuff like “corporate bonds are money because they’re in M7” or whatever. By looking at M2, the authorities plan to keep the money supply as close to demand as possible, since that’s the best proxy for short-term demand for domestic currency we have. I’ve actually written about this topic before, and unsurprisingly, it’s a bad idea.

Let’s go back to the Quantity Theory of Money: M (the supply of money) times V (how much each unit of money moves around) has to equal the total amount of transactions, which is equal to nominal GDP, which is the price level P times real output Q. Hence, MV = PQ. Over the long term, V is fixed and Q are not influenced by the amount of money in the economy (because monetary policy only affects nominal variables, as laid out by Marxist Milton Friedman in “The Role of Monetary Policy”), so all movements in the money supply have to be reflected as movements in inflation. This means that, over the long term, targeting the money supply has to be the same as targeting the price level, or targeting monetary aggregates growth has to be the same as targeting inflation. However, this is over the long term, because there’s two things to note in the short term: first, that monetary policy can actually stimulate the economy for a short period of time (called monetary neutrality), and second, that velocity can shift in the short term. One such example is the pandemic: in April and May 2020 inflation went down significantly because people held onto money more as a precaution, that is, because velocity went down.

The whole idea of having a monetary policy rule is to provide stable expectations of where nominal variables (prices, exchange rates, interest rates) are heading over the medium to long term. That’s what they call a nominal anchor. The problem with inflation targeting is that, well, sometimes inflation increases for non-monetary reasons (for instance, higher oil prices), which puts central banks in a bind of keeping real rates too low or hiking them to not decrease prices (because they don’t print oil) but also increase unemployment. That is, the problem with inflation targeting is that the nominal anchor is actually volatile on its own sometimes. Over the long term, this volatility doesn’t matter, but the business cycle (booms and busts, recessions and inflations) is inherently short term, where volatility does matter - just ask poor Jerome Powell. So you can tell where this is heading: monetary aggregates targeting is a bad idea because you can get fucked over by velocity. There’s pretty good evidence that velocity responds to these business cycle trends: during recessions people spend less, and during “boom” periods they spend more quickly - which means that money growth targeting could amplify, not moderate, economic fluctuations.

One example of this dynamic was Margaret Thatcher’s Britain. Thatcher’s 1979 program to reduce inflation mixed a gradual reduction of the growth rate of M3 with a progressive reduction of the fiscal deficit, coupled with tax cuts on income and tax hikes on consumption. To allow for some way to fund the deficit, the government lifted restrictions on capital controls, so that external borrowing would cover the gap. The problem was credibility: to maintain capital flows, real interest rates would have to be very high, which meant a higher interest bill for existing debt, which would mean an explosive trajectory of debt. Without deeper budget cuts, this would eventually mean that the British government would have to inflate away some of its debt, especially because British pound was fairly overvalued due to complicated issues relating to oil. Hence, the main risk to a monetary aggregate targeting rule is that, if the market requires too high an interest rate, then eventually fiscal dominance kicks in, because the government doesn’t want to go bankrupt and thus would force the central bank to finance it and break the aggregate targeting rule.

The thing is, Milei and his people know about velocity being endogenous - they’re counting on it! Nobody wanted to talk about it or admit it, but there has been some financing of the deficit through the central bank - specifically the “turd plan” of May and June, when the central bank started transferring its net profits to the Treasury (ARS 11.7 trillion), so the Treasury would have cash on hand for paying off debt in case of monetary issues. This was because they wanted to hit what’s known as the “Anker point” (named after the Treasury Secretary and Central Bank President’s old consulting firm), where the financial sector stops lending excess funds to the government or buying dollars and instead goes to town on private credit, boosting the economy. This is driven by an endogenous drop in the velocity of money, corresponding to a decrease in macroeconomic uncertainty and more trust in the peso as a vehicle for savings and investments by regular companies and people. So the main issue is that the “turd plan” only works as long as devaluation expectations specifically are down, because then the money would start pouring onto the USD market to cover itself against a devaluation, actually causing the devaluation (whoopsie! made an oopsie!).

The main problem with Milei’s policy wasn’t really the what (LEFIs and below-the-line bonds are a really bad idea), but the how: the transition from LEFI to aggregates targeting was really sudden, so ARS 10 trillion in LEFI were basically dumped on the market immediately, and the possible allocations for this extra liquidity are very limited. If you take a basic portfolio management perspective, the demand for pesos, dollars, and government assets all compete; but under monetary aggregates control… oopsie made a whoopsie. This was a mistake because, if you fix the amount of pesos, then the trillions needs to go somewhere, and if the central bank is not selling currency (per the agreement with the IMF) then you need Treasury bonds to keep money off the exchange rate. So what happened first was that the cash went to banks, which drove interest rates way down (which would have been inflationary), so the banks started restricting credit, which meant dumping money into dollars - pushing up the exchange rate. the Treasury had to rush out to issue BONCAP and LECAP (the ones that pay below-the-line interest) at above-market implicit interest rates.

But wait, the government has to take on debt to manage monetary policy? That seems like a recipe for disaster - in fact, it’s the whole problem I thought the government was going to run into from the LEFI scheme. Economists call this phenomenon the Unpleasant Monetarist Arithmetic (and it’s quite literally what Thatcher ran into), where if we assume a fixed amount of demand for government debt, plus a fiscal deficit, then eventually that deficit threshold would be reached, meaning that the Treasury would have to either slash its deficits significantly or start printing money to pay off debt. If monetary policy tightened, then the debt program got worse, not better, because it would exhaust the total demand for debt faster.

So the deep issue here is that leaving all major monetary policy variables in the hands of the markets means that the government has extremely limited tools to manage nominal volatility: in particular, the only remaining policy is fiscal, which results in blowing up the credibility of your entire macroeconomic program. But why would this be relevant, if Milei has a relatively aggressive fiscal strategy? Well, the first issue is that obviously he has a primary surplus, but he doesn’t have a primary surplus of 4.8% of GDP. The deficit with all the capitalized interests and the other wack ass below-the-line shit was six times the surplus he got after two years of “chainsaw”. The other issue… is even worse.

Dios es argentino?

The second issue with the new monetary policy scheme is purely external: the government wants to keep money growth down, and doesn’t have reserves (or official permission) to intervene in the exchange rate, which is a key nominal variable - in fact, losing control of it is what went wrong in 2018 and 2019! Why doesn’t the government have reserves? Well, it agreed with the IMF to not intervene as a buyer or a seller in the official market within the bands, which is something I mentioned in the other post: net reserves just stagnated at the negative 8 billion to 5 billion range, and the only thing going in is debt - disbursements from the IADB, the World Bank, bafflingly enough the IMF, and occasionally, the Treasury, which bought dollars in cash (thank the turd plan) and especially after raising USD 1 billion in global capital markets to bolster reserves. This is because the new policy overshadowed the actual plan to get reserves (which I mentioned a bit further ago too), the “mattress plan” (god these people and their stupid names²), was to get the demand for money satisfied by making people sell their USD to the government, which also squares off the problem where the country doesn’t have enough dollars flowing in. Argentina has around USD 200 billion in “mattress dollars”, which would be enough to pay for everything if they started going into the economy (they, spoiler alert, haven’t).

But why is Milei scraping the bottom of the barrel for people’s mattress dollars (or their biblical dollars, since a Milei-aligned preacher tried to get out of money laundering accusations by claiming he transformed 100,000 pesos into 100,000 dollars, or 12 million pesos, through the power of prayer) instead of like, borrowing them? Well, because it would be very expensive. The bond issue mentioned above consisted of an instrument that you buy with dollars but pays out in pesos, and the nominal rate was 31.7%, very high, and the implicit return based on the exchange rate remaining at the midpoint of the exchange rate bands was 19% in US dollars - this return was high because of the high likelihood of an exchange rate jump wiping out the nominal return (whoopsie!). Hence the mattress plan, the government wanted to avoid a jump in the exchange rate by making people sell their dollars. The main issue is that… this doesn’t go along with the central bank not buying any currency and not creating any money to buy currency either. The whole point of taking the economy to Ankerland… was to buy more dollars!

The interesting thing is that July had the greatest agricultural exports of all time for that month. Why did anyone bet (correctly, I’ll add) on the government not having enough dollars? Well, the country has a trade surplus, but it doesn’t have a big trade surplus. The Jan-March balance of payments showed export growth of 8%, a good number, but the payment side was a disaster: a deficit of USD 5 billion in the current account, double what the IMF expected for all of 2025, with an increase in goods imports of 34% annual, service imports increasing 66%, and tourism up 27%. Imports to GDP reached 32% in January through June, the highest percentage in 136 years. People also started buying US dollars in the official market again, around 1 million individuals buying USD 5.3 billion since April. You’ve seen some moderation since, but monthly currency flows show the same thing: everything coming in was going out in imports and services. Tourism and goods outflows all respond to one thing: opening up the economy with a cheap exchange rate (this was in April and May) and with social media showing people everything they were missing out on: Temu, Shein, cool holidays, Miami, Dubai chocolate, etc. The second problem is the government: the country had 10 billion dollars in hard currency debt payments April through December (of which 3.9 billion or so are in the coming four months, and 4.8 billion were in July alone), plus around 100 billion in local currency debt. In 2026, this should be around 28 billion, and wouldn’t go under 20 billion until 2033 (the milennium of the Crucifixion, maybe a miracle saves the economy then), and it still remains above 10 billion until 2036. So a decade of big payments due, without much hope of getting exports up: they’re growing kinda tepidly at the moment (it’s mostly just that one Tiny Desk and maybe the SixSex world tour - Yo tengo 1, 2, 3, 4 desbalances externos a la vez), and the plan right now is to boost fossil fuel and mining exports to just not have the same balance of payments constraints after a few years.

Here’s the real problem, though. The main determinants of the balance of payments are the current account (trade, services, tourism, etc) and the financial account (borrowing and investment). Argentina’s strategy for accumulating reserves was, in 2024, to restrict the financial account using capital controls (basically regulations on foreign investment), and to let dollars get into the country via a current account surplus. However, the trade surplus is now very small, and as mentioned above, the services deficit is now much larger than last year because of looser import and capital controls. Under a fixed exchange rate, the balance of payments (capital + financial + current accounts) comes out of reserves; under a float, the exchange rate adjusts to balance them out eventually. To keep the biblical theme going, this is why Paul Krugman calls the idea that big external adjustments don’t happen with big currency adjustments “immaculate transfer”, because something has to happen to force this. Most of the time, this is a recession, or inflation, that readjusts relative prices along lines that make consumers spend less on imports (or have less money for them). In fact, those with a somehow photographic memory of everything I’ve ever written might know that 136 years ago was 1889, and that in 1890 Argentina had a massive economic recession caused by bank failures caused by currency issues (i.e. a poor approach to conversibility resulting in the liquidity trap, a very large endogenous decrease in velocity due to too lowinflation expectations), which is why Argentina originally split away from Australia and Canada’s economic track.

Right now, the big plan (after the mattress plan kinda went nowhere) is to just borrow it, by making the financial account the “paypig” for the whole thing. But this would mean really high interest rates, because, well, the country still has a bad reputation (as seen in a country risk of 650ish basis points, which means the basis interest rate is 11.00% coupled with a Fed Funds Rate of 4.5%). One example of this would be the carry trade, where a high real rate in US dollars attracts capital via borrowing in low-interest rate developed countries; the problem is that these capital flows are very flighty, and thus the whole thing can unravel really quickly. This is because (as I’ve mentioned before, too), even in countries with low and stable inflation and with deep domestic financial markets, there are substantial aftershocks from Fed policy shifts in their own financial markets, which spreads financial instability since it’s not really possible to coordinate global monetary policy to a large extent. This is why some economists (such as those interviewed by Mr Paul Krugman last week) talk about a dilemma rather than a trilemma: if you don’t impose some form of capital controls, then you can’t set your own interest rate policy or control your exchange rate by yourself, because the financial markets will let you know via the homonymous account of the balance of payments. But also, even with a the not especially serious capital flow deregulation, the mattress plan immediately stops making sense, because otherwise dollars would just go straight from the mattress to Temu.

The thing here is that the really high real interest rates mean three things. The first is big financial instability, because they can just wash you away and screw you over in a snap of the fingers (cough cough 2018/19). This is why a (somewhat famous) economist said that “debt is like crack: first it feels good, then it kills you” back in, well, 2018 and 2019. The second is that they obviously mean a recession: companies and people can’t borrow, so they can’t expand, and debtors stop paying their debts (which is kind of happening), and inflation expectations go down for bad reasons so people stop spending money. The third problem is that the government, as we’ve mentioned somewhere already, had to sell bonds in the market to stop the exchange rate spiral because it doesn’t control any macroeconomic variable besides M2. Well, if the only pillar for macroeconomic stability is the fiscal surplus, and if the government needs higher and higher interest rates to keep paying for the country’s colossal imports bill, then eventually macroeconomic stability is going to end because the government’s debt payments would enter an explosive trajectory. At this point, the Unpleasant Arithmetic kicks in, and everyone worries they’ll just start printing money to pay off the hundreds of billions of domestic currency debt facing them.

Conclusion

Back in February, Milei got in trouble for a “pump and dump” crypto scam: he built up hype around a coin called LIBRA, a bunch of people bought it, and then insiders sold it and profited. A few days ago, Milei’s former Foreign Secretary gave an interview to Al-Jazeera, where she said the only way the saga made sense was if Milei was either stupid, or crooked. I don’t think that the LIBRA thing is especially relevant to the economic program, but I do think it’s kind of analogous to it: raise as much as you can (from the IMF, the World Bank, the IADB, complete dipshits like Niall Ferguson, and from whoever will buy big bonds), hope to stick it out until after the elections, and then either it all crashes down or somehow you pull it off. The “Pull it off” is win big enough in the 2025 midterms to pass a big structural agenda that raises productivity, reduces international doubts about governance to bring down country risk (thus also bringing down interest rates), and then hope mineral and fossil fuel exports carry everything out and not too many people remember they’re unemployed in the 2027 elections. Well, I think it’s a nice theory, but I don’t think it’ll go perfectly. I mean I respect the audacity of ripping off everyone, and I do hope the IMF just loses all the money they put into the fartcoin of countries. But while I do hope it works out (I live here!), I don’t really see it going much of anywhere, really.

1. This is really financially complicated but in a nutshell it works because some bonds, rather than paying fixed interest levels, pay a fixed face value where interest is indirectly defined from market price (the yields). These payments, known as capitalized interests, are not included in the financial deficit because they’re only known ex post, as in, after paying the interest rates. Which is why we only know the Jan-May figure.

2. To be fair the “turd” plan was named by a journalist as a joke because Milei called the peso “excrement” during the 2023 campaign. So the plan is a little bit of excrement, or a turd. The Spanish term is caquita, which is a lot funnier.