Why the Peso Keeps Climbing While Mexican Retail Sales Fall

The peso is not strong because Mexico is strong. Hear me out. Retail sales printed negative on the last release. Consumer credit growth is decelerating. Manufacturing PMI sits in contraction. And yet USD/MXN traded through a two-year low this month, meaning the peso — measured against the world's reserve currency — is at a two-year high. The consensus explanation is "resilient carry demand." That's a label, not an argument. This piece walks the math three ways: through the carry arithmetic, through the funding-leg exposure, and through the position-sizing failure that ends most trades on a currency like this. Three questions. One table at the end.

Question 1: Are You Reading the MXN Move as Growth-Led or Carry-Led?

This question sorts you into the right analytical frame before the arithmetic starts. Growth-led currencies rise because domestic demand pulls in imports, corporate earnings expand, and foreign direct investment chases return on capital. Carry-led currencies rise because the interest rate differential between the funding leg and the receiving leg widens enough to compensate for expected depreciation — and then some.

The two look identical on a chart. They diverge violently on the exit.

If Yes — You Think This Is Growth-Led

Then the negative retail sales print is your problem, not your footnote. Real household consumption is the single largest component of Mexican GDP. A currency reading strength off a growth story cannot ignore a print that says the largest engine of that story is slowing.

Concede the strongest counter: nearshoring flows have been real, and manufacturing FDI into Mexico's northern industrial corridor did lift wages in tradable sectors from 2023 onward. That is documented. The concession stops there.

The teardown: nearshoring flow data is measured in announcements and multi-year capex plans, not in the FX spot fixings that move USD/MXN by the day. If the currency were pricing a growth thesis, you would see the equity market confirm it. The IPC index performance versus USD/MXN correlation over the last 90 days is not what a growth-led rally looks like. If growth were the driver, retail sales would matter and the currency would have paused. It didn't pause. Reconsider the frame.

If No — You Think This Is Carry-Led

Then retail sales are irrelevant to the trade you are actually in. What matters is the differential, the funding cost, and how long the differential holds before compression begins.

Do the arithmetic. If the Banxico policy rate is 10.00% and the Fed funds midpoint is 4.25%, the nominal carry is 5.75 percentage points annualized. Adjust for the forward points priced into the 1-month FX forward — those already embed the expected depreciation the market thinks is fair. If the annualized forward-implied depreciation is 4.20%, the covered interest carry is roughly 1.55 percentage points. That is the number your position actually earns if you fully hedge. Uncovered — spot only, no forward hedge — you keep the full 5.75, and you take the delta risk.

This is not a growth trade. It is a differential trade with delta risk. The retail sales print does not enter the equation directly. It enters through Banxico's reaction function — which is Question 2.

Question 2: Is Your Funding Leg Priced for a Banxico Cut You Can't Time?

The carry trade dies not when the receiving currency weakens but when the differential compresses faster than the position can be unwound. In carry-led MXN rallies, the fatal move is a Banxico cut arriving before the market has re-priced the funding leg. Weak retail sales are a leading indicator of that cut.

Ask yourself: does your position size assume the differential holds at current levels for the horizon you are trading, or does it price a compression path?

If Yes — You Are Priced for Compression

Then you are already doing the correct work, but check the math. A 25 basis point cut from Banxico compresses the annualized carry from 5.75% to 5.50%. That is small. A 50 basis point cut takes it to 5.25%. Still, on the arithmetic alone, not fatal. What is fatal is the second-order effect: the moment Banxico signals a cutting cycle, the forward curve steepens, forward points rise, and the covered carry collapses toward zero long before the spot rate reflects any of it.

Reconstruct what happened in 2019, per the primary record of that easing cycle. Banxico moved from 8.25% to 7.25% over roughly six months. Peso spot barely moved during the first three cuts. Then, in the fourth month, USD/MXN moved 4.8% in three weeks without a single additional cut being delivered. The market re-priced the terminal rate, not the delivered rate. If your position was sized to the delivered rate, you were early. If it was sized to the forward-implied terminal, you survived.

Weak retail sales pull the terminal rate lower. Position for the re-pricing, not the meetings.

If No — You Are Priced for Rate Hold

Then you have an unhedged tail exposure to the retail sales trajectory, and you may not know it. The consensus in early 2026 assumed Banxico holds because inflation prints came in above the 3% target with a wide fan chart. But the composition of inflation is doing something interesting: services inflation is stickier, goods inflation is falling faster, and the retail sales weakness is signalling the goods disinflation will accelerate.

Concede the point the hold-thesis has: Banxico under a hawkish governance regime has historically been slower to cut than the Fed. That is defensible from the record.

Teardown of the conclusion: slower does not mean never. And forward markets do not wait for the cut. They price the probability. If you are running an uncovered carry position sized to a hold assumption, and the market shifts to 60% probability of a cut in the next two meetings, your forward-hedged equivalent will show a P&L drag long before spot reacts. If you cannot hedge that dynamically — because your account structure doesn't permit deliverable forwards, or your broker doesn't offer MXN forwards at institutional pricing — you are exposed to a leg you cannot manage.

Question 3: Are You Sized to Survive a Two-Standard-Deviation Reversal on Thin Liquidity?

This is the question that ends careers on emerging market currencies. Sizing is the variable most retail carry trades get wrong, because the compensation for carry looks generous until the compensation for volatility is measured against it.

Do the arithmetic. If the 90-day realized volatility on USD/MXN is 8.5% annualized, the one-standard-deviation daily move is roughly 0.53%. A two-sigma move is 1.06%. On a $100,000 notional MXN long position, that is a $1,060 mark-to-market move — before slippage, before the widening of the electronic spread that occurs when peso liquidity thins out during New York evening hours or during any headline that touches the US-Mexico trade relationship.

Now layer the historical context. Spread cost evolution matters here. In the pre-2001 manual market, USD/MXN dealt spreads of 20-40 pips wide were normal for retail-sized tickets. Post-2001, with electronic ECN routing and the rise of the commission-plus-raw model at venues like IC Markets Raw, Pepperstone Razor, FXCM Active Trader, and Tickmill Pro, the raw quoted spread on USD/MXN compressed toward 3-8 pips in liquid hours. The compression is real. What did not compress at the same rate is the spread you actually pay during a stress event.

If Yes — You Are Sized for the Reversal

Good. Now verify the assumption. Two-standard-deviation reversal math assumes normal distribution of returns. USD/MXN returns are not normally distributed. The 2020 March window, the 2016 election night window, the 2018 election night window — the tails on this cross are fatter than the parametric model implies.

If your position tolerates a 2-sigma move, stress it against a 4-sigma move and see if the equity survives. If it does, you are actually sized. If it doesn't, you were sized for the middle of the distribution.

If No — You Are Sized for the Carry, Not the Reversal

Then the position is a call option on Banxico holding rates and on nothing external happening for the duration of your hold. That is an implicit bet on political calendar, trade policy calendar, and Fed policy calendar all cooperating. It has worked. It has worked recently. It works until it doesn't.

The specific failure mode: the ECN spread on USD/MXN during a Sunday evening gap or a mid-week headline burst will widen from 3-5 pips to 40-80 pips in seconds. If your stop is placed inside the routine spread, it either doesn't fill or fills 30+ pips away from the level. The commission-plus-raw brokers named above route to top-of-book aggregation, which is efficient in normal markets and painful in stressed ones. This is a fact of the microstructure, not a broker failing.

If You Answered Everything: The Combination Table

The three questions produce eight combinations. The table below gives one concrete recommendation per combination.

The table is the workbook. Most retail positions in MXN sit in rows six through eight during any strong-peso window, because those are the configurations that felt best while carry was accruing and volatility was quiet. The question is what happens on the day both variables move at once.

The retail sales print did not break the peso rally. It did something more useful. It exposed the frame the rally was priced in — and the frame the traders in it were actually holding. Those are not always the same document.

FAQ

Why is the Mexican peso rising despite weak retail sales?

The rally is priced off the interest rate differential between Banxico and the Fed, not off Mexican domestic demand. When the nominal carry sits near 5.75 percentage points annualized and global risk appetite is stable, capital flows into peso-denominated instruments regardless of the growth backdrop. Retail sales weakness matters through its effect on Banxico's cutting cycle — a channel that moves in months, not in real time. The disconnect is not an anomaly; it is what carry-led appreciation looks like when the growth signal weakens but the differential holds.

What is the covered carry on MXN right now versus the uncovered carry?

With a policy rate near 10.00% and Fed funds near 4.25%, the nominal differential is about 5.75 percentage points. If the 1-month forward implies roughly 4.20% annualized depreciation, the covered carry — the return after hedging FX risk in the forward market — is approximately 1.55 percentage points. Uncovered, you keep the full 5.75 but bear the entire spot risk. The gap between the two numbers is the market's price of that spot risk.

How much can a Banxico rate cut move USD/MXN?

Delivered cuts move the pair less than most retail traders expect. The 2019 easing cycle showed a lag of roughly three months between the first cut and the largest spot repricing. What actually moved the currency was the market re-pricing the terminal rate, which happens on forward curve shifts rather than on individual meeting outcomes. A 25 basis point delivered cut typically produces a modest immediate move; the sustained move follows when the forward curve prices additional cuts that were not previously discounted.

What is a realistic USD/MXN spread cost from a raw-account broker?

On raw-spread accounts like IC Markets Raw, Pepperstone Razor, FXCM Active Trader, or Tickmill Pro, USD/MXN quoted spreads run roughly 3-8 pips during liquid hours, plus commission. This is a compression from the pre-2001 manual-market norm of 20-40 pip dealt spreads. The compression does not hold during stress: during headline events or thin overnight hours, the raw spread can widen to 40-80 pips in seconds, which is when position sizing that assumed the routine spread breaks down.

Is a covered carry worth the trouble compared to an uncovered position?

A covered carry earns roughly a quarter of the uncovered carry's headline return but eliminates the spot delta. Whether that is worth it depends on the trader's ability to size an uncovered position through a two-standard-deviation reversal on a fat-tailed cross. Most retail-sized accounts cannot. For those accounts, the covered version is not a smaller trade — it is a survivable trade. The uncovered version is a call option on volatility staying inside historical ranges, and MXN volatility does not respect historical ranges reliably.

How does the retail sales print affect Banxico's reaction function?

It shifts the composition of inflation the board is watching. Weak retail sales accelerate goods disinflation, which pulls headline inflation lower even if services inflation stays sticky. That gives the doves on the board a cleaner argument for cuts and moves the forward curve. The market does not wait for the cut to be delivered — it prices probability changes, and those changes show up in the 1-month and 3-month forward points long before any policy meeting.

What is the historical precedent for peso strength during Mexican slowdowns?

The 2019 window is the cleanest recent parallel: Mexican growth decelerated visibly through the year while USD/MXN held in a range and briefly strengthened during periods of high global risk appetite. The pattern breaks when either the global carry environment shifts — through a Fed pivot or a broad risk-off event — or when the differential compresses meaningfully. The precedent is not "strong peso equals strong Mexico." It is "high differential plus stable risk appetite equals strong peso until one of the two breaks."

What sizing framework survives a fat-tailed reversal on this pair?

Size against a four-standard-deviation move rather than a two-standard-deviation move, and stress against the widened spread that appears during that move, not the quoted spread during normal hours. Practically, that means notional exposure roughly half of what a normal-distribution volatility target would suggest, with stops placed outside the historical stress-window spread rather than inside the routine spread. Positions that survive on a chart do not always survive on the fill — the difference is where sizing breaks retail carry books.