The K-Shaped Quarter

The S&P/BMV IPC closed Friday at 66,383, up 3.2% on the year and down 0.4% on the week, its fifth weekly decline in six. Flat, in other words, and flat is what the index has been all month. Mexico’s second-quarter earnings season, now mostly reported, shows that flatness is an average of two very different outcomes. On one side, companies posted their best numbers in years and raised or reaffirmed their guidance. On the other, they lost customers and cut it. There is no single “Mexican equities” story in this data. There are two, and the line between them is easy to name: whoever could set price, or earned in dollars and hard assets, won; whoever depended on the Mexican consumer making one more trip to the store, lost.

This is the K-shaped quarter. It matters because the index, and most of the passive money that tracks it, treats these companies as one basket. They are not behaving like one, and the divergence is the most useful signal the season produced.

The distortion is not academic. Grupo México, Banorte, and América Móvil are the three heaviest weights in the IPC, and Cemex and Walmex both sit inside the top ten. The index that looked so quiet last week holds both arms of the K, nets them against each other, and prints the average. Anyone who owns Mexico through the benchmark owns the split whether they can see it or not, which is precisely why the single index number is the least informative thing about the quarter.

The side that won: pricing power and dollars

Cemex was the standout. Net sales rose 12% to USD 4,593mm, operating EBITDA rose 24% to USD 1,018mm, a record second quarter, and controlling-interest net income rose 9% to USD 347mm. Read the fine print and the quarter is very good rather than spectacular: on a like-for-like basis sales grew 7% and EBITDA 18%, and the headline EBITDA includes a USD 42mm one-off European claim settlement. Strip that out and underlying EBITDA still grew 19%, with 1.4 points of margin expansion. On the strength of it, management lifted full-year EBITDA-growth guidance to 16% to 17% on a like-for-like basis, from high single digits, crediting Project Cutting Edge, whose savings target it raised to USD 475mm from USD 400mm, and, more tellingly, firm order books and the prices it can charge. One number the headline hides: first-half controlling net income is down 45% to USD 574mm. This is a cyclical selling into strength on both volume and price. It is not yet a cyclical that has repaired its full year.

Grupo México did it with a single input. Controlling-interest net income jumped 90.9% to USD 2,093mm and revenue climbed 34.9%, and it managed that while producing less: copper output at the mining division fell 3.7% in the quarter. The entire swing came from the copper price. COMEX averaged USD 6.16 a pound in the quarter, up 30.5% year on year; on the LME the move was larger still, up 39.8%. The company mined fewer tonnes and earned far more. That is pricing power you do not control but happily bank, and the group is spending it: Southern Copper, its mining subsidiary, priced a USD 1.25bn note in June to fund Tía María in Peru, its own capital program, and general corporate purposes. Tía María was 42% complete at the end of June and is scheduled to begin operations in the second half of 2027.

Banorte showed the financial version of the same idea, though it rewards a closer read than the headline invites. Group net income rose 6% to MXN 15,550mm and group return on equity reached 25.7%, up 209 basis points on the year. The bank’s net interest margin hit 6.9%, above its own 6.4% to 6.8% guidance range. The group’s margin, however, came in at 6.1%, just below the 6.2% to 6.5% guided, and up only 2 basis points year on year. The spread widened where Banorte lends, not across the whole group. The loan book grew 7.8% as reported and 8.6% excluding currency effects, with consumer credit up 10.3%, credit cards up 12.2%, and auto lending up 26.0%. Management reaffirmed a full-year net income target of MXN 62bn to 64bn.

América Móvil rounded out the group, and its tell was cash, not customers. Revenue rose 3.1% as reported in pesos and net income 9.2%, but first-half free cash flow, on the company’s own non-IFRS definition, more than doubled to MXN 43,021mm from MXN 20,054mm a year earlier, and net debt fell by roughly MXN 31bn over the half to 1.31x net debt to EBITDA after leases, from 1.57x. The subscriber line is less flattering than it first looks: 3.5 million postpaid net additions in the quarter were more than offset by 3.7 million prepaid losses as Colombia and Argentina cleaned up their bases, so the wireless base shrank. A business generating far more cash from fewer, better-paying subscribers is still the mirror image of the consumer names’ problem. It is a mix story rather than a growth one.

The common thread is not sector. Cement, copper, banking, and telecom have little in common except this: each sets its own price, or sells into a dollar or commodity market, and none needs the Mexican consumer to add one more item to the cart.

The side that lost: volume and the consumer

Walmex is the clearest case, and we wrote it up on its own. Revenue grew 1.9%, but net income fell 0.7%, transactions fell 1.1%, and the company lowered its 2026 sales-growth guidance to 3.5% to 4.5% at constant currency, citing a slower consumer recovery than it had expected. Even the World Cup, which it had hoped would lift the broad basket, moved only a handful of categories.

Gruma told the same story from the food side, and in dollars. Consolidated sales rose 3% while EBITDA fell 5%, and its US business, which generates just under 60% of group EBITDA, saw volume down 3% and EBITDA down 15%, with foodservice volumes off double digits. Management named price sensitivity, weak consumer sentiment, and more restraint from Hispanic shoppers at retail, and it cut its US outlook to a low-single-digit volume and revenue decline for the year, taking roughly 200 basis points off its margin guidance. We previewed exactly this setup last week; the print confirmed it.

Travel showed the pattern under a flattering headline. ASUR’s revenue rose 9.9%, but that number is construction accounting plus a new US concessions segment; strip construction out and revenue was roughly flat at minus 0.3%. The operating reality was worse. EBITDA fell 8.7% and the adjusted margin compressed 560 basis points to 62.0% from 67.6%, as total passenger traffic fell 2.7% and Mexican traffic dropped 5.0%. The one line that looks like pricing power, commercial revenue per passenger up 12.6%, is an artifact: ASUR US Airports, a retail and food concessionaire at LAX, O’Hare and JFK, adds revenue to the numerator with no passengers underneath it. In Mexico, commercial revenue per passenger fell 8.1% to MXN 145.7. ASUR did not price its way around fewer travelers. It lost the price too.

The dividing line is pricing power

This is not a story about good companies and bad ones. Walmex is an excellent operator that gained share in a weak market. ASUR still runs concessions at margins most industrial businesses would envy, even after this quarter’s compression. The split is structural, not managerial. When final demand softens, businesses that set price hold their profit, and businesses that depend on volume do not.

ASUR is the purest illustration of the mechanism, and the harshest. Same company, same quarter: fewer passengers, lower revenue per passenger in Mexico, lower profit. The one thing it could not manufacture was the passenger, and once the passenger stopped coming, the price went too. Walmex’s ticket rose while its traffic fell. Gruma held its US pricing while its volume slipped. Pricing power buys time when demand weakens. It does not replace demand, and at the volume end of the market it does not hold indefinitely either.

One nuance complicates the clean split, and it is worth flagging. Banorte’s consumer loan book grew 10.3% and its credit cards 12.2%, which means the Mexican consumer is still borrowing even as they shop more carefully. That is not necessarily healthy. A household that trades down at Bodega Aurrerá while adding to its card balance is stretching, not thriving, and it is a tension worth watching as the year goes on.

Guidance is the tell

Management teams voted with their outlooks, and the vote was lopsided in exactly the direction the results imply. Of the four names with pricing power, one raised its bar, two reaffirmed theirs, and the fourth publishes no guidance at all and simply kept generating cash. Both of the companies exposed to the Mexican consumer lowered theirs.

That divergence is more informative than any single quarter. A raised guide is a management team telling you the strength is durable. A cut guide, especially from an operator as disciplined as Walmex, is a team telling you the weakness is not yet behind them.

The economy underneath

This corporate split sits on an economy that is about to look considerably better than it feels. Mexico contracted 0.6% quarter on quarter in the first quarter on INEGI’s revised reading, after a flash estimate of minus 0.8%, and consensus for Thursday’s second-quarter flash is a rebound of 1.3%, which would be the strongest quarter since early 2022 and is expected to be led by industry. That is not a shallow bounce. It is also not the shopper. A GDP print driven by manufacturing and exports does nothing for transactions at Bodega Aurrerá. The peso sits near 17.48 per dollar and the IPC near 66,400, both steady, both averaging over the same divide the earnings expose.

One point cuts against the easy reading, and not in the direction we would have guessed. The strong peso did not fight the winners. For two of them it helped. Cemex and Grupo México both report in US dollars, so a firmer peso inflates the dollar value of their Mexican operations rather than shrinking it, which is exactly why Cemex’s reported 12% sales growth becomes 7% like-for-like and its 24% EBITDA growth becomes 18% once currency and scope come out. América Móvil, which reports in pesos, is the one the currency actually hurt: revenue rose 3.1% as reported against 5.1% growth in service revenue at constant exchange rates. The winners’ quarter was real. It was also flattered at the margin, and the organic numbers underneath are good rather than extraordinary.

For Banxico, the split is a genuine complication. The winners’ strength, copper, cement pricing, bank margins, does nothing to loosen the labor market or revive the shopper, and an industry-led GDP rebound argues against cutting by keeping headline activity respectable. The weakness that would justify a cut sits in exactly the parts of the economy that never show up in a strong Grupo México print. As we argued in “Trapped at 6.50%,” the central bank is already boxed; it held unanimously on June 25 and meets again on August 6. A K-shaped economy is harder to set one policy rate for than a uniformly weak one. Thursday’s flash will be read through that lens: a print near 1.3% is strong enough to keep Banxico on hold and hollow enough to keep the consumer names discounted.

What would break the pattern

Two things could close the gap, and both are worth watching rather than betting on yet. The first is the peso. The median of the 35 institutions in Citi’s Expectativas survey sees it at 17.90 by year-end, and Cemex is underwriting 18.25 to 18.50 in its own second-half guidance. Note that the survey has been drifting stronger rather than weaker, so this is a consensus that keeps not happening. Note also that a softer peso would lift Grupo México’s margins, since it sells in dollars and pays much of its cost base in pesos, while adding imported-cost pressure at the consumer end. It widens the K rather than closing it. The second is the consumer itself. Gruma told analysts it expects the rate of US foodservice volume decline to stabilize by the third quarter, which buys an easier comparison rather than a recovery. If the Mexican shopper follows, the volume-dependent names carry the most operating leverage to a turn, precisely because they gave up the most on the way down. Neither has happened yet. The guidance cuts are management telling you they are not counting on it this quarter.

How to position

The practical read is to stop trading “Mexico” as one thing. The index multiple prices an average almost no company is actually living. Through a still-soft consumer, the pricing-power and dollar-and-commodity side, cement, copper, banks that can defend a margin, and cash-generative telecom, is where earnings are compounding and where guidance is still moving up. The consumer and volume side, retail, packaged food, and tourism-exposed travel, is cheaper for a reason, and the reason has not turned yet.

We are not calling the consumer names value traps. Gruma trades near 6x trailing EV/EBITDA and Walmex keeps taking share, and both will look cheap the day traffic inflects. We are saying the inflection has not happened, and the guidance cuts say management does not think it has either. We would want the turn to show up in transactions and passengers, not in tickets and revenue per passenger, before treating the discount as the opportunity.

The signal to watch is close at hand. Thursday’s GDP print sets the tone into Banxico’s August 6 decision, and the Q3 guidance from the consumer names will tell you whether July marked the bottom or just the latest step down. Until the shopper comes back through the doors, own the companies that set their own price.

One market, two economies. The IPC will keep printing a single number every afternoon. This quarter said there are two Mexicos inside it, and only one of them is having a good year.