McLaren Losing Honda: Four Titles, One Technical Silence, and the Lesson of Power-Unit Dependency

The Honda RA122E V12 sounded nothing like the turbo V6 it replaced. It was denser, more even, almost metallic, like a cable being dragged across concrete. I stood at the end of the Monza pit lane in the autumn of 2026, twenty years old, notebook soaked from the previous night's rain, and wrote down a sentence that seemed silly at the time: this sound is going to disappear.

Weeks later, a short statement from Tokyo confirmed it. Honda withdrew from Formula 1 at the end of the 2026 season. Not because it had lost. Honda had won four consecutive constructors' titles with McLaren from 2026 to 2026, and four drivers' titles in the same window. It left because of numbers far from the track: the collapse of the Japanese asset bubble, a recession, and a racing engine programme circled in red on a boardroom agenda.

There are silences in a technical facility that speak louder than any blockbuster contract. The silence in Woking in the winter of 2026 was one of the most expensive in this sport's history.

Three decades later, readers still ask me the same question: what if Honda had stayed? It is the best and the most meaningless question in F1. Best, because it forces us to talk about what was actually lost. Meaningless, because nobody can verify the answer. Between those two positions there is enough room for serious analysis.

The alliance that had no rival

McLaren entered 2026 with a 1.5-litre Honda V6 turbo, the RA168E. That season the team won 15 of 16 races, a ratio never since surpassed in the modern era. That figure sits in the official record, not in anyone's memory.

Less discussed is the engineering path behind those four titles. In 2026, the FIA banned turbocharging and moved to 3.5-litre naturally aspirated engines. Honda answered with the RA109E V10. In 2026 came the RA100E, still a V10. By 2026 Honda had moved to a V12 architecture with the RA121E, followed by the RA122E for 2026.

Three different engine concepts in four seasons, all attached to championships. A power-unit partner that changed architecture repeatedly while holding a title run is a sign of systemic adaptability, not of a frozen advantage.

On the McLaren side, the car was designed around the engine's envelope. Cooling, exhaust routing, oil tank position, rear weight distribution, gearbox layout, even the shape of the gearbox casing all depended on the physical dimensions and thermal behaviour of the unit Honda delivered. None of it was neutral.

That is the whole technical meaning of a works partnership. Not a component supplier, but a party that shapes the architecture of the car from the first drawing, with top resource priority and a two-way data stream between two factories.

What was actually severed

When a works relationship ends, a team does not simply lose horsepower. It loses a year's worth of accumulated knowledge about how a power unit responds to each tyre compound, each torque map, each track temperature, each gear step. That knowledge base does not travel to the next supplier.

McLaren's path after 2026 proves the point coldly. In 2026 the team ran customer Ford engines, without works status. It still won five races that season, a respectable return without technical priority, but the title went to Williams-Renault. In 2026 McLaren switched to Peugeot and finished the season without a single victory. In 2026 the Mercedes relationship began, and it took until 2026 for the team to win again.

Four seasons. That is how long it took an organisation that had just won four straight titles to restore its ability to win races. The figure matters because it turns a vague phrase about misfiring performance into a measurable quantity: four years, three different engine suppliers, two changes of integration philosophy.

The heaviest damage was not aerodynamic. It was predictive. A team that had optimised everything around one specific engine philosophy had to relearn brake load distribution, rear thermal management and differential setup for every circuit. That is integration cost, and it never appears on a timing sheet.

A cycle, not an accident

What makes 2026 analytically useful today is that it was not isolated. Honda returned as a works team from 2026 to 2026, then withdrew again when the global financial crisis hit. It came back once more as a McLaren engine supplier from 2026, a period both sides would rather forget, before moving to Red Bull and winning titles.

Three entries, three exits, across four decades. None of those decisions was driven by on-track results. The 2026 withdrawal came from macroeconomic pressure. The 2026 withdrawal came from a crisis on the other side of the ocean. That is the pattern fans miss: manufacturers join and leave on a boardroom calendar, not a race calendar.

Based on my experience following races since 2026, no technical regulation the FIA writes can keep a manufacturer in the sport when its leadership has decided to cut costs. That was true in 2026, true in 2026, and it will be true in the next decade.

For a team, the consequence is clear. Works status is not a line in a sponsorship contract. It is a first-order competitive variable outside the team's control. An organisation can design the fastest car on the grid, but it cannot design its partner's balance sheet.

The contrarian angle: would Honda have saved the run?

Here I break with most retrospectives. The most seductive version of 2026 to 2026 is a story about an engine. But even if Honda had stayed beyond 2026, extending the title run would have been very difficult, because three other forces arrived at the same time.

First, Williams-Renault was rising, with a Renault V10 and a new aerodynamic philosophy. The gap between the two teams had already narrowed before Honda announced its exit. Second, the technical regulations were about to change sharply in the mid-1990s, with a wave of electronic systems banned and new refuelling rules, factors that reshuffle the order regardless of who supplies the engine. Third, the internal rivalry between two elite drivers inside the team was a source of organisational energy drain no power unit could offset.

In other words, the what-if question is usually framed around the wrong centre of gravity. The right one is this: why did a team that won four consecutive titles have no contingency plan for a partner walking away?

That was the real weakness. Not that Honda left, but that McLaren had built its entire technical architecture, operating process and even staffing plan around a single supplier, while knowing that supplier was a multinational under shareholder pressure.

Strategy is not a mummy, so do not seal it behind museum glass. The 2026 to 2026 legacy deserves to be remembered for four titles, not proposed as a governance model to copy.

My own mistake

In the autumn of 2026 I wrote a piece arguing that Honda's departure marked the end of McLaren as a championship force. The piece was wrong on timing. McLaren still won five races in 2026, took drivers' and constructors' titles in 2026 and 2026, and won the constructors' championship again in 2026.

The sweetest mistakes are the ones that prove you are still listening. I read the event correctly and the scale incorrectly. A supplier leaving does not erase an organisation; it takes a few years from it, and the number of years depends on what the organisation prepared beforehand.

At 54, I have learned that emotion is also a rare form of data. The sense of loss in the Woking technical area that night was real data. But it was data about people, not about end-of-season results, and I mixed the two in the same paragraph.

What to watch from here

Fans do not remember tables of numbers; they remember the breathing of a race. But for those of us who do this for a living, numbers remain the only tool for verifying memory. And the numbers from 2026 to 2026 are a reminder that in this sport, the power-unit supply chain matters as much as aerodynamics.

In the current transfer window, as teams negotiate engine supply for the next regulatory cycle, I suggest reading contracts in three layers. The first is the length of commitment, because a one-year deal and a five-year deal imply very different levels of technical priority. The second is works versus customer status, because the two carry different data access and upgrade schedules. The third is the parent group's financial health, because that is the only variable never written into a press release.

Let me offer a verifiable prediction. Within five years, at least one team in the title-contending group will be forced to change engine suppliers for financial reasons at the parent-group level, not for performance reasons. When that happens, watch how long it takes that team to win again. If it is shorter than four seasons, McLaren of the 1990s was worse prepared than it needed to be. If it is longer than four seasons, the industry has learned nothing from the winter of 2026.

Both outcomes are worth waiting for. That is why I am still here at 54, listening to every new power unit and wondering which sound will be the last before someone in Tokyo, or Wolfsburg, or Detroit decides it costs too much.