The best return of the five, and mostly not the earnings

What the market paid, and what the company built

Over the five fiscal years to 31 March 2026, Marubeni returned more in yen than any of the other four houses. Read quickly, that is a story about a trading company earning its way out of a bad patch. The ten-year record says something more specific.

Equity went from ¥1.7 tn to ¥4.4 tn, a multiple of 2.6×. Book value per share, which is the measure that matters to an owner, went from ¥970 to ¥2,663 — 2.7×. Profit went from ¥155.4 B to ¥543.9 B, 3.5×. The market capitalisation went from ¥1.2 tn to ¥9.2 tn: 7.8×.

The company said this was the plan

What makes the re-rating worth a chapter rather than a footnote is that Marubeni wrote it down as an objective. The GC2027 strategy says it will expand the market capitalisation by sustaining return on equity while “improving PER (by reducing CoE and elevating growth expectations)” — that is, by moving the price-to-earnings multiple, through a lower cost of equity and higher growth expectations. Of the plans this site has read across the five houses, it is the only one that names the multiple as something management intends to move.

Whose profit is it

62.2% of the latest year’s profit — ¥338.3 B of ¥543.9 B — is the group’s share of profit from associates and joint ventures: companies it influences but does not consolidate. That is the highest of the four houses this site holds a balance sheet for, against 58.5% at Mitsubishi, 53.7% at Mitsui and 35.9% at Itochu.

On the Mitsubishi page that figure is where the analysis has to stop: the equity-method book is disclosed in aggregate and 76.2% of it is not attributed to any named company. Marubeni publishes the same thing split across every segment — both the profit the affiliates earn and the capital invested in them. So the question can actually be answered here.

What recurs

Every company has one-time items. Marubeni is the only one of the five that publishes them segment by segment and quarter by quarter, with the transactions named, alongside how much of next year’s forecast is made of the same thing.

The segments, and how long they have existed

Marubeni reports ten operating segments and an “Other” row for head office costs, inter-segment elimination and unallocated financing cash. Unlike Mitsui’s pair of residual rows, Other here is small — ¥30 B of assets — so excluding it from the shares below barely moves a denominator.

The harder problem is time. Two years of this table exist, and that is all.

The segments have been reorganised twice in two years. Effective with the year ended 31 March 2026, Marubeni collapsed 16 operating segments into 10 and restated the prior year to match — which is why this table starts in 2025 rather than running the ten years the financial table below does. Then, effective with the year ending 31 March 2027, it moved part of Power & Infrastructure Services into Energy & Chemicals and part of Next Generation Business Development's neighbour into Lifestyle, and reclassified the year below again. The consolidated total is unchanged on every basis; four segment figures are not.

Ten years of assets, equity and earnings

Return on assets and leverage are both struck on average balances, so they multiply back to return on equity exactly. Both are blank in the first year, which has no opening balance sheet to average against. The affiliate share is blank in the loss year, where a share of a negative number means nothing.

Ten years of segments

This page argues that the re-rating was the plan rather than the earnings. The segment record is where that gets tested, and it is also where Marubeni is least readable of the five: four reporting bases in ten years, two sources, and not one operating line that keeps its name throughout.

Marubeni has re-cut its reporting segments 3 times since 2017, at 2018, 2019 and 2025. Matching segment names across a break like that splices two definitions together and calls the result a trend, so the eras are kept apart here and each figure says which basis it is on.

Two sources sit behind these ten years, and the join is a change of source as well as of basis: Capital IQ segment disclosure, as reported in each year, then Marubeni, Full-year IR Presentation FYE 3/2026.

2019 carries segment assets but no segment profit, because the vendor's segment profit does not sum to the group accounts for this year. The figure is withheld rather than published wrong or balanced to the total.

No operating segment keeps its name across the ten years. The only lines that run throughout are the residual ones — the corporate and elimination rows that exist to make the others add up. An outside reader cannot watch a single Marubeni business compound over the decade, because the company stopped reporting it as one business more than once.

What the balance sheet holds now

The current basis, ranked by the assets each line carries against what it earns on them. 6 of the 10 earn less on their own assets than the group earns on all of its.

Over the ten years total assets grew 52.7% and equity 159%, against profit 250%. Those three are continuous whatever the segments did, which is why the group figures carry the long view on this page and the segment tables carry the detail.

What management has committed to

Before the current plan, the one before it. Marubeni scores 6 financial targets from GC2024 and reports every one met or beaten.

The same business as Itochu, twice

Two of the five houses on this site have the same founder. Chubei Itoh started a linen trade in 1858; the sales divisions were split in 1918; and the half that became Marubeni re-formed in 1921 and took its present corporate form in 1949. Itochu dates itself from the same 1858.

168 years after that linen trade, and 108 years after the split, the two halves earn 14.6% and 13.6% on equity, and the market pays 2.1× and 2.1× for their books. For two companies that have been separate for more than a century, that is a remarkably small gap — and it arrives by opposite routes. Itochu earns its return from businesses it consolidates; 62.2% of Marubeni’s comes from companies it does not.

Conventions

• Fiscal years. Every column is labelled by the date it ends. The year ended 31 March 2026 is FY2026 to the data vendors and FY2025 to Marubeni, as it is to Mitsubishi and Itochu; Mitsui labels the same year FY2026. The plan years are labelled by their own end dates throughout.
• Net profit. Profit for the year attributable to owners of the parent, which is what Marubeni’s segment note reports and what the group line sums to.
• Return on equity. Net profit over average equity, the mean of the opening and closing balances.
• Return on assets. Net profit over average total assets, on the same basis, so that return on equity decomposes exactly into return on assets times leverage.
• Operating profit. Marubeni’s own measure, which it states is presented in accordance with Japanese accounting practice and is not required by IFRS. It is gross trading profit less selling, general and administrative expenses and allowances for doubtful accounts. It is used here only to show what a segment earns before affiliate profit, which is the one job it does well.
• Affiliate profit and affiliate capital. “Share of profit of associates and joint ventures” and “Investments in associates and joint ventures”, both as the segment note reports them. A share above 100% means the segment’s consolidated businesses lost money.
• One-time items. The company’s own figures and its own definition, rounded by it to the nearest ¥1bn. This page does not classify anything as one-time that Marubeni has not.
• Market capitalisation. At each fiscal year end, from Capital IQ. The February 2026 figure is the company’s own statement and is not a year-end measurement.