Most explanations of the last five years reach for inflation, energy prices or interest rates. Annual reports let us ask a narrower question: what happened to the money, and at which line of the income statement it stopped arriving.
We took every company reporting €1–10 million of revenue in 2021 and followed those same companies through 2025. The median one now earns 2.7% of revenue, against 5.7% in 2021. Slightly more than half the profitability of the Latvian mid-market disappeared in four years. Almost none of it disappeared at the gross margin line.
The test
We used annual report open data published by the Latvian Register of Enterprises on data.gov.lv, covering financial years 2021 to 2025. From it we took companies reporting net revenue between €1 million and €10 million in 2021, and followed each of them forward regardless of what its revenue did afterwards. That leaves 4,794 companies present in all five years.
The cohort matters more than it sounds. Consumer prices in Latvia rose about 34% cumulatively over these five years, so a fixed €1–10m band is not a fixed population: companies drift into it as prices rise, and a chart built that way measures inflation as much as performance. Following the same companies removes that. We did not require a company to stay inside the band, because that would have excluded every grower and every shrinker, which is selection on exactly the thing we are measuring.
We removed the financial sector (NACE 64, 65, 66), holding and head-office entities (64.20, 70.10) and real estate (68), which at this size are predominantly asset-holding vehicles rather than trading businesses. We also dropped NACE 45 and 95, where two revisions of the classification are mixed in the source data. Together the cohort carried €13.9 billion of revenue and 162,000 employees in 2021, and €18.4 billion and 159,000 employees in 2025. We report the median and the middle half rather than averages, because a handful of extreme observations makes the average useless here.
One limitation belongs at the front rather than the back. This is a survivors’ view: a company that failed or was wound up between 2021 and 2025 is not in the cohort. Whatever deterioration the numbers show is therefore, if anything, understated.
The median net profit margin fell from 5.7% in 2021 to 4.8%, then 3.1%, then 2.4% in 2024, recovering to 2.7% in 2025. The direction is the same at every point of the distribution, which is the part worth attention.
Figure 1. Net profit margin: the median, and the two boundaries a quarter of companies fall below and rise above. 4,794 companies followed 2021–2025.
In 2021 a quarter of companies earned less than 1.6% of revenue. By 2024 that same quarter earned less than minus 0.4%, and in 2025 less than 0.0%: a quarter of the Latvian mid-market now earns nothing at all on its revenue. At the other end, a quarter of companies earned more than 11.7% in 2021 and more than 7.8% in 2025, so the companies that were doing well are also doing distinctly less well. The middle half of companies, which covered 10.1 points of margin in 2021, covers 7.8 now. In aggregate the cohort’s profit pool went from €1,056 million to €844 million in 2024 and €995 million in 2025, while its revenue grew by a third in nominal terms.
One in four loses money
The share of loss-making companies rose from 11.3% in 2021 to 26.4% in 2024, easing to 24.8% in 2025. It more than doubled. Over the same period the share earning more than 10% of revenue fell from 31.1% to 19.3%.
Figure 2. Share of companies by net profit margin band. The brass segment is loss-makers.
cross the five years, 51.3% of the cohort posted a loss in at least one year and 14.4% did so in three years or more. The combined losses of the loss-making companies grew from €155 million to €283 million. This is not a tail getting thinner while the middle holds; the whole distribution shifted down and the bottom of it crossed into loss.
The profit disappeared below the gross line
Median gross margin was 15.4% in 2021 and 14.8% in 2025, a fall of 0.6 percentage points. Median net margin fell 3.1 points over the same period.
Figure 3. Median gross and net profit margin, companies reporting by function of expense.
Four fifths of the margin that disappeared did so between the gross line and the net line. That is a specific and useful finding. It says these companies were not beaten on price: they passed input cost increases through to customers well enough to hold gross margin roughly flat through the sharpest inflation in three decades. What they did not manage was the cost base below that line, and there was only about five points of net margin available to absorb it.
Where it went
Employee cost is the obvious candidate, and it is also the hardest number to get. Only about one Latvian company in ten files its income statement by nature of expense, which is the only schedule that discloses employee cost separately. Inside the €1–10m band that is roughly 190 companies a year, too few to carry a trend. We therefore widened this exhibit to every company reporting €100,000 of revenue or more, around 1,300 companies a year, and we report it as direction rather than level.
Figure 4. Median employee cost as a share of revenue, by revenue band. Companies filing by nature of expense, revenue €100k and above.
Median employee cost rose from 16.3% of revenue in 2021 to 20.5% in 2025, an increase of 4.2 percentage points. The increase is very unevenly distributed by size. Companies with €250,000 to €1 million of revenue added 5.0 points, and those with €100,000 to €250,000 added 3.6 points. Companies in the €1–10m band added 1.8 points, and those above €10 million added 1.1 points.
The pattern is what you would expect from a labour market that repriced from the bottom up. Minimum wage rises, and rises at the bottom of the wage distribution, hit the companies whose payroll sits closest to that floor, and those are the smaller ones. A company with €500,000 of revenue and eight staff has no room to absorb five points of revenue; a company with €8 million and forty staff has slightly more.
Every sector, without exception
It would be convenient if this were a rotation, with some sectors losing and others gaining. It is not.
Figure 5. Median net profit margin by NACE section, 2021 against 2025. Sections with 100 or more companies in the cohort.
All eight NACE sections with at least 100 companies in the cohort declined. Agriculture, forestry and fishing fell hardest, from 10.4% to 2.5%, with loss-makers rising from 7.6% to 33.6%. Manufacturing went from 6.9% to 2.9%, wholesale and retail trade from 5.7% to 2.4%, transport and storage from 3.9% to 1.5%. Information and communication fell from 7.6% to 5.5% and is now the most profitable section on the list, having started only third. Construction fell least, from 3.9% to 3.4%, but it began near the bottom and stayed there.
In three sections, roughly a third of companies now lose money: agriculture at 33.6%, transport at 33.5%, manufacturing at 26.3%. A level shift across every sector points at something common to all of them rather than at any sector’s own conditions.
The balance sheet response
Faced with this, companies did something that looks prudent and reads badly.
Figure 6. Median equity ratio and the share of companies carrying any long-term debt.
The median equity ratio rose from 52.4% to 59.1%. The share of companies carrying any long-term debt fell from 59.8% to 52.8%, and long-term liabilities as a proportion of total assets fell from 22.4% to 18.7% in aggregate. The current ratio rose from 1.85 to 2.00. On any conventional reading, these are balance sheets getting stronger.
Median return on equity over the same period fell from 27.6% to 9.9% in 2024 and 10.4% in 2025. That is the other side of the same arithmetic: profit stayed on the balance sheet, debt was repaid, and the asset base did not grow to match. Companies became safer and less productive at once. We should be careful about intent here, because deleveraging can equally mean that credit became expensive or unavailable rather than unwanted. What the data supports is the direction, consistently, in every one of the five years.
Higher revenue per employee, lower output
Median revenue per employee rose from €107,240 in 2021 to €115,910 in 2025, a gain of 8.1%. Deflated by Latvian consumer prices it is €86,377, a fall of 19.5%.
Figure 7. Median revenue per employee, nominal and deflated to 2021 prices.
Employment in the cohort fell 1.6%, from 161,993 to 159,349. The median company grew nominal revenue 14.4% over five years against 34.2% cumulative inflation. Measured in real terms, 64.1% of the cohort has less revenue than it had in 2021, and the median real change is minus 14.8%.
Put the three findings together and the mechanism is arithmetic rather than mysterious. Real revenue per company is flat to down. Employee cost has taken roughly four points of revenue. Gross margin is unchanged, so nothing was recovered on price. Everything below the gross line had to absorb the difference, and in 2021 there were only about five and a half points of net margin there to absorb it.
What a finance function should take from this
The first implication is that pricing is not where the problem sits, which is counterintuitive during an inflation. Gross margins held. The exposed variable is a fixed operating cost base carried against a real top line that has not grown, and the companies least able to flex that base are the smallest.
The second is that a sector benchmark is now even less informative than usual. When every section moves down by two to eight points, a company that has held its position relative to its sector has still lost most of its absolute margin. The comparison that matters is against your own 2021, not against your neighbours.
The third concerns 2025. It is the first year in four in which margins improved, from 2.4% to 2.7%, and loss-makers fell from 26.4% to 24.8%. That is a real improvement and it is also one year, on filings that were roughly 95% complete when we extracted them. Late filers tend to be the weaker ones, so if anything the final 2025 figure will be slightly worse than what we show. We would not call it a turn yet.
AG Capital provides fractional CFO services and Financial Planning and Analysis (FP&A) services to small and mid-size companies in the US, UK, EU and globally, including budgeting, profitability analysis, cost analysis, investment projections, and a cash flow planning. The company thrives in offering high-level financial expertise and leadership to businesses on a part-time or project basis.