Economics and Profitability

Read a vertical farm's economics from the loss rate and selling price by type, not market size

The figures in industry reports are, most of the time, correct. The market is growing; half of it is in the red—both are surely facts. But that is “a story about the market,” not “a story about your plot.” What answers your own economics is the loss rate broken down by type and crop, and the sensitivity by which your break-even scale moves by an order of magnitude depending on the selling price.

The total and the loss rate do not answer whether you turn a profit

“Half of vertical farms are in the red”—have you ever seen this one line and felt your hand freeze over the decision to enter or expand? The market is worth hundreds of millions, even billions, and growing; yet half are said to be in the red. The bigger the number, somehow, the less it answers whether your own operation will turn a profit. Where does that mismatch come from?

You catch yourself noticing that even when you look at a report’s “half in the red” figure, it does not feel like it is about you. PFAL and Greenhouse, leafy greens and fruit-vegetables — the economics of each are an entirely different story, and being shown an average that lumps them all together tells you nothing about which side your own operation is on. On top of that, a small move in the selling price changes the scale you need by an order of magnitude. The bigger the number you look at, the more it starts to feel like something you cannot actually use for your own decision.

That feeling of “it does not feel like it is about me” is the correct one. “Half in the red” is an average that rolls PFAL and Greenhouse, leafy greens and fruit-vegetables, all into one number. But each has a different cost structure, so the average describes an operation that does not really exist, and it tells you nothing about which side you are on. What actually matters is the loss rate within a population broken all the way down to your own type and crop. Look at PFAL leafy greens alone, for instance, and even the same “half” produces a completely different figure. Alongside this, the effect of the selling price is large. A small move in how much a single head of lettuce sells for, and the scale you need to break even changes entirely. So rather than the market-size figure, only once you can translate it into “the loss rate of your own plot” and “sensitivity to the selling price”—these two—does it become material for a decision.

That whether a crop pays off divides by crop is fairly well established in the literature too. What is commercially viable right now in closed vertical farms is a limited range — leafy greens and herbs. (see 1) Staple grains like rice, wheat, and corn, on the other hand—a category that makes up roughly 60% of the world’s food energy—are positioned by one review as not viable for the foreseeable future. (see 2) Behind this lies the weight of electricity costs: in PFAL, power runs roughly 20–40% of production cost, and of that power, lighting alone eats 60–80%, according to reported figures. (see 3) So the more a fruit-vegetable needs more light and heat than leafy greens to swell its fruit, the harder it hits this wall.

Half in the red flips direction by type and crop

So let me add one contrarian angle. That headline of “half in the red” often comes from a population counted mainly around people growing lettuce in PFAL. Even within the same survey, split it by type and the picture changes entirely. In the Ministry of Agriculture, Forestry and Fisheries’ latest field survey (FY2025), only about 50% of PFAL operations are profitable or break even, which means roughly half are on the loss side. The mainstay of that PFAL is leafy greens like lettuce. Yet in the same survey, turn your eyes to the Greenhouse and over 70% are profitable or break even, with the profitable alone exceeding half. The mainstay here is fruit-vegetables like tomatoes (see 5). Count mainly lettuce and half are in the red; on the tomato side, the profitable are many—the same “vertical farm,” yet just re-dividing the population by type and crop turns the picture the opposite way from the headline.

That said, hearing that Greenhouse fruit-vegetables are often profitable, you nod — and at the same time you start to suspect that this too is, after all, just one lumped-together average. Even among the same Greenhouse tomatoes, scale and sales channel surely make it completely different. So rather than “the opposite of the headline,” it is closer to this: the finer you divide the population, the more every layer turns out to have its own circumstances. The question is how far you have to divide before it becomes about you. After type and crop, what is the next cut that matters?

What matters when you divide is, in practical terms, not how much you sell but when and how steadily you sell it—whether it is contract trade across the full year, or one-off spot sales buffeted by the going rate. Even with the same Greenhouse tomato, a layer that has secured a year-round contract with mass retail and a layer whose price swings with the market rate produce profit in entirely different ways, even at the same average selling price. Below that come location (light, heat, climate) and, more than scale, “whether you already have a sales channel that can fill your capacity.” So the order of dividing — roughly, type and crop, then channel stability, then location — is coarse enough to make it about you. Put the other way: divide any finer than this, and in the end all that is left is “the numbers of your own plot.” So rather than keep dividing, it is faster to go and pin down, for your own plot, at what price, how much, and how year-round you can sell.

The view that profit comes out completely differently depending on how you set up your channel holds up in projections too. One study costing out urban agriculture in London found that whether the economics work is strongly governed by whether you can choose high-value crops, whether you can differentiate, and whether you can secure a sales channel — the same facility swings widely in economics depending on how it is set up. The range of economics differs considerably between the startup phase and after the operation is up and running, too. So you cannot leave “which layer you sell to” until later.

What moves break-even scale by an order of magnitude is the selling price

At that point of “pinning down the numbers of your own plot,” the earlier story of “the selling price moving things by an order of magnitude” comes back into play. Say you set an aggressive assumed selling price for lettuce on your own plot and conclude “at this scale we are profitable” — let that selling price slip just 10% lower, and the scale you need to break even can jump by an order of magnitude. So when you set your plot’s numbers too, what you set the selling price at is the most fragile hinge.

This sense of the selling price being “a fragile hinge” shows up clearly as numbers in economic projections. To state the conclusion first: a mere 20% drop in the selling price of lettuce sends the minimum scale at which it pencils out leaping from 38 square meters to 1,700 square meters. These are figures from one model projection (in every case assuming advanced equipment; with average equipment the starting point is 17 square meters. They are projected values premised on a fixed yield and price). From the same baseline, a 35% drop in the selling price and it no longer works without 100 hectares or more. The slightest move in the selling price changes the scale that holds up by an order of magnitude. Conversely, even for a crop like strawberries, whose minimum scale is currently estimated at over 110,000 square meters, one projection has it that a 20% rise in yield drops the minimum scale to 1,200 square meters. (see 4)

The effect of scale itself is, in fact, not that large. Multiply scale by 100 and construction cost falls only about 50% per unit, and what falls is construction cost alone — operating cost does not fall with scale. A slight variation in the selling price or yield moves the break-even scale far more than that. (see 4)

Keep the total as a weather vane and decide economics from your own numbers

If you go all the way down to your own plot like that, can you stop looking at the “market size, X billion yen” total you started with? What is that number good for? The total itself looks like it no longer bears on your own profit. Yet you cannot quite bring yourself to throw it away. That total may be a number for seeing not “whether I can win,” but “whether more people will enter this market going forward.” Where you can sell and what equipment costs are not decided by you alone; they move as entry increases. If so, the total is not your profitability equation itself but a number to keep around for reading the wind direction — which way the selling price and procurement you plug into that equation will swing going forward.

This framing is clean and makes sense. The total and the growth rate do not enter your profitability equation itself, but they show the direction in which the selling price and procurement you plug into that equation will swing going forward. Alongside this, who is entering matters too. Whether the growth in the total is happening in Greenhouse fruit-vegetables or in PFAL leafy greens makes the effect on your own plot completely different. Even the same “it grew,” if the growers entering are chasing the same buyers as you, it works as downward pressure on the selling price as supply rises; if the entrants are in a different layer, the effect is slight. So keep the total as an entry point for grasping the sense of scale and the direction of entry. If you are recovering equipment over many years, it can also be used to see the span of time — whether, over that period, supply and demand in the layer you are selling into keep supporting you. But even the direction you read there, in the end, has to be brought down once more to “at what price, how much, and how year-round you can sell on your own plot,” looking all the way to how the scale jumps when you shift the selling price by 10%, before it becomes a decision.

In order: read the broad trend from the total, decide economics from your own numbers—those two stages.

If you do not yet have your own numbers, borrow from the nearest layer

So far I have proceeded on the premise of “setting down the numbers of your own plot.” But people about to enter often have no buyers and no wholesale account yet, and that crucial number of their own simply is not in hand. Told to divide the population, the layer of their own that the division lands on has not a single line filled in. Where should such a person start?

If you have neither buyers nor an account yet, what you set down first need not be your own numbers but the numbers of the nearest layer. From someone with the same type, the same crop, and the same way of selling, you borrow the known selling price and scale into your own table as a provisional line. Better to have one line, even a borrowed one, than not a single line filled in — then you can run the question of how the scale jumps when you shift the selling price by 10%.

But there is something to watch with borrowed numbers. Cases out in the open tend to be skewed toward the ones that went well, and competitors will not readily tell a new entrant their real selling price or scale. And run that skewed line through the earlier “10% in the selling price moves things by an order of magnitude” sensitivity, and the fragility scales up right along with it. So do not set the selling price at a single point; give it a range on either side, and run it all the way to where the scale jumps.

On top of that, it is best to reverse the order. Normally you build the equipment and then look for somewhere to sell; instead, sound out the buyer and nail down the selling price before you build. Make a trial run on even a single tsubo and confirm, on a small scale, who takes it and at what price. But what a single-tsubo trial run tells you is usually a small spot inquiry, not the contract selling price across the full year. Factoring in that gap too, you gradually turn what was a borrowed line into your own numbers.

So for someone about to enter, the hardest part is probably neither the equipment nor the capital. It is producing, before you build, a single line — provisional is fine — for the numbers of your own plot, which should not yet exist. Once you can set that down, the rest — divide by type and crop, shift the selling price by 10%, and watch how the scale jumps — works as-is on your own plot.