Before planting, write down the margin this crop should leave, and next to it the yield and the price at which it would only break even. The margin is the revenue you expect, less the variable cost of the whole area, less the share of fixed cost charged to this cycle (one crop season, or one batch of animals from purchase to sale). Take 120 hectares yielding 3.6 tonnes each, sold at 1,200 a tonne, with 3,000 of variable cost per hectare and 72,000 of fixed cost for the cycle, in whatever currency you keep your books. The margin comes to 86,400, or 720 a hectare.
The break-even points, called floors here, sit at 3.0 tonnes of yield or at 1,000 of price. Either one can fall 16.7 percent before the cycle stops covering its costs, and that distance is the room. The free crop budget calculator on this site runs these same numbers, and the number worth writing down first is not the 86,400. It is the 16.7.
Most planting decisions get made on a round number one person carries in their head. Cost then arrives in pieces over eight months, yield shows up at harvest and price at the sale, and the result is a surprise nobody can trace. Whether it was price, yield or cost cannot be answered when none of the three was written down first, and a surprise nobody can trace tends to come back the next cycle.
The calculator takes those five numbers plus three ranges, one each for price, yield and cost, written as how far each can move up or down in percent. It returns the margin, the two floors and the room; a ranking of how much 1 percent of each variable moves the margin; and the whole cycle recalculated in three columns: optimistic, base case and pessimistic.
Why is 1 percent of price worth exactly 1 percent of yield?
Because revenue is area times yield times price, so raising either one by 1 percent raises revenue by the same amount, and the fixed cost does not move. In the example, 1 percent of either is 5,184, one hundredth of the 518,400 of revenue. One percent of variable cost is 3,600, one hundredth of the 360,000 the variable cost adds up to across the whole area, and in the calculator it carries a minus sign because it comes out of the margin.
In percentage terms, the margin swings more than any of the three. The 5,184 that 1 percent of price is worth is 6 percent of the 86,400 margin, and that 6 is no accident: revenue is six times the margin, so every 1 percent of price moves the margin six times as much. Thin margins do not just leave less room, they amplify every swing, and a budget written as a single number hides that.
The six and the 16.7 percent are the same fact seen from two sides. Revenue is six times the margin, so the margin is one sixth of revenue, and one sixth is 16.7 percent. Once all the costs are counted, the room down to the floor always equals the margin as a share of revenue.
A farmer who knows the farm keeps 16.7 percent of revenue already knows the price can fall 16.7 percent before it breaks even, and that each 1 percent of price takes 6 percent of the margin, without opening any calculator. On a farm keeping 5 percent of revenue, the same arithmetic gives 5 percent of room and 20 percent of the margin lost for each 1 percent of price, which is the difference between a bad year being a fright and being the last one.
So which variable actually moves your margin most?
On a crop budget, the calculator’s ranking will almost always put price and yield above variable cost, and the reason is arithmetic, not a finding about your farm. Per percentage point, price and yield beat variable cost whenever revenue is larger than the variable cost bill, which is the same as saying the gross margin is positive. The ranking can only come out differently on a budget that is already losing money before the fixed costs are paid.
What the arithmetic does not tell you is how far each one moves in a real year. Rosa and colleagues, agricultural economists who ran fourteen years of Italian yields and prices for seven cereal and oilseed crops, found price more variable than yield in all seven, and concluded that the risk came from “a mix of market (price fluctuation) and production (yield variation), with the higher risk generated by the market”.
Bruce Sherrick, a farm finance economist at the University of Illinois, reached the same place with thirty-seven years of corn yields and prices for every county in the state: “price risk represents about 66% of the cause of revenue variability”. A percent of price and a percent of yield are worth the same money, but price usually moves by more percent in a year.
Split the cost line into its items, though, and the ranking can turn over. Ferreira and colleagues, animal scientists writing in the Arquivo Brasileiro de Medicina Veterinária e Zootecnia, tested the gross margin of 149 feedlot cattle from nine genetic groups by moving one price at a time by ten percent, and found the gross margin more sensitive to the purchase price of the feeder cattle than to the selling price of the beef. Scoring the strongest effect as 100, the beef selling price came second at 38.6, soybean price at 7.4 and corn at 3.0.
That result cannot happen in the crop budget calculator, because the feedlot budget had its cost split into items: feeder cattle, corn, soybean and silage were four numbers tested one by one. The authors do not put it this way, but their ranking means one cost item alone weighed more than anything on the revenue side. A crop budget with one variable cost per hectare cannot find the same about fertilizer, because fertilizer sits inside a total that is never moved on its own. The ranking covers the three numbers you entered, not the items folded inside them.
At what yield and price does the cycle break even?
The calculator gives two floors. The yield floor is the yield at which revenue just covers all costs at the expected price; the price floor is the price that covers all costs at the expected yield. Both leave the same room. Revenue is 4,320 a hectare (3.6 tonnes at 1,200), and cost is 3,600 a hectare: the 3,000 of variable cost plus 600 of fixed cost, which is the 72,000 spread over 120 hectares. 3,600 is 83.3 percent of 4,320, so yield or price can fall 16.7 percent before revenue meets cost.
Underneath sits a second choice: which costs the floor has to cover. The Food and Agriculture Organization of the United Nations (FAO), in its manual on farm production cost statistics, lists a break-even price that covers only variable costs, the gross margin version, next to one that covers total costs, the net margin version, and gives the break-even yield as total costs divided by the projected price.
Both versions are legitimate, and they answer different questions. That is why the fixed cost is optional in the calculator: leave it blank and you get no floor at all; type 0 and you get, on purpose, the floor that covers variable cost only.
Embrapa, Brazil’s federal agricultural research corporation, published the same room under another name, on a real crop. In an onion budget for the Submédio São Francisco region, a total cost of 9,018.70 reais a hectare at 0.68 a kilo puts the break-even at 13,262 kilos a hectare against an expected 20,000. The study calls the gap a margin of safety: the quantity produced or the selling price can fall by as much as 34 percent before the crop starts to record a loss.
The yield that goes into the floor is not the yield off the field. Under the FAO manual’s rule, the production figure should reflect only marketable production, leaving out waste, losses and own consumption. A figure copied from the yield monitor makes the price floor look lower and the room look bigger than they are, by whatever the farm loses between the field and the buyer’s scale, which is a number worth knowing on its own.
Why does the pessimistic column move all three at once?
Because whoever built the tool chose to, and the choice has a published reason. The University of Florida’s Institute of Food and Agricultural Sciences (UF/IFAS) gives it in its extension guide to enterprise budgets: “A primary concern is that sensitivity analysis often evaluates the impact of changing one variable at a time while holding all others constant”, while “In reality, changes in economic conditions often cause two or more input variables to fluctuate together”, so “analyzing them in isolation can underestimate the true financial risk”.
With a range of 15 percent up and down on each of price, yield and cost, all moving at once, the example does not just earn less, it loses money. The optimistic column comes out at 307,584 and the pessimistic one at a loss of 111,456, against a base of 86,400. A spread of 419,040 around an expected margin of 86,400 is what the single-number budget was hiding.
In one respect the column is less realistic, and that has been measured. Sherrick found that in the corn belt “prices and yields tend to move in opposite directions”, closely enough that “51% of an increase in one variable is offset by a decrease in the other”. A column that puts low price and low yield in the same cycle combines two things that tend not to come together across a region.
On a single farm they come together more often. Sherrick also found that the offset “tends to decline” as you move from large regions down to one farm, because it works through regional supply and one farm’s bad year is not regional. Hail on one property, a flood on one road, a crop lost in one valley while the rest of the country harvests normally: in those years low yield comes with no price rise, and those are the years that can end a farm. The pessimistic column is not a forecast. It is the case the natural offset does not cover, which is the case worth budgeting for.
Where do the three ranges come from?
From your records, because ranges set by feel come out too narrow. Cruz Júnior and colleagues, agricultural economists, surveyed 90 corn growers in southern and central-western Brazil in late 2008, asked for the price range they thought possible and compared it with what the market had done. They concluded that farmers perceive less risk than the market actually carries. A range typed from memory is likely to be too tight, and too tight in the direction that makes the budget look better.
Each range has a record behind it that the farm already has, and the three are not the same width. The price range comes from the highest and lowest quote for the same month across recent years in your own price record, a sheet of local prices with the date and source of each (how to set one up is in keeping a price record with a source).
The yield range comes from your last cycles in that field, the same history you would use for planning the next three production cycles. The cost range comes from the spread between the quotes in your hand now. Three equal ranges are almost certainly wrong, and they are what comes out when all three are typed in one sitting.
Rosa found price more variable than yield in all seven crops, and Sherrick puts two thirds of revenue variation on price. So the price range should be the widest, the yield range next and the cost range the narrowest, because cost is mostly committed early and moves least once bought. Typing 15 in all three squeezes the price range to fit the other two, and the pessimistic column looks better than the farm’s real risk.
A cost range can be measured too. Embrapa’s onion study ran 5,000 simulated cost sheets built from real cost items, and the break-even moved from 12,335 kilos a hectare at the low end to 14,160 at the high end, with the margin of safety falling from 38 percent to 29.
What does a cycle budget leave out?
UF/IFAS names three limits. Enterprise budgets “assume fixed input quantities and yields and do not account for risk or uncertainty”, they “typically cover a short time horizon, usually just one year”, and they “generally exclude financing and cash flow considerations, which are critical to a farm’s overall financial health”. The third is the expensive one: a cycle can show 86,400 of margin and still leave the account empty in the month the fertilizer invoice falls due. Catching that takes a budget that runs several years with a cash line for each, like a three-year farm budget built on a written price assumption, in which this cycle is one of the years.
The one-year horizon also leaves out machinery, buildings and establishing a perennial crop, which earn over several cycles and cannot be judged by one. Those need a different calculation over their working life, covered in the article on whether an investment pays for itself.
What comes in here is only the share of fixed cost this cycle carries. That share should come from a rule for splitting shared costs between the farm’s activities, written and dated before the numbers are known, rather than a division improvised in the moment; how to do it is in writing the cost-split rule before the numbers. The largest shared cost is usually land, and what it charges the cycle depends on whether you rent the land or buy it, a decision taken before this budget that shows up here only as a figure.
A published budget you copy from is built on someone else’s farm. UF/IFAS says “Posted enterprise budgets are usually based on average or representative farm values” and “do not represent individual operations”, and warns of a second assumption built into many of them: that farmers run at “ideal efficiency levels, which is often not the case”.
One limit is in the calculator itself. It treats all variable cost as a cost per hectare. On a real farm, harvest, freight and marketing are often paid per tonne and fall with a short crop, so the yield floor the calculator shows comes out a little high.
A published table shows by how much. Argentina’s National Institute of Agricultural Technology (INTA) publishes gross margins for the 2024/2025 season with a break-even yield for each crop, which INTA calls the indifference yield, and notes that these figures leave out overhead and indirect depreciation. For soybean it lists 325 dollars a tonne, 2.2 tonnes a hectare, 479 dollars a hectare of direct costs and a break-even yield of 1.3 tonnes.
Dividing 479 by 325 gives 1.47, not 1.3. INTA’s authors do not explain the gap; Rurivia, which publishes this site, redid the sum from their table. Marketing costs 123.6 a hectare, or 56.18 a tonne over 2.2 tonnes. Because it is paid by the tonne, it works as a cut in the price the farm receives, not as a cost per hectare. Take the 123.6 off the cost per hectare and 355.4 is left; take the 56.18 off the price and 268.82 is left. 355.4 divided by 268.82 gives 1.32, close to the published figure. The same method matches INTA’s wheat, sorghum and corn. For soybean, dividing total direct cost by price overstates the break-even yield by about 13 percent.
Rurivia calculation from the soybean cost, price and marketing figures in INTA’s 2024/2025 gross margin bulletin.
How large the gap is on your farm depends on one thing: what share of your variable cost is paid per tonne. In INTA’s table, marketing is about a quarter of direct cost, which is why the gap is large.
In the worked example here, if harvest and freight are a fifth of the 3,000, the yield floor drops from 3.0 to 2.9 tonnes and the room rises from 16.7 percent to 19.4. That is almost three points of room that did not appear on the spreadsheet. Here it does not change what to do, since the cycle stays in the 10 to 20 percent band covered in the next section; near 10 or 20 percent, it can.
So splitting the cost line is not only about finding the item worth testing on its own. It also lets you see how much of the cost is paid per tonne, which is the difference between a floor you worked out and one the spreadsheet’s layout handed you.
What should you do at each level of room?
Room turns into a decision only when you compare it with the price range you entered, and the calculator gives you both without dividing one by the other. In the example the room is 16.7 percent and the price range is 15 percent. 16.7 divided by 15 is 1.1: the room is barely larger than one ordinary price swing from your own record. A fuller version of this test, with price and yield falling together, gives a smaller answer; it is in the article on how bad a year a cycle can take before it stops paying.
In money: a 15 percent fall in price alone, with yield and cost at base, drops the margin from 86,400 to 8,640, because each 1 percent of price is worth 6 percent of the margin and fifteen are worth ninety. The cycle ends in the black having lost nine tenths of what it was meant to leave. Below 1, an ordinary bad year from your own record reaches the floor; between 1 and 2, it takes a year worse than usual; above 2, only a year outside your record reaches it.
With room above 20 percent, the question is what to protect, not whether to plant. An ordinary bad year does not reach the floor, so the work is keeping the margin: the price floor is the number to take into the sale, the same one you get by working out your lowest acceptable price from the selling side. Write it on the selling plan before the first buyer calls, because after the call it tends to become whatever was offered.
With room between 10 and 20 percent, where the worked example sits, the first task is measuring the ranges, not deciding. An ordinary bad year gets close to the floor without breaking it, so the result depends on how good the three ranges are. Rebuild them from your price record, your own fields and the quotes in hand, and read the pessimistic column again: if it stays positive, the room is real and the work is protecting the margin; if it turns negative, the next step is splitting the cost line into its items.
With room under 10 percent, one ordinary bad year takes the whole margin, and a better guess at the numbers will not change that. What helps is two documents. The first is a target margin: the margin the farm decided this crop has to leave, written and dated before the budget, so the budget’s expected margin has something to be checked against (how to set one is in writing down a target margin). The second is the cost line split into items. The calculator’s ranking will point at price and yield on any crop budget; only the split shows whether one cost item matters enough to test alone. A single 3,000 per hectare is a total, not a budget.
If the base column is negative, that is the case you expect, not the pessimistic one, and what matters is which floor was broken. Type 0 in the fixed cost and the floor covers variable cost only: in the example it falls from 3.0 to 2.5 tonnes of yield and from 1,000 to 833.33 of price.
If the expected price is still above that lower floor, the crop pays its own costs and contributes something toward the fixed costs, and leaving the field empty would not save the 72,000, which falls due either way. What is losing money then is the farm’s fixed cost base, and that is what has to shrink, not this year’s crop. If the expected price is below the lower floor, every hectare planted adds to the loss. With the base column in the red, the difference between the two floors costs real money.
Every result needs a name next to it. One person puts their name to the expected yield, and it should be whoever will answer for it at harvest, not whoever is most optimistic in the meeting. Someone signs the price. Someone answers for the fixed cost figure, which comes from the written split rule, not from a division made on the spot. When the same person signs all three, the budget is one opinion in three columns.
Where to start
Two hours, with last cycle’s invoices, your price record and your field notes on the table, before the seed order rather than after.
The empty column is what pays for the next budget. Filled in at the end of the cycle, it turns this year’s three guesses into three measurements, and after three cycles the farm reads its ranges off its own sheets instead of arguing about them.
A budget that landed exactly on its numbers teaches little, because it cannot show which of the three it got right by knowing and which by luck. From one that missed by 14 percent on yield and 3 percent on price, you can see which way your own estimates lean. Three cycles of that and the farm does its planning on its own numbers instead of somebody’s average. The same habit of writing the number first and checking it against what happened runs through the rest of farm management.