Per unit and per month, and most of the figure is already sitting in bills the farm pays. The cost of storing is not one number. It is a stack of lines, some of them invoiced by a warehouse and some of them paid without any invoice ever being issued, and the only version of it that settles a decision is the one worked out for one named lot over a named number of months.
The lot sits because the offer did not please anybody, and the decision to wait gets taken as though waiting were free. Money that did not come in is paying interest somewhere. Weight leaves. The grade at the door on the way out is not the grade that went in.
Power and hours go into keeping the thing where it is. None of those lines is added up anywhere, so the comparison that would settle the question, what the market pays to take delivery later against what it costs to get there, never gets made. When the lot finally moves, nobody can say whether the wait paid, because the cost of it was never worked out and the record of what went in and what came out is missing or half written.
One more month is the question in the title, and it is not the whole stretch divided by the number of months. Some lines are charged once when the lot goes in and once when it comes out. Others run with the calendar, and the storage costs increase the longer grain is stored. Divide everything by the months and the figure falls as the wait gets longer, which is arithmetic telling a farm that patience gets cheaper the more of it the farm buys.
Why the return quoted for waiting is measured before the cost of waiting
Gross, and the author who published the figure says so in the same article.
Janzen, working from roughly 16,000 farm-year observations of Illinois grain farms between 2003 and 2020, measures the gross return to post-harvest marketing as the percentage difference between deferred and near-to-harvest sales, and reports that farms realize gross returns that are on average roughly the seasonal price rise after harvest while the range of possible marketing gains is wide, and farms often realize negative gross returns. Then he takes the floor out from under his own result: this analysis ignores storage costs inherent in post-harvest grain marketing which are significant relative to observed returns, so the net returns from post-harvest grain marketing must be lower than the gross returns.
Those farms keep their books with a state farm business association by choice, which describes farms that keep books and not every farm in the state, and the spread is not the same every year: in other years the range of the distribution is quite wide. What survives all of that is the shape of the thing. The average anybody quotes at you for holding production is the number before the subtraction, the subtraction is the subject of this page, and the same records show plenty of farms ending up on the wrong side of it.
The line you can work out today, with a price and a rate
Interest on money that did not come in, and it takes one multiplication. Gardner, writing for the University of Kentucky Department of Agricultural Economics in 2023, gives it plainly: the operating loan interest cost per unit is calculated by multiplying the harvest price by the interest rate and dividing the number of months the crop is stored by 12. His worked example holds maize for five months at a harvest price of 5.00 US dollars a bushel, which is the American unit of 56 pounds, about 25.4 kilos. At a four percent rate the line comes to 0.08 US dollars a bushel. At ten percent it comes to 0.21.
Read those two figures as shares of the value sitting there and they travel to any currency: about 1.6 percent and about 4.2 percent of the harvest value, for the same five months, which is our own division of his published figures. Two things move that line and neither of them is the warehouse. It grows with every month the lot stays where it is, and costs increase as the price of the commodity increases.
A farm that pays cash instead of borrowing does not escape the line, it changes what the line is worth, and Gardner names that too when he notes the effect could be mitigated/stopped by paying your operating loan off early or using cash reserves instead of loans.
Every line here is already priced on somebody’s published tariff
Published, itemized, and charged separately, which is the proof that these are real lines and not an accounting exercise. In Brazil the storage tariff of the Conab warehouse network, approved by Resolution 1.609 of 11 April 2023 and in force from 15 June 2023, prices each of them on its own, and its first heading already names the billing unit: storage and/or space reservation, by indivisible civil fortnight.
A reader outside Brazil looks for the same document under a local name. In the United States the Department of Agriculture points at it by name when it says the Commodity Credit Corporation will accept the handling and storage rates stated on the Warehouse Operator’s Public Tariff or Statement of Charges assessed to their commercial customers, and fixes a fallback where none is posted, capping handling at a maximum of 12 cents per bushel for both receiving and loadout handling charges.
The Brazilian table, read as a list of lines rather than as a price list, is the model of the sheet. Values below are in Brazilian reais and belong to that jurisdiction and that date. What copies is the left column.
| Line on the published table | How it is charged |
|---|---|
| Storage or space reservation, bulk, other agricultural products | 3.31 per tonne, each fortnight |
| Storage or space reservation, bulk, rice, barley and malt | 4.30 per tonne, each fortnight |
| Insurance, on every product except those charged the surcharge | 0.02432 percent, each fortnight |
| Surcharge, on rice, maize, beans, sorghum, soybeans, wheat, barley, rye and triticale | 0.15 percent, each fortnight |
| Intake, bulk | 2.69 per tonne, on the operation |
| Dispatch, bulk | 3.36 per tonne, on the operation |
| Drying, up to 16 percent moisture | 11.76 per tonne on natural gas, 16.45 on other fuels |
| Cleaning or pre-cleaning, up to 5 percent foreign matter | 3.59 per tonne, on the operation |
| Minimum charge for storage | 10.00 per fortnight, whatever the tonnage |
Which lines run with the calendar, and which stop at the door
Five of those nine lines have a fortnight in them and four do not, and that split is the whole answer to the question in the title. Intake, dispatch, drying and cleaning are charged once on the operation. Storage, insurance, the surcharge and the minimum charge are charged again every fortnight the lot is still there. One more month of waiting buys two more fortnights of the second group and nothing at all from the first, so the cost of one more month is built only out of the lines that keep running.
On the Brazilian table that is 3.31 per tonne per fortnight for bulk of other agricultural products, which the table’s own unit makes 6.62 per tonne per month, plus 0.15 percent per fortnight of surcharge on the listed grains, which is 0.30 percent per month. Both doublings are ours, and doubling the fortnight is the only thing done to the published table.
The minimum charge is the line that refuses to behave. The table sets it as ten reais per fortnight for the storage charge whatever the tonnage, so it lands on a small lot at a rate per tonne that no rate table shows. This is the reason the number is worked out for the lot that is actually sitting there rather than read off an average. Where there is no room to be wrong about a few units, the arithmetic has to be tighter, not looser.
The loss that is easy to price, and the one that is not
Two losses, and they behave differently in the account. Kumar and Kalita, reviewing postharvest storage for the journal Foods in 2017, separate them: storage losses can be classified in two categories: direct losses, due to physical loss of commodities; and indirect losses, due to loss in quality and nutrition, and the second kind ends in money as well, since the loss in quality results in value loss of the product, and sometimes leads to total rejection also. The first is a weight, and a weighbridge ticket at each end settles it. The second is a grade, and nothing settles it except what the buyer paid.
The gap between the two is measured, and it is the reason the record has a grade column and not just a quantity column. Santos and colleagues, computing an economic damage threshold for stored wheat for the Revista Brasileira de Engenharia Agrícola e Ambiental in 2002, put the cost of control and the commercial value of the product on either side of the same equation, and had to fix that commercial value twice over, once for imported wheat and once for domestic, because identical physical damage on two wheats worth different money is not the same loss.
Their own caution is the part that copies: quantifying the quality loss and attaching a monetary value to it is harder than the calculations for the weight loss. Keeping product in condition is not this page’s business and never will be, and it belongs to the reader and to his agronomist.
Why ten percent is a country’s figure and never your lot’s
National estimates describe a country, and no buyer ever paid anybody against one. Lorini, writing for ESALQ at the University of São Paulo in 2015, reports that in Brazil the average grain losses estimated by the Ministry of Agriculture and by the Food and Agriculture Organization reach around 10 percent of everything produced in a year, and in the same passage sets out the loss that is not weight at all: quality losses that compromise the use of all the grain produced or reclassify it to another use of lower added value.
Ten percent is an estimate by two institutions about a whole country in a whole year, which is why it cannot be the loss line on any sheet. Kumar and Kalita reach the same wall from the research side, reporting that availability of consistent and reliable postharvest loss data is still a challenge. The number that goes on the line is the weight that went in minus the weight that came out, on the farm’s own tickets, for the lot in question. Everything else is somebody else’s average being asked to do work it cannot do.
The record that makes the number checkable at both ends
Date, quantity, moisture and grade, twice, and the second time is the one that gets skipped. The reason for the second reading is that the commercial state of a lot moves while it sits still. Lorini reports the sharpest version of that in Brazil, where for wheat the product is disqualified for sale if a single live insect is found in the lot. A lot that entered saleable and left disqualified changed while nobody was writing anything down.
The grade is not an opinion in any of these markets, which is what makes it recordable. In Argentina, Resolución 1075/1994 approved the quality standards for grain and declared that the standards approved above are of a mandatory character, unless the parties expressly agree different specifications, and it prices the grade directly: on malting barley, delivery at Grade One earns a bonus of one percent on the official quotation, with a deduction of 1.50 percent if Grade Three is delivered.
Moisture above the base the norm sets is priced the same way, through a percentage weight deduction under the table in force, with the agreed or fixed drying tariff to be paid. Two jurisdictions, two documents, the same three columns. Writing them down at both ends is the check step of farm management, and the only reason it feels optional is that nobody is billed for skipping it.
The list of what you left out belongs on the same page
Underneath the figure, named one by one, because a cost sheet with no declared limit gets read as complete. Gardner does this in his own publication, footnoting every chart with the same warning: costs do not account for physical storage costs such as bin depreciation, transport, drying, etc. His number is one line of the stack and he says which one, which is why anybody can use it without being misled by it.
Own storage is where the missing list gets longest, because it invoices nothing and costs the same. The structure was paid for once, it wears out, it takes power and hours, and none of that reaches a ledger by itself the way a warehouse note does. Deciding whether that structure should exist at all is a different page, and a written investment plan is where the condition for building it gets written as a number somebody can check. What belongs here is narrower and blunter: if depreciation, power, labor or freight did not enter your figure, write their names under it, so that the next person to read the sheet knows what the figure is missing rather than guessing.
What the number changes, once it exists
Three things, and none of them is a decision to store or to sell. All three live in the marketing axis, and all three go wrong in the same way when the cost of storing is missing from them. The figure per unit per month raises the price the farm can accept, because every month held adds lines to it, and the break-even price for a lot held six months is not the one for a lot sold at harvest. It gives the comparison something to compare against.
What the market pays to take delivery later is the distance between two quotes read the same way on the same day, which is the discipline a price record with a source imposes with a header naming one source, one weekday and one hour. That distance either covers the lines that keep running or it does not. And it enters the year’s reconciliation as a line of its own, because the sales against the plan with the cost of storing left out produce a weighted average price that flatters every lot that waited.
Holding is a position, and it is the one position most farms take without writing anything. The channel and the trigger for each tranche belong on a written selling plan before the first sale, and holding a lot past the date the plan named is a decision that deserves the same treatment: a date, a figure, and the name of whoever is allowed to reopen it. Janzen notes why the record has to be the farm’s own and not the screen, since farms can forward sell grain for delivery at almost any future date, the price received is not necessarily the cash price on the day delivery occurs. The cash price on a screen is not what anybody got.
Where to start
Two to three hours the first time, with the power bill, the loan statement and the weighbridge tickets on the table. Nothing here needs a warehouse, a broker or a price forecast.
The figure changes what gets argued about on the phone. A farm that knows one more month costs a named amount per tonne stops asking whether the offer is good and starts asking whether the deferred bid clears that amount, which is a question with an answer. The other one never had one, and that is why the same conversation happens every cycle with the same lot and a different price.