Before you lock in a buyer’s forward bid, find out what it charges you. Subtract the delivery-month futures price from the bid: what is left is the basis the bid carries. Compare it with the basis you expect at delivery. If the bid is 120, futures trade at 131 and you expect a basis of minus 6 at delivery, the bid carries minus 11, and locking costs you 5 a unit.
When the bid’s basis sits lower, the gap, in cents a bushel or dollars a ton, is the buyer’s fee for taking your price risk, and it is not part of the price of the crop. Whether that fee is worth paying depends on what you owe before delivery.
The fee is easy to miss because no line on the contract names it. A merchant calls on a Tuesday with a forward bid good until the end of the day, quoted as one price. Whoever built that price took on your price risk and charged you for it inside the quote. Keep one sheet of paper for the offer and write each number on it as you go. By the end it holds the bid, the futures quote, both bases, the fee, the range of waiting, and the volume you decided, with a date and a name. The free forward pricing calculator on this site runs the same subtractions.
Kansas State University put a figure on that fee for Kansas wheat. Working from forward bids collected every Wednesday at 18 elevators, Taylor and two colleagues estimated the cost of forward contracting at “$0.086/bushel for the years 2002 to 2007”, and then at $0.327/bushel for the period 2008 to 2012 once local basis turned volatile.
Per bushel that looks small, and a fee built into a price never arrives as a bill. The same authors multiplied it over a whole contract: $1,635 on a 5,000-bushel contract in the later period, against $430 on the same contract in the earlier one. No contract carried a line for either figure. It reached the farmer as “a wider implicit basis bid”: a bid whose basis sat further below the futures price.
Why does the forward bid already have a fee inside it?
The buyer takes two risks off your hands and is paid for both. Sign a forward contract and you hand over the futures price risk and the local basis risk together. The elevator on the other side normally covers itself by selling futures against the bushels it just agreed to buy, and keeps the basis risk. The Kansas State authors describe elevators guaranteeing a harvest price to farmers “in exchange for a fee, which we call a risk premium”, and note that the more volatile the local basis, the higher that premium runs.
That fee is what you pay for certainty. This article measures it two ways: the basis test, which compares the bid’s basis with the one you expect, and the waiting test, which compares the locked price with what waiting is likely to bring.
In a bid quoted as one price, the fee can only sit in the basis. Townsend and a co-author, in a peer-reviewed study of Oklahoma wheat, wrote this out: “the cost of forward contracting is measured as the expected change over time in the forward basis bids”. In plain terms, it is how much better the basis in bids tends to get between the day of the forward bid and delivery.
At an Oklahoma river terminal, with bids from 1986 to 1998, they put the cost between six and eight cents a bushel at 100 days before delivery, and around 30 cents a bushel for a farmer who priced at planting instead of selling at harvest.
Both figures are wheat at one terminal. The authors add one caution, about the 30 cents at planting: bids that early were rare in their data, and the result might not extrapolate to corn and soybeans, whose shorter growing season may let elevators provide forward contracts at planting at a lower cost.
Buyers can miss the fee too. Of the elevator setting those bids, the authors record: “The elevator manager believes that no risk premium is included in the forward bids, but the data show that there is.”
How do you find the fee nobody prints?
Look for it in the basis, not in the price, because basis moves far less from one year to the next. In Kansas State’s extension primer on basis, Dhuyvetter compares two wheat years at Dodge City and reports an average difference of $1.42 per bushel in price levels between the two years against only $0.17 per bushel of average difference in basis. That is why he treats a three to five year basis average as a usable forecast and never says the same about price.
Nobody can tell you next June’s price, but from your own price record you can estimate next June’s basis fairly well. If you do not keep one, you can start with the article on an annotated price record, a list of quotes with the source and date of each.
So subtract the futures price from the bid, and hold the result against the basis you expected. Basis is the local cash price minus the futures price, and Dhuyvetter is careful about the sign: “A negative basis implies the futures price is greater than the cash price”, which is the usual state of grain away from a delivery point. A basis further below zero is called wider. He also turns basis into a test for an offer: “If an expected basis is known, a forward contract bid can be evaluated.”
Here is the basis test in round numbers, in whatever unit and currency you sell in. The buyer offers 120 a unit, delivery-month futures trade at 131 today, and from your price record you expect a delivery basis of minus 6. The bid carries a basis of minus 11 (120 minus 131) against the minus 6 you expected. That gap of 5 a unit is the fee.
Rurivia calculation from a hypothetical forward bid of 120 a unit, a futures quote of 131 and an expected basis of minus 6 on the same lot.
Why doesn’t the expected price settle the decision?
The waiting test compares the locked price with what waiting is likely to bring. You expect 127 a unit at delivery, and waiting costs you 6 a unit. Count in that 6 only what you would carry by waiting and would not carry under the contract, such as interest on money you would otherwise have sooner. If the grain sits in your bin until delivery either way, storage is the same on both sides and drops out. So waiting is worth 121 on average, against the 120 you can lock today: a gap of 1.
Now write the lowest and highest delivery prices that would not surprise you, say 113 and 141. After the 6 of waiting, the result lands anywhere between 107 and 135, with the 120 well inside that range. A 1-unit edge means little against a 28-unit spread, so the expected price is not enough to decide on.
What decides is which end of that range you can live with, and the two ends do not weigh the same. Write down what falls due between now and delivery: the loan installment, the input bill, the land rent. At the top of the range the extra money makes a better year. Near the bottom, the crop may stop covering a payment with a fixed date, and then the problem is not a thinner margin but a talk with your lender. A price 14 below the middle can hurt far more than a price 14 above it helps.
That difference decides whether the fee is worth paying. A farm with fixed payments due before delivery is using the fee to remove the low end of the price range, the only end that can cost it the farm, and that is a sensible thing to pay for. A farm with no such payments pays the same fee to avoid a swing it could have ridden out. So the question is not whether the fee is fair but whether anything at the low end needs protecting.
Check the low end against the installment due that month and against the break-even price, the lowest price that covers what the crop cost you to produce; its own article walks you through working it out.
The protection has two costs besides the fee. A locked price does not rise when the market does, so you give up the top of the range. And the contract adds a delivery obligation: harvest less than you promised and you buy the shortfall at that day’s price. The ceiling on volume, in the section on what to do with each result, handles that one.
Where does your expected price come from, and what does it usually get wrong?
The expected delivery price is the one number in these tests that comes only from your own guess, and it is where research has found the mistake. In a peer-reviewed Brazilian study, Cruz Júnior and four colleagues asked 90 maize farmers in southern and central-western Brazil, in late 2008, to spread their price expectations across a list of price ranges.
Of the 81 whose answers could be compared, 62 gave a spread narrower than the futures market had actually produced, about three in four. Measured instead against the price history of their own regions, 44.4% gave a spread narrower than that history had produced. The authors read this as farmers holding a perception of risk below the risk the market carries.
The 90 were volunteers, reached as clients of a consultancy, as members of a state farm economics institute or through a state university, not a random draw of farms. The authors say this does not guarantee a statistically representative sample of Brazilian maize growers. So do not lean on the exact share. Take the pattern: many of these farmers pictured a narrower price range than the market delivered.
That pattern lands on the low and high prices on your sheet, not on the expected price. So when you write your own 113 and 141, do not trust a range just because it feels comfortable. Take the low and the high off the same price record the expected basis came from, and let a bad year you actually lived through set the low.
You can test your own optimism with one more subtraction. Futures at 131 plus your expected basis of minus 6 give 125 at delivery: the market’s implied price, two below the 127 you expected. Use 125 instead of 127 and waiting drops to 119, one below the 120 on offer, so the waiting test flips from wait to lock. When the verdict flips on a number you guessed, do not let that number decide alone.
The two tests are not two charges to add up. With the market’s implied price, locking costs you the 5 in the basis but spares you the 6 of waiting, so it comes out 1 ahead. The fee is already inside the waiting test, and adding the two would count it twice.
When the two tests disagree, go by the basis test, because of what goes into it. The futures quote is public and anyone can check it, and the expected basis comes off a record you kept. The waiting test rests on a price you guessed this morning, and the Brazilian researchers found that people guessing this way often picture too narrow a range.
The same researchers found selling ahead uncommon. Asked how else they sold their crop, only eight of the 90 farmers said they had sold part of their production in advance, and none of them had sold all of it. The eight who sold ahead sold part, which is the practical form of this decision: how much to lock, more than whether to lock.
What do the two tests leave out?
Three things sit outside the arithmetic. The first is quantity. MATba, the Buenos Aires grain futures exchange, explains in its teaching manual that a forward contract fixes the quality and quantity of the goods, and the moment, place of delivery and price, and that, unlike a future, its performance depends on the good faith of both parties.
Harvest less than you promised and you buy the shortfall at whatever the market asks that day. The Oklahoma authors note the same exposure from the elevator’s side: local elevators contract mostly with farmers, so they are liable when a farmer does not deliver on a contract.
The second is the buyer on the other end of the phone. In a 2021 online survey of 84 Italian arable farmers, Penone and two colleagues measured the intention to sign a marketing contract, not contracts actually signed. One concern they measured was “the risk of buyers not honouring contracts when prices fall”, yet security concerns made no substantial contribution to the intention. What counted most was the farmer’s own leaning toward contracts and how compatible the contract was with the farm’s quantity, quality, management and size.
So spend the call on the specifications, where the contract touches your operation and a question gets an answer you can check: which grade, what moisture, which delivery window, and what happens to the price if the load comes in below the agreed grade. A buyer who means to walk away from a contract will still say the right things about honoring it.
The third is the cost of waiting, which is easy to leave out. Leave it at zero and waiting looks free. You can work that cost out line by line with the article on whether it is worth waiting to sell. It is also the number most likely to flip the waiting test.
What do you do when the locked price falls inside, below or above the range?
Decide what you say on the phone from the basis test, not from the expected price. When the locked price falls inside the waiting range, the expected price gives no verdict, so decide in volume. Lock the share of the crop that has to pay for things due whatever happens, such as the installment and the input bill, and leave the rest open. In the example, a fee of 5 a unit against a 1-unit edge shows the expected price was never the argument.
Two other articles help here: with one you list the volume that already has a price fixed, contract by contract, and with the other you draw up a written selling plan with the share to price by each date. Check the share you are about to lock against both before you agree to it.
When the locked price falls below the whole range, either your range is too narrow or the bid is poor, and where your low and high came from tells you which. If they came off your price record, it is the bid. If they came from memory, the Brazilian finding may apply to you, and the first move is to widen the range, not to refuse the offer.
When the locked price falls above the whole range, which you will see fairly often in the calculator, it is tempting to read the bid as generous. What it tells you is that your own expectation sits below what the buyer will pay. The bid can still carry a fee in its basis, so run the basis test before you treat it as a gift. It is a reason to commit volume with some confidence, not to commit all of it.
Whichever way the verdict lands, the volume has two limits, and usually only one gets written. The floor is the price: nothing signed below the break-even price. The ceiling is the quantity: the bushels or tons a bad harvest still delivers, not the amount you expect, because the contract binds the quantity and the weather does not. Sign between the two.
The fee can also come out negative, and that is not an arithmetic error. It means the bid’s basis is narrower than the one you expect, so the buyer is paying more for delivery than, by your own record, the market will. Dhuyvetter describes this case in an example with cattle, where with a bid above the expected one the buyer is providing a strong financial incentive to sell today for future delivery. Take it as that incentive: a reason to commit volume, never a reason to sign past the ceiling.
When the fee is the number that stands out, change what you ask for rather than whether you sign. Ask the buyer to quote the delivery-month futures and the basis separately instead of a single price. An offer in two numbers can be negotiated, and a basis you were told is one you can file and compare next year. That is how the basis history behind this whole test gets built. Before you decide the fee is small, multiply it over the whole contract, as Kansas State did: a fee is quoted per bushel and paid per contract.
If you want price protection without the delivery promise, hedging with futures on an exchange handles the same risk another way, and you can work through it in the article on what price you keep after hedging. There you still carry the basis risk, instead of paying a basis fee.
The Oklahoma authors compared the two and did not find a tie. They conclude that this study and others have now found that “the costs of forward contracting are substantially larger than the cost of hedging”, setting their six to eight cents a bushel against the two cents a bushel the paper cites from an earlier study as the cost of hedging that wheat. That finding is about one terminal in those years, not advice about your buyer. And a buyer committed to take your grain on a set date can be worth something on its own.
Where to start
Half an hour, with the offer on the table, your price record open and a sheet of paper for this offer. Nothing here needs a broker, and none of it survives the call unless it is written down.
A forward bid comes as one number, and one number cannot be taken apart in front of the person quoting it. Asking for the futures and the basis behind it turns the conversation from whether the price is good to how it was built, and the question costs nothing.
Whichever way you decide, the answer becomes a line in your price record. Kept offer after offer, those lines are part of farm management: next year you can see what each lock cost you and whether the fee bought protection you needed. Other articles in the library’s marketing section build on the same record.