Futures contracts

A dated commitment, not a commodity

A futures contract is a commitment to exchange a set quantity of an asset, on a set date, at a price agreed today. You are not buying the thing, you are buying the commitment that bears on it. The distinction sounds academic, yet it explains everything else: a commitment can be resold, transferred, extinguished by an opposite commitment, and it needs no warehouse anywhere.

The idea is old and well documented. Rice already traded forward at the Dojima market in Osaka, in a form the shogunate recognised around 1730: standardised coupons backed by stored rice, passed from merchant to merchant. The modern shape appears in Chicago in the middle of the nineteenth century, when the exchange founded in 1848 gradually replaces private agreements with contracts identical to one another.

Standardisation is the central invention, not an administrative detail. While every deal carries its own quantity, its own grade and its own date, the only way out is to find the exact person who signed it with you and negotiate your exit. The moment a thousand contracts are strictly identical, they become interchangeable: anyone can take yours off your hands, because theirs is the same. Liquidity comes from that, not from how fashionable the underlying happens to be.

You can check the effect straight away in the way you leave a position. On an over the counter market, exiting is negotiated, and the price of exiting depends on who is facing you. On a futures market, exiting means selling what you bought, at the price on the screen, to a stranger you will never meet. Your position vanishes from the books because an opposite commitment extinguishes it, with nothing moving anywhere in the physical world.

The contract specification, line by line

Every contract has a specification published by the exchange that lists it, and that page is the first thing to read before trading one. It fits on a single screen, it costs nothing, and it holds everything you need in order to work out what you are risking.

Six lines truly matter. The underlying, what the commitment bears on. The contract size, meaning the quantity covered. The tick, the smallest price increment the exchange allows. The tick value, expressed in money. The trading hours, which are not the hours of the share market in the same country. Finally the settlement method, physical delivery or cash settlement.

The arithmetic that follows is the only piece you will use every single day. Take a contract covering 50 units and quoted in steps of 0.25 point: one tick is worth 50 × 0.25, so 12.50. A stop placed 8 points away sits 32 ticks away, and it costs 400 per contract. You need nothing else to size a position, and you cannot size one without knowing it.

⚠️ Two lines of that page are routinely misread, and misreading them costs real money. A physically deliverable contract must be closed out before first notice day, which falls before expiry rather than on the day of expiry. The evening settlement price, second, is computed by the exchange over the end of the session and is not necessarily the last price traded. That price, not the one you watched on your screen, decides the amount moved on your account that evening.

Print the page or paste it into your journal. A specification changes rarely, but it does change, and every serious trader on that contract knows the day it happened.

Delivery exists, and it is not for you

Delivery is not a threat, it is the reason the market exists at all. A farmer who will sell a harvest in six months, a manufacturer who will buy copper in the spring, a carrier that will burn fuel all year long: these people are not speculating, they want to know their price in advance. The futures contract sells them that certainty.

The mechanism can be followed by hand. Whoever will have to buy later buys a contract today. If the price rises in the meantime, the raw material costs more and the contract earns roughly as much. If the price falls, the raw material costs less and the contract loses. Either way the total barely moves, which was the whole point: the hedger was not trying to win, the hedger was trying to stop depending on the price.

Two economists put names on what happens there. John Maynard Keynes, in A Treatise on Money in 1930, describes normal backwardation: the party who hedges gives up part of the expected price to whoever accepts the risk instead, so that insurance has a cost and that cost is the other party's income. Holbrook Working, in the nineteen fifties, corrected an idea that was too simple: the industrial user does not cancel risk, he swaps it for a far smaller one, the gap between the price of his own goods and the price of the contract, which is called the basis.

What you should keep fits in one sentence. You will never take delivery of anything, because you will close before that, like the overwhelming majority of positions ever opened on these markets. Delivery concerns those who have a genuine use for the goods, and it is their steady presence, day after day, that gives these markets a depth nothing else imitates.

The clearing house changes everything

Here is the piece almost nobody explains to beginners, and it changes the nature of the market. When your order is filled, you are not tied to whoever took the other side. A clearing house steps between the two of you immediately: it becomes the buyer to every seller and the seller to every buyer. The legal name for that substitution is novation.

The consequence is considerable. Nobody is anybody's counterparty. You no longer depend on the solvency of a stranger, you depend on the clearing house, which takes no directional position of its own and demands a guarantee from every member. If one of them defaults, that guarantee is seized and the position is liquidated. Should it fall short, the clearing house draws on a fund every member has paid into, then may call on them for more: the solidity you are buying is that stack, not a promise.

This architecture was built slowly, and in Chicago it took its complete form in the nineteen twenties. It also explains why margin exists at all: a clearing house that guarantees everyone can only do so by measuring every evening what each member owes it, and by collecting immediately. Margin is not a nuisance invented to slow you down, it is the price of a guarantee you enjoy without ever thinking about it.

⚠️ Compare that with a market where your own intermediary is your counterparty. There, your gain is its loss, and no neutral third party stands between you. This is not an accusation, it is a structural difference: on a cleared market, the place where the price is formed and the place where your account is held are two separate things, and nobody along the chain pockets what you lose.

Size is fixed, and that is where it hurts

On most markets open to a retail trader you choose your size freely, sometimes down to a hundredth of a unit. On futures, you do not. A contract is a contract, its size is printed on the specification, and you can only trade a whole number of them. That is the first difficulty of a small account, and it has nothing to do with the difficulty of reading a chart.

Run the arithmetic, it is brutal. Take a contract whose point is worth 5, and a 4,000 account on which you allow yourself 1% of risk, so 40. A 40-point stop costs you 200, five times your budget. You cannot take half a position: either you take one whole contract and risk 5% of the account on a single trade, or you do not trade. Plenty of accounts destroyed on these markets were destroyed by that constraint alone, without a single analysis error along the way.

Micro contracts exist for exactly that reason. They track the same underlying, follow the same price, and are worth a tenth of the corresponding mini. Exchanges introduced them on stock indices in 2019, then extended them to other families. The same 40-point stop now costs 20, which fits inside the budget and even leaves room for a second contract. Granularity comes back, and with it the ability to manage a risk instead of enduring it.

⛔ One consequence is usually discovered too late: fixed size is not worked around by tightening the stop, it is worked around by changing contract. Pulling a stop closer so that a position fits a budget is the precise manoeuvre that turns a plan into a bet, since the stop stops depending on the chart. When the position does not fit, the only two honest answers are to step down a size in the contract, or to let the trade go.

Margin is not a down payment

The word misleads, because it names something else elsewhere. Here, initial margin is neither a part payment nor a deposit on a purchase. It is a performance bond, locked up, still your property, answering one single question: are you able to absorb an unfavourable day?

Its size is not an arbitrary percentage. Exchanges compute it from market scenarios, using portfolio methods of which the best known was introduced in Chicago in 1988, and it moves when volatility moves. Margin that jumps overnight is therefore not your broker being difficult, it is the exchange having just measured that the typical day has become wider.

Every evening the clearing house strikes a settlement price and actually moves the cash: the day's losers pay, the day's winners are paid, on the account, in cash. That is daily settlement. Your position is never an unrealised loss quietly piling up in a corner of the screen, it is closed out every evening and reopened at the settlement price. The system lets no debt grow, which is precisely what it is built for.

Maintenance margin is the floor your balance cannot go under while you keep the position. Below that floor comes the margin call: you must bring the account back up to initial margin, not merely back to the maintenance level, or cut the position. Without an answer inside the deadline your broker liquidates, and it liquidates at the price of the moment, not at the price that would have suited you.

⚠️ Intraday margin deserves a warning of its own, because it fools a great many beginners. Many brokers ask for only a fraction of the normal margin, sometimes a tenth, for a position opened and closed inside the same session. That reduction does not cut your risk by a single penny: the contract is the same, the point is worth the same, and a forty point move costs exactly the same amount of money. It only cuts what you are asked to post in order to carry it. An account that could hold one contract on normal margin and three on intraday margin is risking three times as much with the three, and it finds out on the day the market moves fast.

Expiry, the roll, and the chart that lies

A futures contract has an end date, known in advance. That single property carries three consequences no other reachable market imposes, and all three are paid for when they are ignored.

The first is that the crowd migrates. As expiry approaches, participants carry their positions over to the next contract across a few days: that is the roll. The expiring contract empties out, the gap between bid and offer widens on it, and a position left there becomes expensive to leave. The number that tells you where everybody went is not volume but open interest, the count of contracts still alive, published by the exchange: the day the next contract's open interest passes the current one's, the market has already moved across.

The second is that rolling costs money. The two contracts do not trade at the same level, because they do not carry the same financing, storage or expectations. Rolling means closing one and opening the other, so trading twice, across a gap that owes nothing to your reading of the market.

⛔ The third is the trap of continuous charts. A history several years long cannot come from one contract, since no contract lives several years. Charts therefore stitch successive expiries together, and to avoid a step at every joint most of them shift the whole older history by the observed gap. The prices you read on the left of your screen are then no longer prices: they are real prices plus the sum of every roll gap since. Over a long history such a series can even dip below zero, which proves better than any explanation that these are no longer quotes.

The practical consequence is straightforward. A historical level taken from a continuous chart gets checked on the contract that actually traded it before it becomes a reference, and any test run on a stitched series must state its splicing convention, failing which its results compare to nothing. Put your contract's roll date in your calendar, next to the opening hour of your session.

Volume, and why it is worth something here

Here is the major advantage of these markets, the one that justifies putting up with everything else. Each contract trades on a single, central order book. There are not ten venues where the same contract changes hands, there is one. Every transaction goes through it, and the exchange publishes the number of contracts traded.

That figure is a real volume: a count of contracts actually exchanged, identical at every broker, checkable against the exchange itself. It is not an approximation, not an estimate, not a tally of price changes seen by one server. The difference is anything but academic, since every reading that leans on volume, the confirmation of a break, accumulation at a level, a move running out of participants, becomes a measurement here instead of remaining an impression.

Book depth is visible too, with the quantity resting at each price, and it is the same for everyone. That is what makes order flow work on these markets and wobble everywhere else. You can watch a quantity disappear at a price, watch it come back, measure what was absorbed, and reach a conclusion your neighbour will reach from the very same numbers.

⚠️ Crossed with the open interest met earlier, that volume says one more thing. A move carried by high volume and rising open interest is made of positions being put on: new money is taking a side, and whatever it took will have to be undone one day. The same move with collapsing open interest is made of positions being closed, so of participants walking away, and the fuel is running out instead of building up. Note the lag while you are here: the session's volume is known the same evening, open interest is only published by the exchange the following day, and in a preliminary form at first. Reading tonight's session against yesterday's open interest is not a mistake, as long as you know that is what you are doing.

Practise: read a specification, then follow a roll

The exercise for this lesson needs no money and no open position, and it takes twenty minutes tonight. It produces measurable progress, because it replaces a vague notion with numbers you copied out yourself.

Pick a contract, any contract, and open its official specification on the website of the exchange that lists it. Copy seven lines into your journal by hand: the underlying, the contract size, the tick, the tick value, the trading hours, the settlement method and the last trading day. Then work out what a twenty-tick stop costs, and a fifty-tick stop. Those two amounts are your unit of measure on that contract, and you worked them out yourself instead of reading them somewhere.

Next, open the two nearest expiries side by side, the current one and the next one, and note three numbers for each: the last price, yesterday's volume and open interest. Repeat that reading every day for two weeks. You will watch the migration happen in front of you, day after day, and you will know for good what a roll looks like, which no explanation ever replaces.

Keep those readings in your journal, with the roll date written out in plain words, and write next to it what you will do with a position still open that day: carry it over to the next contract, or close it the evening before. A decision taken coolly two weeks ahead never looks like the one taken on the morning the gap between the two contracts jumps out at you.