The S&P 500 and the E-mini S&P 500 future almost never print the same number. The difference is not lag, not a data error, and not a sentiment reading. It is a financing calculation. A future hands you index exposure later, so you keep today the cash you would have spent on the basket of shares, and you give up the dividends those shares would have paid. The gap between the two prices is what the market charges for that swap over the time left until expiration.
An index future differs from its cash index because the futures buyer keeps the purchase money and earns interest on it, but forgoes the dividends the actual shares would have paid. Fair value is approximately the cash index plus financing cost minus expected dividends over the time to expiration, and basis is the observed difference between the futures price and the cash index. When financing exceeds dividends the future trades above the index, and that premium decays to zero at expiration regardless of what the index does.
Why does ES trade at a different price from the S&P 500?
Because both routes to index exposure deliver it at different moments, and the delay has a price. Buy the 500 stocks today and your cash leaves the account immediately, while every dividend those companies declare lands in your account. Buy one E-mini S&P 500 contract instead and you post margin, which is a performance bond rather than a purchase, the rest of your cash stays put earning whatever your account pays, and you collect no dividends because you own no shares.
Both routes end with the same exposure, so they cannot cost the same. If they did, a desk with a balance sheet would buy the cheap one and sell the expensive one until they matched. The futures price is the cash index adjusted for the two things the futures buyer does differently: keeping the money (a benefit worth the financing rate) and skipping the dividends (a cost worth the payout stream).
That observed difference is basis. Sign conventions split, which trips people up at the primary source. The CFTC glossary defines basis as "the difference between the spot or cash price of a commodity and the price of the nearest futures contract for the same or a related commodity," a word order that reads as cash minus futures. Equity index desks and most trading platforms quote it the other way round, futures minus cash, because that reads as a premium. This article uses futures minus cash throughout. A sign flip between two sources is usually nothing more than a convention difference.
What fair value actually measures
Fair value is the theoretical futures price at which neither route to index exposure is cheaper. It is a cost-of-carry calculation, the same family of arithmetic the CFTC glossary describes under carrying charges: the cost of holding a financial instrument over a period of time, including interest on the deposited funds. An index has nothing to store and nothing to insure, so carry collapses to two terms, interest earned and dividends missed.
The standard form is Fair Value = Cash x [1 + r(x/360)] - Dividends, where r is the financing rate, x is days to expiration, and Dividends is projected dividends over the period expressed in index points. Two details matter more than they look. The day count is 360, the money market convention for short-term US dollar rates, not 365. And the dividend term is properly built from a projected schedule of actual ex-dates, not an annualized yield.
That is why two fair value quotes for the same contract at the same second can legitimately disagree. Each one embeds a choice of financing rate and a choice of dividend forecast. The worked example below substitutes an annualized dividend yield to keep the arithmetic transparent. It gets the direction and the rough magnitude right, and it will not match a professional desk's number to the decimal, because theirs is built from a dated dividend schedule and their own term funding rate.
Hold the fork in your head. Down the stock path, cash leaves the account on day one and a dividend arrives on every ex-date between now and expiry. Down the futures path, the cash never leaves and no dividend ever arrives. Both paths land on the same index level at expiry. Fair value is the price adjustment that makes the two landings economically identical.
The arithmetic, worked exactly
Every rate below is illustrative, chosen so the arithmetic stays readable, not a claim about current market conditions.
Inputs: cash index at 5,000.00, annualized financing rate of 4.0%, annualized expected dividend yield of 1.5%, and 90 days to expiration, which is 0.25 years on the 360-day convention.
Financing component = 5,000.00 x 0.04 x 0.25 = 50.00 index points. Dividend component = 5,000.00 x 0.015 x 0.25 = 18.75 index points. Fair value premium = 50.00 minus 18.75 = 31.25 index points. Fair value = 5,000.00 plus 31.25 = 5,031.25.
There is a faster route to the same answer. Net carry is the financing rate minus the dividend yield, 4.0% minus 1.5% = 2.5% annualized. Over a quarter that is 0.625% of the index. 5,000.00 x 0.00625 = 31.25 points. Identical. The premium is nothing more than net carry applied to the index over the remaining term, which means you can estimate it in your head.
In currency: the E-mini S&P 500 is valued at $50 times the index price, and the Micro E-mini at $5 times the index, per the CME contract specifications. So 31.25 points x $50 = $1,562.50 of premium embedded in one long ES contract at 90 days out. On the Micro E-mini, 31.25 x $5 = $156.25.
| Days to expiration | Year fraction (360) | Premium (index points) | Per ES contract | Per MES contract |
|---|---|---|---|---|
| 90 | 0.2500 | 31.25 | $1,562.50 | $156.25 |
| 45 | 0.1250 | 15.625 | $781.25 | $78.13 |
| 30 | 0.0833 | 10.4167 | $520.83 | $52.08 |
| 0 (expiry) | 0.0000 | 0.00 | $0.00 | $0.00 |
All rows use the same 5,000.00 index and 2.5% net carry. Dollar figures are rounded to the nearest cent.
The 360-day year is a convention, not a law of nature. Run the same 90 days on a 365-day year: 5,000.00 x 0.025 x (90/365) = 30.8219 points, against 31.25 on the 360-day basis. The difference is 0.4281 points, or $21.40 per ES contract. Small, but every quoted fair value carries an assumption about which calendar produced it.
Why the premium decays to zero
The premium shrinks every day because what it pays for is time, and time is running out. At 90 days there are 90 days of net carry left to price. At 30 days there are 30. At expiration there are none, and the future settles to the index itself. The CFTC glossary calls this convergence, the tendency for prices of physicals and futures to approach one another. For a cash-settled index contract it is stronger than a tendency: the final settlement price is an index calculation, so the basis is forced to zero by the contract terms rather than by traders.
In the example, $1,562.50 per ES contract of premium at 90 days goes to zero by expiration even if the S&P 500 finishes exactly where it started. Averaged over the period, 31.25 points across 90 days is 0.3472 index points per day, or about $17.36 per contract per day.
A future trading above the index is a rate quote, not an opinion.
Treat that per-day figure as an average, not a schedule. Financing accrues on calendar days. Dividends do not arrive smoothly, because constituent ex-dates cluster into bunches rather than spreading evenly across the quarter, so the dividend component steps down in chunks. The real convergence path is a staircase with a downward drift, not a straight line.
Most explanations skip the next part. The decay is only a genuine economic cost if the cash you did not spend on shares is sitting idle. The entire justification for the premium is that the futures buyer earns financing on unencumbered cash. An institution does. A retail futures account may earn little or nothing on its cash balance depending on the broker and the account type, in which case the trader absorbs the convergence without collecting the interest that pays for it. In a funded prop account the question does not even arise, because the capital is the firm's and so is the financing. Check your own statement or your firm's agreement rather than assuming either way.
Rollover, and why back-adjusted charts need adjusting
The price gap between contract months at rollover is basis, nothing more. The deferred contract has an extra quarter of net carry left to accrue, so it prices above the expiring one whenever financing exceeds dividends. On the same numbers, 5,000.00 x 0.025 x 0.25 = 31.25 points, or $1,562.50 per contract.
That gap is exactly what continuous back-adjusted charts remove. Subtracting the roll gap from all prior history makes the series continuous, which is what an indicator needs, at the cost of making the historical levels synthetic. A support level read off a back-adjusted ES chart from two years ago is not a price anyone traded. Roll date choice and gap handling are covered in the guide to futures contract rollover. The narrow point here is that the roll gap is not a market move, it is the carry differential between two expiries.
When the future trades below the index
Change one input and the sign flips. Same 5,000.00 index, same 90 days, but a financing rate of 1.0% and expected dividends of 2.0% annualized. Financing = 5,000.00 x 0.01 x 0.25 = 12.50 points. Dividends = 5,000.00 x 0.02 x 0.25 = 25.00 points. Fair value premium = 12.50 minus 25.00 = -12.50 points, so fair value is 4,987.50 and the future trades below the cash index.
That is not a hypothetical. The sign of the basis comes down to a single comparison, short-term financing rates against the index dividend yield, and whichever is larger wins. Extended near-zero-rate periods put the yield on top and pushed index futures under cash. It is the cleanest available proof that neither a premium nor a discount carries directional information: same market, same index, same participants, and the sign of the gap decided by the rate environment. If you want to know which side applies today, compare a current short-term rate against the current index dividend yield instead of trusting a remembered number.
Why real basis wanders from theory
Observed basis is fair value plus whatever supply and demand distortion exists at that moment, and the distortion term is real. Five things push it around.
Order flow in the future itself. A large hedger or an index fund adjusting exposure can move the future away from theory faster than arbitrage capital chooses to respond. Borrow cost and shorting difficulty. The arbitrage is asymmetric, for reasons spelled out below. Dividend expectations. A cut, an initiation, or a shifted ex-date changes the dividend leg directly. Financing conditions. The rate that prices the basis is the arbitrageur's own term funding or repo rate, not a headline published number, and it differs by counterparty. Balance sheet. Carrying a cash-and-futures package consumes capacity, and capacity tends to get more expensive around quarter-end.
What keeps the two prices tethered at all is index arbitrage, which the CFTC glossary defines as the simultaneous purchase or sale of stock index futures and the sale or purchase of some or all of the component stocks that make up the index. Nothing in that definition pins the gap to zero. Arbitrage compresses large, persistent deviations and leaves small ones alone, because every leg costs money and carries execution risk across the basket.
The loop is also lopsided, and the lopsidedness is where it stalls. When the future is rich, the desk sells the future and buys the basket: one order into the futures book, one program into the equity book, no borrow needed. When the future is cheap, the desk has to buy the future and short the basket, which inserts a locate-and-borrow step on every name before the equity leg can fire, at a cost that varies by name and can be unavailable outright. That one extra hop is why a discount can sit uncorrected longer than a premium of the same size.
The overnight gap is usually not basis at all
The gap you see between ES and the S&P 500 at 3 a.m. has nothing to do with carry. ES trades nearly around the clock, Sunday evening through Friday afternoon Central time with a daily maintenance break. The S&P 500 cash index is computed from live constituent prices, and those constituents only trade during the US equity session. The NYSE core trading session runs 9:30 a.m. to 4:00 p.m. Eastern.
So overnight, the "S&P 500" number on your platform is very often a frozen last value from the previous close. The 40-point gap you are staring at is the index having moved since the cash market shut, measured against a number that stopped updating. Check what your own platform's index quote actually is, because some vendors display a last-computed value and others publish a derived or extended-hours estimate, and the two behave completely differently overnight.
Outside the US cash equity session, a displayed S&P 500 level may be a stale last value rather than a live calculation. The gap to ES in those hours mostly reflects index movement the frozen number has not recorded.
What this actually changes in your trading
Comparing a futures price directly to a quoted cash index level compares two different instruments, and it will mislead you about where the market is. Three consequences follow.
Reference levels do not transfer one for one. A prior-day high, a moving average, or a strategist's target expressed in cash index points sits at a different number on the ES chart. The offset is the basis at that moment. Translate it rather than assume it away.
The offset is not a constant you can memorize. It moves with short-term rates and with dividend expectations, both of which change continuously. The Federal Reserve H.15 release posts selected US interest rates daily Monday through Friday at 4:15 p.m., including the effective federal funds rate and Treasury bill yields, which is a fair gauge of how fast the financing input moves. A trader who memorizes "ES runs about 30 points over the index" will be wrong within weeks and badly wrong within a rate cycle.
Options on the two underlyings are not interchangeable. SPX index options are cash settled against the index itself. Options on ES exercise into the futures contract. The same strike number is not the same moneyness on both, because the underlyings differ by basis. Read an options chain against a chart of the other instrument and you are off by roughly the premium.
Copy trading carries the same trap. If a leader account trades an S&P 500 CFD or an index-tracking instrument while a follower runs ES futures, identical trades show a persistent price offset. That offset is basis plus whatever financing spread the CFD product embeds. It is not a copier fill error, and comparing raw prices across two different instruments will generate false alarms all day. Judge execution quality against each instrument's own book, on fill timing and slippage rather than absolute price. The same discipline applies when you watch correlated index products together, which pairs with the analysis of correlation risk across ES, NQ, YM and RTY. And because carry accrues against daily marks, it helps to know how the CME daily settlement price resets your position each session.
On contract mechanics: ES is cash settled on a quarterly March, June, September and December cycle, and final settlement is to a special opening quotation of the S&P 500 on the third Friday of the contract month rather than to a closing print. The $50 ES and $5 MES multipliers are long standing and stable, but confirm expiry dates and settlement terms on the CME contract specifications page before relying on them for a dated position.
When none of this matters
For an intraday scalper, basis is effectively a constant. It shifts the level of the ES chart relative to the index, but it does not move within a holding period measured in minutes in any way that touches a scalp. If you are flat by the close every day, the entire carry argument is a rounding error against your first tick of slippage.
The decay only bites on multi-week and multi-month holds, and the reference-level problem only bites if you import cash index levels onto a futures chart. Those are the two cases worth attention.
The honest limit is bigger than that. Understanding basis will not generate a single trade for a retail futures trader. Compressing a deviation requires simultaneous execution across hundreds of names, institutional financing, stock borrow, colocation and balance sheet capacity, and none of those requirements are negotiable. The value of everything above is defensive: not misreading the futures-versus-cash gap as sentiment, not treating a stale overnight index print as basis, not assuming a fixed offset, and knowing that a long-dated long future carries a component that erodes toward zero whatever the index does.
Frequently asked questions
Why is ES trading higher than the S&P 500 index right now?
Because the financing benefit of not owning the shares currently exceeds the dividends you give up by not owning them. The futures buyer keeps their cash and earns interest on it, so the future prices above the cash index by roughly that net carry over the remaining term. It carries no directional information about where traders think the market is heading.
What is basis in index futures?
Basis is the observed difference between the futures price and the cash index. Equity index desks quote it as futures minus cash, while the CFTC glossary phrases it as the difference between the spot or cash price and the price of the nearest futures contract, so check which convention a source uses before comparing numbers. Basis equals theoretical fair value plus whatever supply and demand distortion exists at that moment.
What is fair value for S&P 500 futures?
Fair value is approximately the cash index plus financing cost over the time to expiration minus the dividends expected over the same period. The standard form is Cash x [1 + r(x/360)] minus projected dividends in index points, using a 360-day money market convention. Any fair value number depends on which financing rate and which dividend schedule the calculator used, so two published figures can differ legitimately.
Does a futures premium over the cash index mean traders are bullish?
No, and reading it that way is one of the more expensive misreadings in index futures. The sign of the gap is set by whether short-term financing rates are above or below the index dividend yield. In a low-rate environment where dividends exceed financing, the same market with the same sentiment will show the future trading below the index.
Do index futures always converge to the cash index at expiration?
Yes, because the contract settles to the index itself, so the premium has nowhere to go. The CFTC glossary calls this convergence, the tendency for prices of physicals and futures to approach one another. For ES specifically, final settlement is to a special opening quotation of the S&P 500 on the third Friday of the contract month, so confirm current terms on the CME contract specifications page before trading a dated position.
Why does my platform show a huge gap between ES and the S&P 500 at 3am?
Because the cash index is only computed from live constituent prices while the US equity market is open, so overnight your platform is very likely showing a frozen last value. The NYSE core trading session runs 9:30 a.m. to 4:00 p.m. Eastern while ES trades nearly around the clock. That overnight gap mostly reflects index movement the stale number has not recorded, not basis.
Does holding a long futures position cost me the premium?
You lose the premium to convergence, but whether that is a real cost depends on whether you are earning the financing that justifies it. An institution earns interest on the cash it did not spend on shares, which offsets the decay by construction. A retail futures account may earn little or nothing on its cash balance depending on the broker, and in a funded prop account the capital and the financing both belong to the firm, so check your own statement or agreement.
How much premium is embedded in one ES contract?
Take the net carry rate, apply it to the index over the remaining term, and multiply by $50. Using an illustrative 5,000.00 index, 4.0% financing and 1.5% dividends at 90 days, the premium is 31.25 index points, which is $1,562.50 per ES contract or $156.25 on the Micro E-mini. Those rates are illustrative and real inputs move continuously.
Why does the December contract trade above the September contract?
Because the deferred contract has an extra quarter of net carry left to accrue, so it prices higher when financing exceeds dividends. On the illustrative numbers above that gap is 31.25 index points, or $1,562.50 per contract. This is the same gap that continuous back-adjusted charts have to remove, which is why historical levels on those charts are synthetic rather than traded prices.
Can a retail trader profit from trading the basis?
Realistically no, because compressing a deviation requires simultaneous execution across hundreds of stocks, institutional financing, stock borrow, colocation and balance sheet capacity. Index arbitrage is a professional business with hard capital and latency requirements. The value of understanding basis for a retail futures trader is defensive: not misreading the futures-versus-cash gap and knowing that long-dated exposure carries a decaying component.