The instrument expires; the chart does not
A futures contract is an agreement with a stated expiry. In India's equity derivatives market several contracts on the same underlying trade side by side — the near month, the next, and one further out — and each has its own price, its own liquidity and its own last day.
The exact expiry day is set by the exchange in the contract specification, and it has been changed more than once. Read the current specification rather than a remembered rule; this lesson depends only on the fact that expiry exists, not on which weekday it falls.
What this Part cares about is narrower than the instrument itself, which belongs to Part 24. It is that a multi-year futures chart has to be assembled from dozens of separate contracts, and how they are joined changes what the chart says.
Why the join is not a move
When one contract is near its end and attention shifts to the next, the two are usually trading at different prices. That difference is ordinary: a contract expiring next week and one expiring in two months are not the same promise, and they are not worth the same.
So lay the contracts end to end in date order and the series jumps at every changeover. Nobody gained or lost anything at that jump. It is the seam between two instruments, not a movement in one.
A long chart built this way has a seam every month. Any return, average or level measured across one of them is measuring the seam as though it were the market.
Three contracts, one line
Illustrative contracts laid end to end, then joined. Tap each to see what the join costs.
The prices are unchanged between the two views. Only the offsets applied to the earlier contracts differ.
What back-adjustment buys and what it spends
The common repair is back-adjustment: shift every earlier contract so that its last price meets the next one's first, removing the seam. The result is one smooth line, and the shape of every move inside each contract is preserved exactly.
What it spends is the prices. Once the older contracts have been shifted, the numbers on the left of the chart are no longer prices anybody paid — they are prices minus an accumulated offset. A level read off that part of the line never traded.
Pushed far enough back, the arithmetic can drive the adjusted series to zero or below. A negative price on a long back-adjusted chart is not a data error; it is what the method does when the offsets outgrow the price.
There is a second method that multiplies rather than subtracts, keeping percentage moves exact instead of rupee moves and never reaching zero. Whichever is used, the answer to "what was the price then?" is the same: not this.
- The individual contract
- a real price history, for its own short life
- Contracts laid end to end
- real prices, with a seam at every roll
- Back-adjusted
- one continuous shape, prices no longer real
- A historical level from a back-adjusted chart
- never traded
- A return measured across an unadjusted seam
- includes the seam
A ten-year continuous futures chart shows a level that the series has "touched four times". What should you establish before using it?
Which of these does a long continuous futures chart genuinely support?
A continuous futures series is defined by three choices: which contract it rolls to, on what day it rolls, and how it joins them. Two providers making different choices produce two different histories of the same market, and neither is wrong. A chart that does not state all three is not evidence about the past — it is a shape, and 03.02 already established what a shape on its own is worth.
A stitched series at least contains everything that happened. The next two lessons are about data that is missing on purpose — the companies a list quietly leaves out, and the facts a test was never entitled to know.