FSP fallback methods

The Final Settlement Price of a commodity contract is normally the simple average of the polled spot prices on three days. Those days are the expiry day E, and the two days before it.

That rule works only if the spot market was open on all three days. Sometimes it was not. Mandis shut for festivals, strikes and floods.

So SEBI and the exchanges keep a ladder of fallbacks. You climb down it one rung at a time, and only when the rung above has failed.

The order is the examinable part. Each rung is a worse approximation than the one above. The bottom rung is a stale price from days earlier.

The ladder exists because a contract has to settle at some number. A missing poll is not a reason to leave money unpaid.

The normal rule first. FSP is the simple average of the polled prices of the last 3 days of the contract including the expiry day — E, E-1 and E-2.

Rung 1 — SEBI's shortened average. One methodology prescribed by SEBI is to take the average of the last 3 daily polled prices, or 2, or a 1-day polled price, of whichever date is available, going back as far as E-3 day — and this applies only provided the polled price on the expiry day E is available. That proviso is the hinge of the whole ladder.

Rung 2 — an alternate basis centre. If the polled price on E day is not available, exchange circulars take over. The workbook's illustration is NCDEX, where the E-day polled price for an alternate basis centre is used, subject to conditions designed to arrive at a fair FSP for the main basis centre on E.

Rung 3 — extrapolation from the futures-spot trend. If those conditions are not fulfilled, another alternate methodology applies: extrapolate futures traded prices to arrive at the spot price on E, on the basis of the trend of futures-versus-spot movement in the last 30 days.

Rung 4 — futures prices themselves. If extrapolation is also unworkable, the futures prices of the last 3 days, after removing outliers, may be used to arrive at a fair spot price for settlement — again subject to conditions.

Rung 5 — the last available poll. If none of these alternates can be used, the last available polled spot price is taken as the FSP.

Note who owns which rung. Rung 1 is SEBI's prescribed methodology. Rungs 2 to 5 come from exchange circulars, and the sequence above is NCDEX's.

One product never enters this ladder at all. Commodity index futures are not polled. Their FSP is computed after 5:00 pm from the weighted average traded price of the constituent futures between 4:00 pm and 5:00 pm, and where no constituent trades, an appropriate close-price method is used instead.

A gold contract expires on 20 October. Two scenarios, and the ladder sends them to different rungs.

Scenario A — the E-day poll exists. The mandi was shut on 19 October only.

The normal E, E-1, E-2 average is impossible. Rung 1 applies, because the E-day price is available. Take the last three available daily polls going back to E-3 — 20, 18 and 17 October:

(50,100 + 49,700 + 49,500) / 3 = Rs 49,766.67

On a 1 kilogram gold lot quoted per 10 grams, the seller invoices 49,766.67 x 100 = Rs 49,76,667.

Scenario B — no E-day poll. This is the workbook's own case: spot markets are closed for a week up to 20 October for Diwali, while the futures expiry is 20 October. Rung 1 is unavailable, because it requires the E-day price.

The exchange now works down rungs 2 to 4 — an alternate basis-centre poll for 20 October, then extrapolation from the last 30 days of futures-spot movement, then the last three days of futures prices with outliers stripped. Suppose none of the conditions attached to those methods is satisfied. Rung 5 applies, and the last available polled price, as of 13 October — say Rs 49,200 — becomes the FSP for a 20 October expiry.

The same 1 kilogram seller now invoices Rs 49,20,000. That is Rs 56,667 less than Scenario A, on a price seven days stale, and it is the settlement number for every open position in the contract.

Chapter 3 (Commodity Futures), section 3.10, in the two paragraphs that follow the E, E-1, E-2 rule. Chapter 2 supplies the index-futures carve-out.

This page teaches the exceptions; Final Settlement Price teaches the normal rule, and the two are examined together. Expect: how far back may the shortened average reach (E-3), what condition must be satisfied for it (the E-day polled price must be available), what is the trend window for extrapolation (30 days of futures-spot movement), and what is the last resort (the last available polled spot price). A scenario question that hands you a missing E-day poll and four options is the standard hard version.

• SEBI's shortened average needs the E-day poll. Lose E-1 and E-2 and you can still average; lose E itself and you drop straight to the exchange's own circulars. This proviso is the single most tested point on the page.
• E-3 is the floor. The shortened average reaches back three days before expiry, not five, and not to the start of the month.
• Rung 1 is SEBI's; rungs 2 to 5 are the exchange's. If asked who prescribes the alternate basis-centre route or the extrapolation, the answer is exchange circulars, and the workbook's example is NCDEX.
• The ladder never uses the Last Traded Price. Even at rung 4 it uses three days of futures prices with outliers removed, not the final print.
• The extrapolation window is 30 days of the futures-versus-spot relationship, not 30 days of futures prices on their own.
• The last resort can be badly stale. In the workbook's Diwali illustration a 13 October poll settles a 20 October expiry.
• Index futures are outside this ladder entirely. They are never polled, so no fallback for missing polls can apply to them.