A bit of a rant but curious as well. We live in a microsecond HFT world, so why is it that we don't know the actual closing level of the S&P 500 until sometimes 4:15-4:20? On Friday, for example, the final closing number was almost a full point off the number that all the real-time feeds showed at 4:00, and we didn't show close to that final number until after 4:20. Does anyone know the mechanics of why the actual final number, or at least one that's within a tenth of a point, can't be determined and disseminated until so long after the actual close? I would think that it's nothing more than a simple math and computer science problem to gather up all the price feeds from everywhere a component stock trades, figure out the last one, and plug it into the weighting formula. I understand that you might occasionally have a busted trade, but this happens fairly consistently and at a magnitude that exceeds one or two busted trades on even the heaviest weighted index components. Thousands of HFT algos do this math millions of times a day, why can't the consolidated feed manage to pull it off?
You are using an out of date browser. It may not display this or other websites correctly.
You should upgrade or use an alternative browser.
You should upgrade or use an alternative browser.
Why so long to calculate the S&P 500 closing?
- Thread starter Sig
- Start date
A bit of a rant but curious as well. We live in a microsecond HFT world, so why is it that we don't know the actual closing level of the S&P 500 until sometimes 4:15-4:20? On Friday, for example, the final closing number was almost a full point off the number that all the real-time feeds showed at 4:00, and we didn't show close to that final number until after 4:20. Does anyone know the mechanics of why the actual final number, or at least one that's within a tenth of a point, can't be determined and disseminated until so long after the actual close? I would think that it's nothing more than a simple math and computer science problem to gather up all the price feeds from everywhere a component stock trades, figure out the last one, and plug it into the weighting formula. I understand that you might occasionally have a busted trade, but this happens fairly consistently and at a magnitude that exceeds one or two busted trades on even the heaviest weighted index components. Thousands of HFT algos do this math millions of times a day, why can't the consolidated feed manage to pull it off?
Are you referring to ES futures or SPX cash?
A bit of a rant but curious as well. We live in a microsecond HFT world, so why is it that we don't know the actual closing level of the S&P 500 until sometimes 4:15-4:20? On Friday, for example, the final closing number was almost a full point off the number that all the real-time feeds showed at 4:00, and we didn't show close to that final number until after 4:20. Does anyone know the mechanics of why the actual final number, or at least one that's within a tenth of a point, can't be determined and disseminated until so long after the actual close? I would think that it's nothing more than a simple math and computer science problem to gather up all the price feeds from everywhere a component stock trades, figure out the last one, and plug it into the weighting formula. I understand that you might occasionally have a busted trade, but this happens fairly consistently and at a magnitude that exceeds one or two busted trades on even the heaviest weighted index components. Thousands of HFT algos do this math millions of times a day, why can't the consolidated feed manage to pull it off?
Calc comes from primary exchange. So market makers and specialists have to process all the market on close orders for all the stocks and then finalize the closing price. Then the calculation can happen.
Needless to say, it's a very important number as billions (and possibly trillions) of notional are tied to it. So the process has to be precise. There are times when there isn't an official closing print - there was one like 7 years or so ago. It had to do with a technical glitch. I believe sept 11 didn't have an official closing price as well.
Calc comes from primary exchange. So market makers and specialists have to process all the market on close orders for all the stocks and then finalize the closing price. Then the calculation can happen.
Needless to say, it's a very important number as billions (and possibly trillions) of notional are tied to it. So the process has to be precise. There are times when there isn't an official closing print - there was one like 7 years or so ago. It had to do with a technical glitch. I believe sept 11 didn't have an official closing price as well.
You are referring to cash. The futures are settled based on the VWAP from 3:14:30 to 3:15:00 central time based on the "big" ES contract. Then the price is converted to the E-mini price which uses a .25 tick level vs the .10 tick level of the big contract. The settlement price on the futures is based on the futures, not the stocks.
You are referring to cash. The futures are settled based on the VWAP from 3:14:30 to 3:15:00 central time based on the "big" ES contract. Then the price is converted to the E-mini price which uses a .25 tick level vs the .10 tick level of the big contract. The settlement price on the futures is based on the futures, not the stocks.
Yes. I think sig was referring to the cash.
Yes. I think sig was referring to the cash.
In that case, what you posted above is correct, the MOCs are a tedious process. My old firm use to trade them actively and we often had to wait a long time to get the final MOC print when there was a huge market on close imbalance.
You are referring to cash. The futures are settled based on the VWAP from 3:14:30 to 3:15:00 central time based on the "big" ES contract. Then the price is converted to the E-mini price which uses a .25 tick level vs the .10 tick level of the big contract. The settlement price on the futures is based on the futures, not the stocks.
Wrong on the futures settlement.
Why is it so difficult for people to get simple things stated correctly?
All you have to do is go to the CME website and read what is there for the "Normal Daily Settlement Procedure".
"E-Mini S&P 500 Futures
Normal Daily Settlement Procedure
The lead month is the anchor leg for settlements and is the contract expected to be the most active.
The volume-weighted average price (“VWAP”) of all trades executed in the full-sized futures contract on the trading floor and in the E-mini futures contract executed on CME Globex will be calculated for a the designated lead month contract from 15:14:30 – 15:15:00 Central Time (“CT”), the settlement period. A multiplier of 5 will be applied to the quantities traded in the full-sized contract to reflect the 5 to 1 relationship between the full-sized and the E-mini contracts.
The combined VWAP for the designated lead month will be rounded to the nearest .10 index point."
Simply and clearly stated and correct.
Wrong on the futures settlement.
Why is it so difficult for people to get simple things stated correctly?
All you have to do is go to the CME website and read what is there for the "Normal Daily Settlement Procedure".
"E-Mini S&P 500 Futures
Normal Daily Settlement Procedure
The lead month is the anchor leg for settlements and is the contract expected to be the most active.
The volume-weighted average price (“VWAP”) of all trades executed in the full-sized futures contract on the trading floor and in the E-mini futures contract executed on CME Globex will be calculated for a the designated lead month contract from 15:14:30 – 15:15:00 Central Time (“CT”), the settlement period. A multiplier of 5 will be applied to the quantities traded in the full-sized contract to reflect the 5 to 1 relationship between the full-sized and the E-mini contracts.
The combined VWAP for the designated lead month will be rounded to the nearest .10 index point."
Simply and clearly stated and correct.
That is EXACTLY what I stated. I actually read it from the CME website. LOL.
Sorry for not being specific, I was talking cash. Thanks for the insight, curious as to why the MOC process can't be automated to take place more quickly? You'd think the specialists could set up algos to figure this all out within a couple seconds, especially since the MOC orders need to be in 15 min before the market closes? I'll admit I have no experience with the behind the scenes mechanics of MOC and am curious to hear from someone who has.
Wrong on the futures settlement.
Why is it so difficult for people to get simple things stated correctly?
All you have to do is go to the CME website and read what is there for the "Normal Daily Settlement Procedure".
"E-Mini S&P 500 Futures
Normal Daily Settlement Procedure
The lead month is the anchor leg for settlements and is the contract expected to be the most active.
The volume-weighted average price (“VWAP”) of all trades executed in the full-sized futures contract on the trading floor and in the E-mini futures contract executed on CME Globex will be calculated for a the designated lead month contract from 15:14:30 – 15:15:00 Central Time (“CT”), the settlement period. A multiplier of 5 will be applied to the quantities traded in the full-sized contract to reflect the 5 to 1 relationship between the full-sized and the E-mini contracts.
The combined VWAP for the designated lead month will be rounded to the nearest .10 index point."
Simply and clearly stated and correct.
The end of each month, ES futures settle at the 4 pm price, rather than 4:15. The reason for this is so that month-end statements, for money managers, etc., which often include stocks, ETFs, and futures, are aligned. If not, the extra 15 minutes trading could distort NAV, TE, etc.
CNBC never seems to understand this, so not surprised others are aware as well.