Well, a few things here. Dividends do change, often in fact. So that uncertainty has to be factored in. Second, when a dividend is paid out, are you properly accounting for the fact that you can "re-invest" that cash flow back into the S&P? So let's call SPX cash. If one is selling SPX, they are losing this benefit. They need to ask a higher price to be compensated for the loss in cash flow. This higher price will be reflected in the difference between SPX and ES. One way to evaluate this is to set up the arbs and see where you lose money. Long ES/Short SPX. Long SPX/Short ES. Yes, there is a borrowing cost for SPX which right now is zero or close to it. So using this theory ES should trade lower then SPX. If every stock went ex-dividend tomorrow this benefits the SPX holder! Why? Cash flows today are worth a lot more then then cash flows 3 months from now. Hence there should be a premium to SPX over ES.
I'm talking a week or two in the future here, many dividends are announced this far out (http://www.dividend.com/ex-dividend-dates.php) enough for my purposes at least. Just to make sure I've got it right, if you assume interest rates are 0, then cash flows today aren't worth any more than cash flows 3 months from now, right? And the SPX index takes into account reinvesting dividends into the entire index (not just the stock that paid out the dividend). So the main reason for today's reality, where Dec ES is in fact trading 7 point below SPX, is the fact that the Dec ES is taking into account that all dividends due between now and the Dec expiration will have been paid by that Dec expiration, and today's SPX price is reflecting only the dividends that have been paid so far. If you paid out all the dividends today, SPX price would match the Dec ES price (assuming 0% interest)?
No, the prices would not match. If you paid out all the dividends today the SPX would trade at a premium since those dividends would be re-invested today and earn the index return from now till expiration.
The index return between now and then is unknown and could be negative or positive. More importantly, why is Dec ES in fact trading 7 point below SPX today if not for the intervening dividend payments? And why does the delta between the Dec ES and SPX get smaller the days where a significant number of components go ex, and stay the same on days that few do?
You do realize how little 7 pts is right? It's 32 bps. How exactly are you trying to exploit this again? Honestly man, I have no idea what you are trying to do. Maybe if you explained what you are seeing here and how you are trying to exploit it, it might help in finding where the error is if there indeed is one. Do you have any idea what the implied differential is on a cash index basket i.e. lifting offers on 500 stocks or hitting bids on 500 stocks? Here is a hint, it ain't 32 bps.
I'm not trying to exploit anything related to the difference between SPX and ES. I am working with 5-15 point spreads, so 7 points is extremely relevant to what I'm doing and if the ES/SPX delta decays an unexpected point during a 2 week period on a 5 point spread that's a big deal for me. In some cases completely unrelated to the delta between SPX and ES it makes sense for me to use ES and in some cases it makes sense for me to use SPX (related to cash vs delivery of underlying on expiration). I hold to expiration so I need to have a good an idea as I can of what that delta will be between the two on the expiration day. I figured I can't be the only one who cares what the difference between ES and SPX will be on some given day in the future, hence the question.
You do know that ES and SPX have completely different option settlement procedures right if we are talking about options since obviously you can't hold SPX cash.
Let's assume I know wtf I'm doing when it comes to ES and SPX and have held thousands of the options on both to expiration over many years. Precisely because SPX options settle to cash and ES options settle into the underlying ES you can imagine that it's useful to use one in some instances and the other in other situations. This isn't my first rodeo, I'm just trying to fine tune something I've been doing for a while and for that I needed a compiled set of data on dividends on S&P 500 component stocks, that's all. I guess I could have said I was doing an academic paper on the relationship of bacon prices to the S&P 500 and my question would hopefully be equally valid. If you've got any pointers on where to find the data I'm seeking I'd be most grateful, but if it doesn't exist I'll just do the work to compile it myself.
When asking for help, gratitude and appreciation usually offers nice dividends (pun intended). Getting historical dividend data should not be that terribly difficult. Are you saying a simple google search turned up no results? http://www.dividend.com/dividend-history/