No registration required! (Why?)

Charting condundrum


Ok, we all know of the benefits and limitations inherent in using a continuous contract construct for determination of support/resistance levels. My question, though, is this:

If a portion of the value of determining a given level (say the 200 day sma) lies with the fact that a lot of people are watching that particular level, what process should one use to determine what said level should be?

Here are the options, as I see it:

1) Slap a 200 dma on a daily chart of a continuous contract. The problem is that your calculation of a continuous contract may well (no, will probably be) different from the next person's. So your beautiful 200 dma may be yours alone.

2) Take the 200 dma from a daily chart of, say, the SPX and add the day's premium to it (premium being, in this case, ES - SPX). Obviously, you'd only be able to do this eo establish the 200 dma for the previous day's close, but it seems to me, that this is a less arbitrary way of establishing such a value.

The problem with method number 2 lies with the phrase "it seems to me". Just because it makes sense to me, doesn't mean it makes sense to "thee" -- and I'm particularly interested in what makes sense to thee, especially if thou art tossing around 1000 lot orders.

By the way, if #2 is the way to go, does this logic, then also apply to fib levels (i.e., take the fib level on the cash, then add the premium from the previous day's close to the future's price to get the analogous level).

By the same token, how about for the study of "super-profiles" -- large time-frame market-profile style distributions for establishing areas of support and resistance? Most certainly, it would seem, such levels would be affected (and distorted) by the compromises inherent in adjusting contract prices to account for premium decay in the creation of a smooth-appearing continuous contract. SPX, which lacks true transactional dimension, would not be a good fit for this, but SPY certainly might, right. Would one then take these levels off SPY, and add the "spy premium" (es - spy) from the previous day's close to come up with the appropriate levels on the ES?

Is anyone still with me here? This one's really left me stumped...

Cheers
I must confess that I got a little bit lost with what you're looking for towards the end of the post. I think that it's my long weekend brain not functioning fully.

Your questioning about which level is the one that most traders are watching is really one of reverse engineering the support and resistance levels of each day and seeing what math behind those levels consistently allows you to select them. On any non-trending trading day (which accounts for about 80% of trading days) the two undisputed levels are the support at the low of the day and the resistance at the high of the day. Any other levels that were created during the day may or may not have worked as tradable s/r levels depending on the money management that you applied at the time.

So the quest w.r.t. the levels is to try and find with X day MA most consistently hits those highs and lows, or something similar to this.

The dilemma of the futures contract premium and its decay is correctly addressed by you, in my opinion, by looking at the underlying cash price and using that to determine the future's level (by adding the previous day's premium) to the levels calculated. However, if you want to take into account overnight trading then this idea falls away because the cash doesn't trader overnight.

Fair enough. I got a little carried away. How about a specific example, then. Let's say we want the 200 day sma on the es contract, as the 200 day is a well-known benchmark. Given this fact, we want to make sure we're using the same benchmark as the big guns, right?

So, what do we do?

1) Run a 200 day sma on our continuous contract chart?

2) Run a 200 day sma on the SPX, then add the premium from the previous day's close to determine the "benchmark" level?

Sorry if I'm persisting to hammer at a point that is obvious to most people. Thanks...
poster: I would say that running the 200 SMA on the SPX is the accurate and correct option because the continuous contract (futures) will have an "unatural" jump at rollover which is around every 60 trading days. This means that the futures will have at least 3 "unatural" discrepencies for the 200 SMA.

In order to calculate the most accurate offset I would take a rather unusual approach. I would look at the difference between the 2 futures contracts at the time that they share the same liquidity which is usually just before and around the open of the rollover day.

This is when the expiring contract is nearing the cash price (about 8 trading days away) and the new contract will be trading at a premium to the expiring contract which is most representative of the offset. This would be an "observerd" and "real" premium as opposed to a theoretical and calculated premium.

I would then take that figure and extrapolate it through to the next expiration date and reduce it by a fractional amount each calendar day. I would calculate that fraction by dividing the premium by the number of calendar days to the next expiration. Assuming no interest rate changes during that period I believe that this figure would be the most accurate way of reducing the premium each day.

As an example, say the new/old premium was 1.80 points and the time to next expiration was 90 calendar days. I would reduce the premium by 0.02 points each calendar day. So from a Monday to a Tuesday that would be 0.02 points and over a weekend you would use 0.06 etc.


Yes, this makes sense. Ultimately, there is no "one" way to go about these things, is there? I suppose that is the reality. Thanks...