The first chart plots all exponential moving average values between 1 and 200, left to right on the x-axis, as of Friday's close (10/09/2009). The chart includes a "best fit" straight trendline for the curve, and the R2 value (closeness of curve fit) for the closing price. Back in late February/early March, this R2 value was at 0.99+... it descended to zero as price rose through its transition period and then climbed steadily upward again as price continued climbing. Currently at 0.8777, R2 is at its highest point since the spring reversal and has never reversed.
The next three charts plot an x-axis price range between 1000 and 1100, and a y-axis R2 range between 0.7900 and 0.8900. These x & y ranges are zoomed ranges that are relevant to current prices and R2's. In a nutshell, these charts show a price curve where the maximum R2 (trend fit) can be obtained, and show where actual closing price is located relative to the "peak trend capacity" for a given moment. The further away price is from the curve peak, the more it is deviating from the trend, either upward (right of the curve peak) or downward (left of the curve peak).
The three charts are plotted respectively for Monday 09/28, Monday 10/05 and the coming Monday 10/12. Don't get confused on how I can plot next Monday vs. the R2... all I have to do is enter each of the x-axis prices shown into the Monday time frame chart to plot and see the resulting curve that's upcoming.
Two important things to recognize here:
The curve peak rose from 0.8350 to 0.8675 for the week between 09/28 to 10/05. but slowed significantly for the week between 10/05 to (coming Monday) 10/12, from 0.8675 to 0.8875:
1- {(0.8875-0.8675) / (0.8675-0.8350)} = 38.5% drop in R2 acceleration, a rather interesting number for Fibonacci followers.
And, the price range for the R2 curve peak is now moving backwards.
09/28 price range for peak R2 = 1031 through 1036 (forward from 9/21)
10/05 price range for peak R2 = 1043 through 1048 (forward from 9/28)
10/12 price range for peak R2 = 1038 through 1043 (backwards from 10/05)
(For reference but not shown because it runs below and off my selected chart scale: chart for 09/21 price range for peak R2... the range was 1014 through 1017, peak R2 was 0.7960, demonstrating that there's been a progressive uptrend to this point. With a price location of 1064.66 on the curve, it was quite an upward deviation at that point in time).
Peak zones mentioned above are enclosed within the red vertical lines.
Monday 09/28, with closing price and its R2 indicated:
Monday 10/05, with closing price and its R2 indicated:
Monday 10/12 (I've leadered Friday's close and R2 levels on the curve for the sake of having some reference points for this curve for Monday). See the R2 curve's peak zone moving backwards on this chart?:
A third thing to recognize is how higher prices plowed into the right-side of the curve deviation from the curve's peak, and doing so under a condition of a broad trend deceleration to a Fib threshold for Monday.
My take: there's nothing more devious than kicking off a new negative IT trend reversal with counter-trend higher prices.
Disclosure: I'm fully short for a position trade as of Friday's close. And my hamster is back (re-introduced last week and a bit early but by personal request and favor
) for an encore presentation.
How you came up with it is beyond me, but it is interesting nonetheless. 
I hope you are right, too.
