Do we really need to establish a pattern of previous highs and lows to identify a point of resistance or support. I don't think that's required. Further I believe that the markets will not make things that simple for us. Projecting forward significant chart-based volatility levels can be the answer to the challenge.
This past week the S&P 500 Index reached significant daily support at 1012, when it got to within one half ATR(40) of the lower N Band.
Simultaneously the index sunk to the weekly SR 5 support line. The upper SR lines, 1 through 4 are tightly bunched together, which indicates heavy overhead resistance.
With all of this in mind I think there is a likelihood of a trading range setting up between 987 and 1208. The duration of the range should persist through September. I say 987 because that is roughly one ATR below the Weekly SR 5 support; which is the SR 6 support line.
To read more about projecting support/resistance forward, below is an excerpt from Volatility-Based Technical Analysis:
Support/Resistance: Flat is Out
Perhaps the most effective method of detecting support and resistance in the markets of old was a horizontal price line. A simple straight line sits at a previous significant high or low price. In the correct circumstances using this extreme price point of a past swing functioned well. Its strength was clearly in high volatility environments. In this first decade of our new millennium a previous high or low has given way to volatility based support/resistance.
In Figure 7.6 a chart of Gilead Sciences (GILD) shows horizontal alignment of prices at two levels. In circle 1 the low price for the day is set at 16.03 after a sizable gap down the previous day. Going forward it set the extreme low for over two months. In circles two and three this same low is touched and used as support. In circle 4 an upward price swing sets a high of 17.85. Two months later it appears to be used as a level of resistance. These horizontal alignments appear to be straight and true.
By contrast the same chart as viewed in MetaSwing (Figure 7.7) shows that support and resistance levels were due to different causes:
Circled Area 1 shows that price closes at the lower N Band locations. This price action and sets the peak in the N Band volatility measurement. Consequently and new S/R 6 level is set. One month later it serves as support for the low on November 22nd.
Circled Areas 2 and 3 make it clear that the dominant support areas are due to S/R 5 and 6. It certainly does not hurt that there was a previous price low set which lines up with the S/R line. S/R 5 and 6 were set because of the price action that occurred during December, which caused the lower N Band to turn down and form a peak.
Circled Areas 4 and 5 show a very interesting distinction. The highs in area 4 line up with S/R 3. Months later S/R 3 is reset higher and is a bit more accurate at defining resistance.
It isn’t that previous price levels don’t have an effect on current levels of support/resistance. They do but it is very limited, and is declining as time pushes forward.
It’s important to note that many high and low prices, represented by long upper and lower candle wicks, are present because of a few transactions. There is often only a tiny fraction of the day’s volume that sets this extreme, and therefore renders the price action at that level as insignificant. Just as often there were no actual transactions at that those prices; it was a data error reported by the exchange and remains uncorrected by your data service. If we have to wonder if the price was real or significant, it makes it that much more difficult to rely on a previous price level.