back start next
[start] [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [ 52 ] [53] [54] [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] [65] [66] [67] [68] [69] [70] [71] [72] [73] [74] [75] [76] [77] [78] [79] [80] [81]
52 138 Part IV: Bollinger Bands with Indicators  3/00 4/00 5/00 8/00 7/00 8/00 9/00 10/00 11/00 12/00 1/01 2/01 Figure 17.4 BB, MFI, and AD%, Marsh & McLennan, 150 days. Several tags of the upper band accompanied by weak indicator readings suggest caution, but a sell doesnt come until late on the chart. 
Chapter 17: Bollinger Bands and Indicators tags are accompanied by a succession of weaker indicator readings, resulting in a series of cautions, until you finally get an outright sell. Strong stocks are often hard to hold onto. Staying with strong stocks during a series of upper Bollinger Band tags can be a nerve-wracking process. But in the presence of indicator confirmation, strong stocks should be given the benefit of the doubt. Earlier we looked at M and W patterns and how Bollinger Bands could be used to clarify them. This relative framework allowed you to act even if a new low or high occurred in the second half of the pattern. The key was relative highs and lows, that is, highs and lows as a function of the Bollinger Bands, not of absolute price levels. Here we consider a second qualifier, an indicator that acts to increase our confidence. Take a W2 bottom as an example. The second leg down makes a new low, but if that new low does not break the lower band as did the first low, you now have a relative W4 and should be ready to buy on the first strong up day (Figure 17.6). Buying shortly after a | | 1---- --- | | | | | | | | | | | | | > | | | | | 1 1 | i j | | | si / I | 1 1 jllIIjiIIHn | TlJTiT]j | M • | | w-JI1 | bill | ,.] | lilH„.Jllll. | " j j | | | | | i j | | 1 1 | | 1 1 | | | | | | | | | | | | | | | llilihllil 1 | 11 ill | 11 III,llli | Mlii E | i Mil | | | llll III | llll | llli |
6/00 7/00 8/00 9/00 10/00 11/00 12/00 Figure 17.6 W2, relative W4,11% confirms, Dow Chemical, 150 days. A new absolute low in price but no new relative low; plus the indicator turns positive.
Part IV: Bollinger Bands with Indicators new low is made can he very scary, hut the fear can he reduced and confidence increased if the indicator used for confirmation does not go to a new low. In that situation we have two positive signals: one, no new low relative to the bands and, two, indicator confirmation. A sign of strength such as an up day on above-average volume and above-average range would be the third confirmation. The natural thing to ask is, If two pieces of confirmation are better than one, why not use three or even four? No reason at all why you cant. In fact, using more than one indicator for confirmation can improve your results. However, a problem arises if the indicators are collinear, which is to say, if the indicators are all singing the same song. That is the great trap of multicollinearity- several series containing very similar information used as though they were separate, independent inputs. Many a sorry trader has been caught in this trap, some completely unaware. The multicollinearity trap can be easily avoided; all that is required is a bit of discipline. Use only one indicator from each category, one momentum indicator, one trend indicator, one volume indicator, etc. (see Table 17.1)-similar to the approach of certain American Chinese restaurant menus, one from column a, one from column b... with no substitutions allowed. Using three different momentum indicators drops you directly into the collinearity trap; they are all saying the same thing derived from the same source. If the waiter were doing his job, he wouldnt take this order, but most market software is not that smart. It is possible to use indicators from the same category and avoid the multicollinearity trap if those indicators are not closely correlated with each other. But this requires testing and careful consideration, and should be avoided unless there is an overriding Table 17.1 Indicator Categories with Two Examples of Each Category Example Indicators Momentum Trend Sentiment Volume (open) Volume (closed) Overbought/oversold Rate of change, stochastics Linear regression, MACD Survey, put-call ratio Intraday Intensity, Accumulation Distribution Money Flow Index, Volume-Weighted MACD Commodity Channel Index, RSI
[start] [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [ 52 ] [53] [54] [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] [65] [66] [67] [68] [69] [70] [71] [72] [73] [74] [75] [76] [77] [78] [79] [80] [81]
|
Menu
