Название | Encyclopedia of Chart Patterns |
---|---|
Автор произведения | Thomas N. Bulkowski |
Жанр | Ценные бумаги, инвестиции |
Серия | |
Издательство | Ценные бумаги, инвестиции |
Год выпуска | 0 |
isbn | 9781119739692 |
Every time I see a bearish bat pattern, my first thought is it's a Big W chart pattern. It's as if the Big W was computerized and the developer used Fibonacci ratios to determine the turning points. The bat pattern joins other Fibonacci‐based patterns: AB=CD, butterfly, crab, and Gartley.
I measured performance of Fibonacci‐based patterns differently than I do other chart pattern types. That's because we're looking for a reversal at the end of the pattern and not an upward or downward breakout. Therefore, the layout of this chapter is different from most other chapters in this book.
The above Results Snapshot gauges the downward move after the pattern ends. The stocks I looked at saw price turn lower 86% of the time after reaching point D. That percentage is not good but great!
The average decline in bull markets is a hair over 14%. The breakeven failure rate isn't too bad, at 17.7%. That means an average of almost 18% of the bearish bats will fail to see price drop more than 5% below the high at D (end of the pattern).
All of this may be confusing unless you can see what a typical bat pattern looks like. So let's take a quick tour.
Tour
Figure 4.1 shows an example of a bearish bat. The pattern begins at X, which is any significant turn. The locations of points A and B follow, but their placement is determined by one of two Fibonacci numbers. Once points XAB are found, the search for C can begin using one of six Fibonacci numbers. Following that, if you're lucky, D will appear close to one of four Fibonacci numbers. One last number, .886, helps validate the pattern. All of that gibberish means this pattern is complicated, and that makes it rare.
If you're handy with a slide rule (remember those?), then you'll be wasting your time searching for a bat. Even using a fancy electronic calculator will take too much time to find a bat. Use a computer with pattern‐finding software. I have a free one on my website called Patternz which will find this pattern and dozens of others. Try using it if you're serious about finding bats.
Even for a bearish bat, this example looks weird because of the uneven bottoms (A and C). Volume trends upward (E) when Table 4.2 says the average pattern shows a downward trend. As bad as this bat looks, it performs exceedingly well, with price dropping to F. This bearish bat is indeed bearish.
Figure 4.1 The five turns, X, A, B, C, and D, complete the bearish bat.
A swing trader taking a position just after D forms and watching the stock go all the way down to F might start dreaming about bats. Maybe even build some bat houses and hang them on his (or her) property. Here's a hint: a nearby pond is a plus. Let's return to trading bats. First, we have to find them.
Identification Guidelines
Table 4.1 shows identification guidelines. The guidelines are similar to other Fibonacci‐based patterns except the turns may use different numbers.
Figure 4.2 shows another example of a bearish bat.
Table 4.1 Identification Guidelines
Characteristic | Discussion |
---|---|
Appearance | Looks like a big W with turns located by Fibonacci ratios. |
BA/XA retrace | The ratio of BA/XA is either .382 or .5 |
BC/BA retrace | The ratio of BC/BA is one of .382, .5, .618, .707, .786, or .886 |
DC/BC extension | The extension of leg DC to BC is one of the Fibonacci numbers: 1.618, 2, 2.24, or 2.618. |
DA/XA retrace | The ratio of DA to XA is .886. |
Volume | Volume is downward the majority of the time, but this is an observation, not a requirement. |
Duration | I limited patterns to 6 months, but this is an arbitrary limit I use for most chart patterns. |
Figure 4.2 This bearish bat sees price dip after D but still breaks out upward.
Appearance. My guess is that the name comes from the picture which forms when you connect turns XABCD with lines, but also includes lines connecting XB and BD. The pattern looks like two wings. When trying to visualize this, consumption of alcohol may be a plus.
As I mentioned, the pattern's turns are found using Fibonacci numbers. Let's go through this pattern to see how the bat in Figure 4.2 qualifies.
BA/XA retrace. The ratio of leg BA to XA should be either .382 or .5. Here are the prices used in the retrace or extension calculations for the bearish bat. The high price of point X is 56.93. The low at point A is 47.81, B peaks at 52.00, C bottoms at 49.71, and the pattern ends at D with a high price of 55.98.
To determine the retrace, I use the height (high–low range) of the price bar at the target to determine whether or not it spans a Fibonacci number. For example, the BA/XA turn using the high price of B would be (52.00 – 47.81)/(56.93 – 47.81) or .46. Using the low price at B (50.90) in the equation gives (50.90 – 47.81)/(56.93 – 47.81) or .34. The range .34 to .46 straddles the .382 Fibonacci number, so I allow it as a valid XAB turn.
Using this method gives a lot of leeway to the pattern's turns but even so, the pattern is rare. Trying to narrow the turn window would limit even more patterns from being found, so this is the algorithm I chose. The software you use may think differently.
BC/BA retrace. In a manner similar to the BA/XA retrace, I plugged in the numbers for the BC/BA retrace (using the low at C). They are (52 – 49.71)/(52 – 47.81) or .55. Using the high at C (51.65) gives (52 – 51.65)/(52 – 47.81) or .08. The range of .08 to .55 encompasses .382 and .5, so I allow the turn as valid.
DC/BC extension. In a similar manner, I find the DC extension of BC. That's (55.98 – 49.71)/(52 – 49.71) or 2.74. Using the low at D (54), we get an extension of 1.87. The 1.87 to 2.74 values straddle all of the listed numbers (any one of which