Skip to main content

Teaching You to Trade Stocks 78: Continuing the Discussion on Line Segment Division

2007/9/6 22:28:31

In Memoriam

Pavarotti

(Though history has seen many better than you, in our era, you were the best)


Denouncement

Ding Yi

(Being ugly isn't a crime, but coming out to deceive people is unacceptable)


Dai Yuqiang

(Slippery voice, affected stage presence, too old for Super Boy — just go buy some tofu or grind against a wall)


Fan Jingma

(Originally far more skilled than the previous two, but once you started hyping the so-called "Three Chinese Tenors" with them, all hope is lost for you)



Note: everyone, don't be so neurotic sometimes. This ID will not buy any new stocks now — if anything, only working with stocks bought earlier at low prices. This ID's last stock — everyone here knows — was 002149, bought on its first day of listing. The transaction records were posted that day and deleted a few hours later. What happened afterwards, everyone saw.

This ID only drinks juice at lunch, and only fresh-squeezed — nothing to do with any brand.

I had said I'd start a new lesson, but seeing that many people clearly still haven't figured it out, and since today wasn't supposed to be about stocks anyway — this is essentially borrowing time from other things for this remedial session.

The division of line segments comes down to the two types described in previous lessons. Based on the complete classification of these two types, there is nothing that cannot be uniquely divided. But the moment it comes to actual division, many people get confused. Why? Because the most basic concepts still haven't been grasped.

First, both line segments and strokes have direction. A stroke starting from a top must end at a bottom. Likewise, a line segment beginning with an upward stroke must end with an upward stroke — it's impossible for a line segment to start with an upward stroke and end with a downward stroke. Since the starting fractal of an upward stroke is a bottom, and the ending fractal of a downward stroke is also a bottom, in other words, a line segment cannot go from bottom to bottom or from top to top. This is the most basic concept.

Similarly, just as within a single stroke the top cannot be lower than the bottom, within a single line segment, the top endpoint must be higher than the bottom endpoint. If you draw a line segment that doesn't meet this basic requirement, it's definitely drawn incorrectly.

Since charts extend continuously, unless it's the very first segment after a new stock's listing, every line segment is one that breaks its predecessor. If your division can't ensure that every preceding line segment is broken by the one that follows, then your division is definitely wrong. Line segment breaking can be transmitted backward in time — meaning a line segment broken by a later one must itself have broken the line segment before it. If this principle is violated, the line segment division definitely has problems.

Of course, in actual practice, you don't need to start from the first day of listing. Generally, you start from the most recent highest or lowest point visible on the K-line chart. For example, if you're just starting to divide the 1-minute chart today, you could start from yesterday afternoon's plunge low of 5,224 points. But doing so obviously prevents correct understanding of the larger trend. To have a clear analysis of this market move, even if you don't start from the July 6 low of 3,563, you should at least start from the August 17 low of 4,646.

Once you've chosen the starting point, you can begin dividing into segments. If you're experienced, you can divide directly into segments, since fractals and strokes can be mentally computed — you can proceed straight to segmentation. But if you're not yet experienced, start with fractals, then strokes, then line segments — this is more reliable.

In actual division, you'll encounter some peculiar-looking line segments. In reality, so-called peculiar segments aren't peculiar at all. It's just that most people have a mental impression that line segments should be waves getting progressively higher or lower, very simple-looking. In fact, line segments don't need to be like that at all. Generally, in near-unidirectional movements, line segments are simple without overly complex situations. But in oscillating movements, the likelihood of so-called peculiar line segments increases dramatically.

All peculiar line segments are caused by the first type of stroke-level break appearing in a line segment but ultimately not developing into a line-segment-level break in that direction. This is the sole reason for peculiar line segments. Because if the line segment could be broken at the line segment level in that direction, it would be perfectly normal — nothing peculiar about it.

Note: there's a detail that requires attention here. A line segment will always eventually be broken by another line segment, but after a stroke-level break appears, it doesn't necessarily develop into a line-segment-level break in that direction from that stroke.

From the most basic concepts, we know every line segment has direction. For example, line segment B has a downward direction — meaning it's a line segment starting with a downward stroke, so its ending stroke must also be a downward stroke. Therefore, a first-type stroke break of this line segment must be an upward stroke. But after this stroke, if no fractal forms in the characteristic sequence, the first type of line segment break can't be satisfied, and thus no line segment break forms in this direction.

A line segment cannot be broken by a line segment of the same direction. Any line segments of the same direction either have no relation to each other, or one is actually a continuation of the other — meaning the earlier line segment actually never completed.

But when a first-type stroke break of a line segment ultimately fails to develop into a line-segment-level break in that direction: in the example above, after the upward breaking stroke completes, the next stroke is surely downward. This downward stroke must form a downward line segment — otherwise, it would mean the earlier upward breaking stroke could extend into a line segment, contradicting the assumption.

If this downward line segment breaks below the bottom of that upward stroke, then the original line segment B hasn't ended and is continuing to extend. In this case, if that upward stroke exceeded the high point of line segment B, you'll see a situation where the line segment's starting point isn't the highest point. (Note: consistent with this situation, in yesterday's chart, point 81 should be at the 5,268.74 position of 09051101, while point 82's position remains unchanged. This is because the originally marked position was a sharp drop where the data collection at the time may have been somewhat chaotic. After using the data correction function, the actual price was found to be higher than at 09051101, so this correction is necessary.)

If this downward line segment doesn't break below the bottom of that upward stroke, then we can confirm that the upward stroke can indeed extend into a line segment, and at this point, line segment B has definitely been broken.

Note: there's a crucial prerequisite in this example — line segment B must have already confirmed that it broke the line segment before it. If line segment B's break of the preceding line segment hasn't been confirmed, then confirm that first — the analysis here doesn't apply.

From this example, you can understand the similarities and differences between stroke breaks and line segment breaks. For the second type of line segment break: for example, if line segment B is the second type relative to line segment A, and line segment C fails to form a second characteristic sequence fractal and directly makes a new high or low, then these cannot be considered three separate line segments — line segments A, B, and C combined can only count as one line segment.

Additionally, you must pay attention: for the second characteristic sequence fractal judgment in the second type, inclusion relationships must be strictly processed. The requirement from the first type — where inclusion relationship processing cannot cross the hypothetical boundary point — doesn't apply here. Why? Because in the first type, if inclusion relationships appear across the boundary point of the characteristic sequence, it proves the reversal force against the original line segment is particularly strong, so naturally you can't use inclusion relationships to mask the presentation of that strength. But in the second type's second characteristic sequence, its direction is consistent with the original line segment. The appearance of inclusion relationships means the original line segment's energy is abundant, and the second type itself already implies insufficient reversal energy against the original line segment. Given all this, inclusion relationships must of course be applied.

Through the above explanation, no line segment problem should be able to stump anyone — provided, of course, that you can understand the content above.

Note: a reminder must be given here, as this has been mentioned before: if within a line segment, the highest or lowest point isn't an endpoint of the line segment, then in any analysis using line segments as the foundation — such as constructing the smallest-level hubs from line segments — the line segment can be standardized so that its highest and lowest points are at the endpoints. Because in analysis based on line segments, each line segment is treated as a basic component without internal structure, so you only need to care about the line segment's actual range, which means you only look at its high and low points.

After standardization, all upward line segments start at their lowest point and end at their highest point, and all downward line segments start at their highest point and end at their lowest point. This way, all line segment connections form a continuous, end-to-end connected zigzag line. Complex charts thus become very standardized, providing the most standard and fundamental components for subsequent analysis of hubs, trend types, and so on.