How a gold base gets built, and what each event in the range is testing

What an accumulation range is claiming

The underlying claim is about inventory. During a decline, holders who need to sell do so. At some point the people still willing to sell at these prices are outnumbered by patient buyers who are happy to take the other side quietly, and the transfer happens sideways rather than at a single price. The range is the mechanism of that transfer, not a pause in a story.

Three ideas support the reading. Supply and demand determines direction, but only net of what the other side absorbs. Cause and effect says the time and activity spent inside a range is the cause, and the move out of it is the effect, which is why a long base is treated differently from a short one. Effort versus result compares how much activity a move required against how far it travelled, which is the lens that makes the dull middle readable. A large down candle that produces almost no net progress is informative precisely because the effort did not pay. That asymmetry is what you are hunting for inside the range.

Phase A: the decline stops

Phase A is about stopping, not reversing. Four events get named. Preliminary support is the first point in the decline where buying shows up clearly enough to produce a visible reaction, usually on expanding activity. The decline then resumes and reaches a selling climax, a wide range push lower on heavy activity that closes well off its low. The climax is where supply finally meets a buyer willing to take all of it.

What follows is the automatic rally. With sellers temporarily exhausted, price lifts easily because almost nothing is in the way. The height of that rally sets the top of the range, and the climax low sets the bottom. The last Phase A event is the secondary test, a return toward the climax low that should arrive with less activity and a narrower range than the climax itself. If the secondary test produces as much activity as the climax, supply has not been exhausted and Phase A is not complete. On gold this sequence often compresses around a scheduled release, which makes the activity comparison harder to read than textbook diagrams suggest.

Phase B: the long and uninteresting middle

Phase B is where the cause is built and where patience is tested. Price works between the two boundaries set in Phase A, visiting each edge more than once. The purpose of the repeated visits is testing. Each push toward the range high asks whether sellers are still waiting there. Each push toward the low asks whether supply remains. The informative pattern is a gradual decline in the activity and range of the down moves while the up moves hold their character.

Two practical points. Phase B has no fixed length, and anyone who tells you how long it should last is guessing. And the boundaries are rarely clean on gold, because the market overshoots both edges regularly without that meaning anything. Rather than treating the edges as lines, treat them as areas and watch what happens after the overshoot. This is ordinary range behaviour with a purpose attached to it, and the purpose only becomes visible in retrospect, which is the honest limitation of the whole phase.

Phase C: the spring and what it tests

Phase C contains the one event most people know by name. A spring is a push below the range low that fails to attract follow through and is reclaimed. Its function is a final test of supply. If very little selling appears when price trades under the obvious level, the people who would have sold there have already gone, and the market has learned something it could not learn any other way.

Two details matter. First, a spring is defined by the reclaim, not by the dip. Price trading below support is a question, and the answer is whatever happens over the next few candles. Second, not every base contains one. Some ranges move into Phase D without ever undercutting the low, and waiting for a spring that never arrives is a common way to miss the move entirely. On gold the undercut is frequently a stop run below a visible low, which is the same event described in different vocabulary. The useful question in both languages is identical: did the market accept the lower price or reject it?

Phase D: demand taking visible control

Phase D is where the evidence becomes mechanical rather than interpretive. You are looking for a sequence: a sign of strength, which is an advance that covers ground with noticeably less effort than earlier advances in the range, followed by a last point of support, a pullback that holds above the previous pullback low. Repeat that pair and you have higher lows forming inside the upper half of the range while the up moves widen.

The contrast with Phase B is the whole point. In Phase B the moves are symmetric and the range holds. In Phase D the symmetry breaks: down moves get smaller and shorter in duration, up moves get larger. This is where absorption stops being a theory and starts showing up as a change in candle character. If you are marking this live, the practical approach is to note each pullback low as it forms and require the next one to be higher. The moment a pullback takes out the previous one, the Phase D read is finished and you are back to a range.

Phase E: out of the range, and what a failure looks like

Phase E is simply the trend that follows. Price clears the range high and the old resistance area starts behaving as support on the pullbacks. Nothing in the method says how far the move travels, and the cause and effect idea is directional only in the loosest sense, so treating range width as a target is an assumption rather than a rule.

More useful is knowing what failure looks like. A base that fails does not usually collapse immediately. It clears the high, fails to attract continuation, returns inside the range, and then takes out the Phase D pullback lows one by one. That sequence is your evidence that the handover did not happen. The honest caveat for the entire framework belongs here: every phase label is applied with the benefit of knowing what came next, and in real time you are always one event behind. The only defence is to require the next piece of evidence before adding risk, and to define in advance which price makes the read invalid. Marking the range boundaries on the live chart and letting price come to them is more honest than deciding which phase you are in and waiting to be proved right.

Traps that belong specifically to bases

Three recur. The first is the spring that is not one. Price drops through the low with an expanding range and a close at the bottom, and never comes back. Nothing about the dip told you which it was, only the reclaim did, which is why acting on the undercut itself is a coin toss dressed up as analysis.

The second is the premature phase label. A single wide down candle that closes well is called a selling climax while the decline is still in progress, and the trader spends the next stretch of the downtrend looking for Phase B. The guard is the secondary test: no lower activity on the retest, no Phase A.

The third is range width. Gold can build a base that is wide enough that a stop beneath the range low makes the trade unworkable at any sensible size. A correct read on an untradeable structure is still not a trade. Recognising that early saves more money than any refinement of the phase labels.