The main Elliott wave count expected a strong upwards day for Wednesday, but this is not what has happened.
Summary: It is still more likely that the correction is incomplete, but this has decreased in probability today and should be discarded below 1,196.24. The target for a wave up is still at 1,243. The target for the alternate idea, a strong fifth wave down, remains at 1,146.
Click on charts to enlarge.
Main Daily Wave Count
At this stage I judge this main wave count to have an even probability with the alternate below. I will let the structure of downwards movement, and momentum, tell us which wave count is correct over the next few weeks. At this stage they both expect more downwards movement so there is no divergence in the expected direction.
This wave count sees a five wave impulse down for cycle wave a complete, and primary wave 5 within it a completed five wave impulse. The new upwards trend at cycle degree should last one to several years and must begin on the daily chart with a clear five up.
The first five up may be a complete leading expanding diagonal. Within leading diagonals the second and fourth waves must subdivide as zigzags. The first, third and fifth waves are most commonly zigzags but they may also be impulses. This wave count sees minor waves 1, 3 and 5 as zigzags.
Leading diagonals are almost always followed by deep second wave corrections, sometimes deeper than the 0.618 Fibonacci ratio. This wave count expects a big deep correction downwards, and it should subdivide as a clear three on the daily chart (the alternate below expects a five down).
My biggest problem with this wave count, and the reason I will retain the alternate, is the structure of intermediate wave (2) within primary wave 5. This is a rare running flat but the subdivisions don’t fit well. Minor wave C should be a five wave structure, but it looks like a clear three on the daily chart. If you’re going to label a running flat then it’s vital the subdivisions fit perfectly and this one does not. This problem is very significant and this is why I judge the two wave counts to be about even in probability.
Intermediate wave (5) looks like a zigzag rather than an impulse, and has a corrective wave count. This is also a problem I have with this wave count.
Intermediate wave (2) is most likely to subdivide as a zigzag, which subdivides 5-3-5 at minor degree. When this 5-3-5 is complete then how high the following movement goes will tell us which wave count is correct.
It is possible to move the degree of labelling within intermediate wave (2) all up one degree and see it as over. However, I will not publish this idea as it is extremely unlikely and publishing it would give it undue weight. That would see intermediate wave (2) as far too brief and too shallow in comparison to intermediate wave (1). Only if price breaks above 1,251.97 will I seriously consider this idea.
Intermediate wave (2) of this new cycle degree trend may not move beyond the start of intermediate wave (1) below 1,131.09.
From January 23rd onwards, since the expected trend change, volume is highest on down days. This supports the idea that we may have seen a trend change and the trend is now down. Volume for Silver is even clearer, with a big spike on the down day there for 29th January.
The two hourly charts below work for both daily wave counts.
Main Hourly Wave Count
Further downwards movement confirms that the last correction must be a rare triple zigzag, which is the only structure now that fits and has the right look.
The only difference today between this main wave count and the alternate hourly below is the degree of labelling within minute wave iv. This main wave count sees minute wave iv as still an incomplete expanded flat correction.
Expanded flats are very common structures. They subdivide 3-3-5. The maximum common length of a B wave within a flat is 138%, but I have seen plenty of flats with B waves longer than this. There is no rule stating a maximum length for a B wave with a flat, but convention states when it becomes twice as long as the A wave the wave count should be discarded because the probability is extremely low. Minuette wave (b) would reach twice the length of minuette wave (a) at 1,196.24, and if we see a new low below this point I will discard this wave count in favour of the alternate below.
Minute wave iv has now lasted four days. It is possible that it may be able to complete in just one more day to total a Fibonacci five, or it may not exhibit a Fibonacci duration. The most likely point for minute wave iv to end is still the 0.382 Fibonacci ratio of minute wave iii at 1,243.
Minute wave ii was a deep double zigzag correction. In order to exhibit alternation it is still more likely that minute wave iv is an incomplete flat correction which should be shallow.
Minute wave iv may not move into minute wave i price territory above 1,251.97.
Alternate Hourly Wave Count
This alternate will become the only hourly wave count if price breaks below 1,196.24.
This wave count sees minute wave ii as a deep double zigzag lasting a Fibonacci three days, and minute wave iv a shallow triple zigzag also lasting a Fibonacci three days. There is alternation in depth but little alternation in structure. This is unusual for Gold.
Because minute wave iii was shorter than 1.618 the length of minute wave i, and because there is no Fibonacci ratio between minute waves i and iii, I would expect to see a Fibonacci ratio between minute wave v and either of i or iii, and I would expect minute wave v to be a long strong fifth wave typical of commodities. At 1,146 minute wave v would reach 1.618 the length of minute wave i. Minute wave v may break below support at the lower edge of the pink channel about minor wave 1 or A.
Alternate Daily Wave Count
The maroon channel about cycle wave a from the weekly chart is now breached by a few daily candlesticks, and is now also breached on the weekly chart by one weekly candlestick. If cycle wave a is incomplete this channel should not be breached. For this reason and this reason only this wave count, despite having the best fit in terms of subdivisions, only has an even probability with the main wave count. It will prove itself if we see a clear five down with increasing momentum on the daily chart.
Draw the maroon trend line on a weekly chart on a semi-log scale, and copy it over to a daily chart also on a semi-log scale (see this analysis for a weekly chart).
Within primary wave 5 intermediate wave (1) fits perfectly as an impulse. There is perfect alternation within intermediate wave (1): minor wave 2 is a deep zigzag lasting a Fibonacci five days and minor wave 4 is a shallow triangle lasting a Fibonacci eight days, 1.618 the duration of minor wave 2. Minor wave 3 is 9.65 longer than 1.618 the length of minor wave 1, and minor wave 5 is just 0.51 short of 0.618 the length of minor wave 1.
Intermediate wave (2) is an expanded flat correction. Minor wave C is a complete expanding ending diagonal. Expanded flats are very common structures and ending diagonals are more common than leading diagonals.
This wave count has more common structures than the main wave count, and it has a better fit.
For this alternate wave count the diagonal is an ending diagonal for minor wave C. Within an ending diagonal all the sub waves must subdivide as zigzags. The fourth wave should overlap first wave price territory. The rule for the end of a fourth wave of a diagonal is it may not move beyond the end of the second wave.
Although Gold almost always adheres perfectly to trend channels, almost always is not the same as always. This wave count is still entirely possible. The trend channel breach is a strong warning that this wave count may be wrong and we need to heed that warning with caution at this stage. For this wave count once minor waves 1 and 2 are complete (which would be labelled minor waves A and B for the main wave count) minor wave 3 downwards should be very strong and extended, and would probably take price below 1,131.09.
A new low below 1,131.09 would confirm that a third wave down is underway.
At 956.97 primary wave 5 would reach equality in length with primary wave 1.
Intermediate wave (2) may not move beyond the start of intermediate wave (1) above 1,345.22. If this invalidation point is passed this wave count would be fully invalidated.
The short to mid term outlook for both wave counts is identical. The structure and labelling on the hourly chart is identical, so I will publish only the one hourly chart while the wave counts do not diverge.
This analysis is published about 02:10 p.m. EST.