Upwards movement to a short term target for Friday was expected. The target at 1,164 – 1,168 was met and slightly exceeded by 0.41.
I have a new wave count. Volume analysis favours it, and some problems with the prior Elliott wave count are resolved.
Summary: Short term some downwards movement is expected from here. The new wave count expects it to be very choppy and overlapping for a minor degree fourth wave correction which may not move below 1,119.23. The old wave count, which is now an alternate, would be confirmed if price breaks below this point. At that stage, a third wave down would be confirmed. Now, while price remains above 1,119.23, we may be in a bull market to continue for a further 10 weeks up to 1,400.
To see weekly charts and analysis click here.
Changes to last analysis are bold.
MAIN ELLIOTT WAVE COUNT
The bigger picture at super cycle degree is still bearish. A large zigzag is unfolding downwards. Along the way down, within the zigzag, cycle wave b must unfold as a corrective structure. This wave count now sees cycle wave b ending with primary wave C up. Cycle wave b is seen as a big expanded flat correction.
This wave count resolves a few problems I had with the prior main wave count, and is better supported by regular technical analysis at this stage. For these reasons, it must be my main wave count.
Ratios within cycle wave a are: primary wave 3 is 12.54 short of 1.618 the length of primary wave 1 and primary wave 5 is 13.85 longer than 0.382 the length of primary wave 1. There is perfect alternation between the deep flat of primary wave 2 and the shallow triangle of primary wave 4. Primary waves 2 and 4 are both very close to a Fibonacci 55 week duration, giving this part of the wave count the right look.
Cycle wave b is an expanded flat. Within cycle wave b, primary wave A also subdivides as an expanded flat. Primary wave B is 1.89 times the length of primary wave A, longer than the common length up to 1.38 but within the allowable convention of up to twice the length of primary wave A. This deep B wave is the only part of this wave count which is not absolutely typical. It breaks no Elliott wave rule; I have seen plenty of expanded flats with B waves this deep or even deeper, but it is not the most common look. It still looks reasonable.
At 1,400 primary wave C would reach 2.618 the length of primary wave A.
The alternate would be invalidated with a new high above *edit: 1,232.49, which would provide full confidence in this wave count.
Because cycle wave a is a five wave structure, cycle wave b may not move beyond its start above 1,912.57.
RSI shows some long held bullish divergence going back to April 2013. This strongly supports this wave count.
Cycle degree waves should take one to several years to unfold. At its end, cycle wave b would probably have lasted just over one year; cycle wave b may complete in a total Fibonacci 55 weeks. Thus far cycle wave b has just ended its 45th week, so it may now be expected to continue for a further ten weeks, give or take up to three either side of this expectation.
The daily chart focusses on all of primary wave B.
Primary wave B subdivides perfectly as a zigzag. Draw a channel about it using Elliott’s technique for a correction. When this blue channel is breached, then that shall provide trend channel confirmation that primary wave B is over and primary wave C upwards has begun. The channel may provide resistance before that happens. After the channel is breached, then a throwback to test it for support may also be expected as very likely.
Primary wave C must subdivide as a five wave structure, either an impulse or an ending diagonal. An impulse is more likely. When the first wave up is complete, then we shall know which structure primary wave C is unfolding as because an impulse requires a five up and an ending diagonal requires a zigzag up for intermediate wave (1).
Within intermediate wave (1), minor wave 4 may not move into minor wave 1 price territory below 1,119.23 (the high of minor wave 1 is taken from the hourly chart).
If primary wave C is to be an ending diagonal, then this upwards movement looks too short to be intermediate wave (1) in its entirety; it would be only minor wave A, which must be a five wave structure. The invalidation point for both ideas, at this stage, should be at the same place at 1,119.23.
Price is finding resistance at the upper edge of the blue channel copied over here from the daily chart. This may be where minor wave 3 ends and minor wave 4 begins.
Minor wave 3 is 4.42 short of 2.618 the length of minor wave 1.
Within minor wave 3, there are no adequate Fibonacci ratios between minute waves i, iii and v.
There are 23 possible corrective structures that minor wave 4 may unfold as. Although I am labelling minor wave 4 as a possible expanded flat correction, it does not have to unfold like this. Minor wave 4 may be a triangle, combination or multiple flat correction to exhibit structural alternation with the double zigzag of minor wave 2.
Minor wave 2 was very deep at 0.97 of minor wave 1. Minor wave 4 should be shallow, reaching down to either the 0.236 or 0.382 Fibonacci ratios. Minor wave 2 lasted a Fibonacci three days. Minor wave 4 may be 1.618 or 2.618 the duration lasting a Fibonacci 5 or 8 days in total.
Minor wave 4 is likely to be very choppy and overlapping; it may move sideways or have a slight downwards slope. While minor wave 4 unfolds, the wave count within it will change as its structure becomes clearer. Analysis of minor wave 4 should focus on seeing when it is over. There may be a breach of the pink channel copied over from the daily chart.
If minor wave 4 is an expanded flat or running triangle, then it may include a new high above 1,168.41. If it does that, then the alternate below would further reduce in probability.
The best fit channel about this last fifth wave up, here labelled minute wave v, is breached and now provides resistance. There is more than one way to draw a best fit channel, and the hourly charts show two possibilities. Here, use the lower orange line for resistance.
ALTERNATE ELLIOTT WAVE COUNT
This was yesterday’s new main wave count, but today this count is relegated to an alternate.
This wave count is now an alternate, because it has the following problems which the new wave count resolves neatly:
1. Intermediate wave (2) is much bigger in size than an intermediate correction within Gold’s impulses normally are.
2. The maroon channel has been breached on the daily chart twice. This is an early indication that cycle wave a may be over.
3. Minor wave 2 is much longer in duration than a minor degree correction within Gold’s impulses normally are.
4. Volume no longer supports this wave count.
5. Gold often exhibits very short fifth waves to follow its fourth wave triangles.
A new low below 1,119.23 in the short term would confirm this wave count. Full and final confirmation would come with a new low below 1,072.09.
If primary wave 5 reaches equality with primary wave 1, then it would end at 957.
This wave count now sees a series of three overlapping first and second waves: intermediate waves (1) and (2), minor waves 1 and 2, and now minute waves i and ii.
The blue channel is drawn in the same way on both wave counts. The upper edge will be critical. Both wave counts expect some downwards movement from here to bounce down from resistance about the upper blue trend line. Here the blue channel is a base channel drawn about minor waves 1 and 2. A lower degree second wave correction for minute wave ii should not breach a base channel drawn about a first and second wave one or more degrees higher. If this blue line is breached by one full daily candlestick above it and not touching it, then this alternate wave count will substantially reduce in probability.
Minute wave ii may not move beyond the start of minute wave i above 1,232.49.
Today, at the hourly chart level, this alternate wave count has another problem. Downwards movement during Friday’s session looks best as a three wave structure. While it may be a zigzag within a leading diagonal for a first wave down, leading diagonals are not nearly as common as impulses for first waves.
This sideways movement for Friday looks too big to be a micro degree correction within subminuette wave v.
So far the channel drawn here about subminuette wave v is providing support. When it is breached on Monday, then it would provide trend channel confirmation that minute wave ii should be over. This channel drawn here as a best fit about subminuette wave v is more conservative than the best fit channel drawn on the main hourly chart. For confirmation of a minute degree trend change the channel should be as conservative as possible.
Further confirmation is absolutely required for confidence in this wave count. A clear five down on the hourly chart followed by a new low below 1,119.23 is required.
Only when this wave count is confirmed would I calculate a target for minute wave iii. Minute wave iii would be expected to show a strong increase in downwards momentum.
Subminuette wave ii may not move beyond the start of submineutte wave i above 1,168.41. A new high above this point would indicate that minute wave ii may be continuing higher, but it should still find resistance at the blue channel on the daily chart. If that channel is breached, then the probability of this alternate would be substantially reduced.
Click chart to enlarge. Chart courtesy of StockCharts.com.
Weekly Chart: This week has ended with slightly lower volume. A small decline in volume for another upwards week indicates at least a near term correction; the rise in price was not fully supported by volume this week.
On Balance Volume has has plenty of room to move higher. When it touches the purple trend line, then that may provide resistance for a larger multi-week downwards movement.
RSI is within normal. There is plenty of room for price to rise or fall.
Click chart to enlarge. Chart courtesy of StockCharts.com.
Daily Chart: Volume for Friday shows a third upwards day on rising volume. Friday’s volume is higher than 11th August. This rise in price is now supported by volume.
ADX has finally turned up: above 30 and pointing upwards. The +DX line is above the -DX line indicating a new upwards trend. ADX does tend to be a lagging indicator, but when it shows a trend we must take notice. This supports the main wave count.
On Balance Volume did not find resistance at its pink trend line, which was broken through. This is further bullish indication today which supports the main wave count.
RSI is close to overbought. There is still room for price to rise.
While price made new highs for Friday Stochastics turned lower. Near term Stochastics is expecting a correction against the trend indicated by ADX to resolve its overbought condition. The divergence between price and Stochastics for Friday is near term bearish. This supports the main wave count, which expects a correction agains the trend to unfold from here.
Friday’s candlestick has a longer upper wick than lower, indicating the bulls may be at least temporarily exhausted. Bulls did not manage to keep price close to Friday’s highs; the bears managed to push it lower than the bulls could hold it.
This regular technical analysis overall supports the new main wave count, and plays a large reason for my change of wave counts today.
Finally, an additional note to answer the question I am sure will be asked (and has in fact been asked once already): What could I have done differently to come up with this new wave count sooner?
Firstly, it must be admitted that although I expected this upwards movement was a correction, I have for the last three days correctly predicted direction and provided targets each day which were met. That’s not perfect by any means, but direction is important. We have not had confirmation upwards movement is over for the alternate wave count.
Up to 11th August volume data from FXCM clearly showed the strongest day in the sideways chop was a downwards day (look at the TA chart at the bottom of that analysis). I rely on volume data heavily to indicate the direction of a breakout, but data changed on 12th August, so I learned that FXCM volume data is unreliable, so have changed to using StockCharts data. That data change was a major change from bearish to bullish.
So what could have been done differently? I needed to be more flexible with where cycle wave a may have ended, remembering that cycle wave b should be choppy and overlapping and may include a new low below the end of cycle wave a. I needed to take more time to find an alternate at the weekly chart level once the maroon trend line was breached. I have learned from this exercise that trend channels should be given more weight plus volume data must be reliable.
If I’m right, then we still have reasonably early indication of a new upwards trend. My job now will be to find when corrections within the trend are over to pick the next strong impulsive movement.
This analysis is published about 07:17 p.m. EST.