Forecasting Trends: USDJPY – Long Term Cycles

Hey traders!

Another week over…

Just look at the current markets… recently I was talking about ‘double tops and bottoms’…

This week the AUDUSD briefly broke through its old major high made in July 2008, showing a glimmer of hope for the bulls, but by end of week trading it didnt manage to ‘close’ above it… so a reversal could be imminent.

But before we jump the gun we need some sort of confirmation… next we look for a reversal bar to confirm the resistance, like that perfect ‘doji’ candle prior to the market reversing at the July 2008 high; a classic reversal bar forcing the market down.

The USDCHF is nudging a previous major ‘low’… during the week it also broke through the old low price, but finally closed almost smack on that previous low… the bulls may come in soon to turn this market around.

Getting back to our USDJPY Cycle analysis, last post we looked at how to identify the major lows on a monthly chart. We always need a minimum of 4 cycle lengths, as this gives a higher degree of precision and this depends on how much data you have available, you may find you need to start on the weekly chart if you cant get 4 or more cycle lengths (low point to low point).

Now you can also see in the original USDJPY monthly chart that those major lows are not exactly the same length.

  • Cycle 1 (A-B)            = 54
  • Cycle 2 (B-C)            = 55
  • Cycle 3 (C-D)            = 52
  • Cycle 4 (D-E)            = 57

To find the monthly cycle we need to find the ‘average length’ of the valid cycles we have available. (When I say ‘valid’ Im talking about ensuring that they all adhere to our ‘Check List’, which you can find I summarize at the bottom of each post).

So to find the ‘average cycle length’ we add all the cycles together and then divide by the number of cycles, where we get; 54.5… then round the number down or up, and here we get a final figure of 55.

Next we need to check if all those cycle lengths fit our next criteria of ‘a minimum 80% fit’.

As you can already see the lows never fall exactly 100% on time, there is a variation, that is why we have a buffer called a ‘cycle zone’ where the market makes a high probability turn.  In W.D.Gann’s work he uses a buffer of 16.67%.

So to find our cycle zone we need to add and subtract 16.67% from the cycle average figure we found earlier.

  • 55 plus 16.67% = 64
  • 55 minus 16.67% = 46

Therefore our cycle zone is between 46 – 64 months.

Check back to see if all our monthly cycles fit within that zone… I have drawn the zones on the chart below.

To forcast future cycles we need our historical cycles to ‘fit’ within our cycle zone 80-100% of the time, where the market makes a low at some stage within the outer time frames of the zone buffer.  Looking at our monthly chart we have a 100% hit rate, which is typical for the monthly chart. 

Ok, now I can already tell you are thinking that somewhere between 46 – 64 months is a long time… where is the accuracy in that? In a time frame that large of course a large low is likely to occur!!

But remember we start with the monthly, then go down to the weekly and then daily charts… where the accuracy is found.

Soon you will see how all this ties together, and how we project the cycles forward… but more on that next post.

So to rehash our cycles, this is what we need;

Check List

  • Start on the Monthly chart (larger cycles dominate smaller cycles).
  • Identify all significant/major ‘Lows’ in a market.
  • Rising markets = lowest point must be cycle start.
  • Falling markets = lowest point must be cycle finish.
  • Minimum 4 cycle lengths.
  • 80% + accuracy fit within cycle zone (plus/minus 16.67%).

Enjoy the weekend!

Leave a comment

Your email address will not be published. Required fields are marked *


You might be interested: