I have spent a good deal of this week wondering whether we were bouncing out of a 40-week or 18-month cycle trough, and John’s post about the outstanding 9-year FLD projection and the discussion surrounding that prompted me to sit down and seriously consider the alternatives.
First here is the analysis that I have been watching which has last week’s trough as one of 40-week magnitude:
What I like about this analysis is the regularity of the 18-month cycle troughs, and the fact that most of the shorter cycles phase fairly well (in other words they have troughs on suitably prominent troughs in price). What I don’t like about it is the 54-month cycle trough in November of 2012, and the fact that the recent 40-week cycle trough is so much more prominent than the 18-month cycle trough in February of this year.
Of course the latter objection might prove to be of no importance IF the market forms a much deeper trough near the middle of next year, and for that reason this is the more immediately bearish option.
Here is what this analysis looks like in the short term:
The important points to note here are:
- That the VTL that was crossed by price in early October was the 18-month VTL. This means that the 54-month cycle peak has probably formed, which is at odds with the outstanding 9-year FLD target that John discussed.
- Price has bounced back up to that VTL, as it very often does (as I mentioned in this post)
- Over the next two weeks price faces resistance from both the 18-month VTL and the 40-week FLD.
It is also very possible that we have just witnessed an 18-month cycle trough, as shown here:
There is only really one thing that I prefer about this analysis: that the February 2014 and recent October 2014 troughs have magnitudes that match the prominence of those troughs in price. Things I don’t like about it are the irregular cycle wavelengths with fairly wide ranging changes from one cycle to the next, and the 54-month cycle trough in June 2013 also troubles me.
Here is what this analysis implies in the short term:
And the important points here are:
- That the VTL that was crossed by price in early October was the 40-week VTL. This means that the 18-month cycle peak was confirmed, which might or might not be the 54-month cycle peak, and therefore allows for the possibility of another new high as suggested by the outstanding 9-year FLD target that John discussed.
- Price has bounced back up to that VTL, as it very often does (as I mentioned in this post), and so for immediate trading purposes there is very little difference.
- Over the next two weeks the market still faces fairly substantial resistance from both the 40-week VTL and the 40-week FLD.
I suppose that highlights one of the reasons that I enjoy trading using Hurst’s cyclic principles so much: this debate about the magnitude of the recent trough is a fairly academic one. When it comes down to actually making trading decisions they would have been very similar no matter which analysis one chose.
Hurst wrote that whereas different analysts would reach different analysis results, the end result of trading decisions would most often be the same. How I wish he could see the evidence of that in our discussions today!
I mentioned that the debate about the magnitude of the trough is a fairly academic one, but of course there are differences, which can be summarized as:
- If the recent trough was of 40-week magnitude that is more immediately bearish, because we do not expect a higher price than the high of late September, and we expect a much deeper (lower) trough in about 9 months time.
- On the other hand if the recent trough was of 18-month magnitude then the outlook is more bullish in the near term, and we would expect to see new highs (up to the target of 2150 as discussed by John). And the trough in 9 months time would possibly not be at a lower price, as the market will probably hold up for longer. However even if this is a more immediately bullish outlook we still expect the bearish move down into the next 54-month (and 9-year) cycle trough, it will just happen a little later.
As a matter of interest here is hybrid analysis that resolves some of those issues I brought up above, but it worsens others, as is so often the case!
I will leave you to ponder the differences. Let me know what trough you think we’re bouncing out of!
4 thoughts on “18-Month or 40-Week?”
Hi David,
I had a look at the $TSX this morning. One can clearly make the case for an 18 month low in the Toronto Stock Exchange. It is not at all obvious in the $SPX. But based on the principle of commonality, I would lean towards this month’s low as an 18 month for most markets. The $TSX will be weighed down by the energy and commodity sectors which look very weak.
cheers,
john
Hi John
Out of curiosity I ran a filter on the Toronto index extracting the 40 week wave using the same parameters I use on the S&P500. The phase response of the 40 week wave is a little different than the other waves and may not correspond to your expectations. There is some obvious frequency modulation for which I did not adjust but it is close enough for government work! The 10 week wave (not shown) separates beautifully and would be very easy to use as a trading cycle on the Toronto index.
Enjoy
Hi William,
Thank you for posting your work on the $TSX. I was really hoping you would do that. So if I read it correctingly, counting 40 week Hurst cycles per your bandpass filter, it does suggest the June 2012 low as the 4.5 year Hurst low (at something like 39 months – very short). This is a great help. I looked at the $TSX again today and this seems to be the most obvious place for it.
So now looking past the 4.5 year low, June 2013 is the next 18 month cycle low (at 12 months), and we have June 2014 as the last 18 month cycle low (at 12 months?). This does work for me.
I am not optimistic on the $TSX. I do not think it will make a new high due to the bearish behaviour in commodities since the summer highs.
The other interesting thing your work suggests is that in 2011, an 18 month cycle low was made just prior to the April 2011 high which then lead to a double top. This time around seems to be different. I am amazed that the cycles are running so short at 12-13 month periods (I count 11,13,15,12,and 12 months off the 2009 lows).
In 2008 the $TSX was the last man standing due to the bubble in oil/natgas. The $TSX made impressive new highs (July 2008) while US markets had struggled for months prior and had topped out in Oct. 2007. This time around it seems the US markets will be the last to top IMO. TWT.
Don’t get me wrong. Regardless of how this current rally ends in these markets, a great swing short is setting up that can be held into late next year.
cheers,
john
Hi John,
Glad I could help, a word of caution though.
It is not advisable to extrapolate the longer waves of Hurst’s simplified harmonic model from the output of a single bandpass filter. There are a couple of fundamental differences between the outputs of bandpass filters and the approach used by ST, namely the lows are not perfectly synchronized (mathematically impossible) at the most visually evident absolute lows and the waves do not always exhibit a simple harmonic relationship. The chart below is a good illustration. It is showing the 40 week and the 4 year waves. There are 18 oscillations of the 40 week wave for the 3 oscillations of the 4 year wave shown, a simple harmonic relationship, but the phasing might be a little different than expected. Fortunately these differences are usually very small and are immaterial at the trading level as David always points out.