Natural Gas

[Monte15]

Wanted to start a conversation on people's ideas on the current Natural Gas price - I have a small position which I have taken on the latest dip as looks very cheap at the moment but can't help but feel that whilst the arguement for the milder weather and lower levels of useage but with the current red sea oil situation I cant help but feel at some point we will need to stock pile - whats everyone else's feedback

 

[THT]

Wanted to start a conversation on people's ideas on the current Natural Gas price - I have a small position which I have taken on the latest dip as looks very cheap at the moment but can't help but feel that whilst the arguement for the milder weather and lower levels of useage but with the current red sea oil situation I cant help but feel at some point we will need to stock pile - whats everyone else's feedback

Its on my watch list - Notice the cycle L-L (The H-H cycle came bang in on time)

 

[THT]

What time interval do you consider the most effective for identifying cyclical turning points?

Hi MattK

For Natural Gas above its a 4 year (ish) cycle - MANY commodities have a 4 year cycle

BUT, and it deserves capitals - cycles aren't really static - if you imagine you could go inside the chart, then price is being twisted and moved by TIME, this is what creates the triangles in price action every now and then with "twisted" pivot points - so this makes it absolutely impossible to be able to nail down a static cycle exactly all the time

However, there's always something that is there or there abouts in any chart, you just have to work out the time periods - some of these will be Hurst cycles, others differing figures etc - You just need to try to work out what the time-frame is working to, so for example on a daily chart it might be working to a 20 day cycle etc

But at some point that cycle will "disappear"! Then reappear etc

If you research Hurst's work, he produced a list of numerous cycles that he found in over 1000 samples of stocks/commodities etc

To answer your question though:

Depends on the time-frame you're looking at and it will differ from market to market, stock to stock etc - the easiest thing to do as I did with the Nat Gas chart above was to just "eyeball" the Intervals and average out and not be too fixated on it being exact

The cycle low expected 2010 actually came in late 2009, nearly a year early, so you have to use some variance if that makes sense

 

[1nvest]

What time interval do you consider the most effective for identifying cyclical turning points?

to @THT's point about Hurst, this is a useful link and talks about hurst's "Nominal Model" where he identifies the core cycle lengths and their average wavelength duration (taken many years ago

Revisiting the Nominal Model. – Hurst Cycles

Just bear in mind where Hurst mentions days, these are calendar days not trading days
In my limited experience, i find the 40 day cycle most easily observed and its higher cycle lengths the 80day/20 week etc
The site refers to the tools that Hurst used such as the FLD-Future Line of demarcation and VTL - Valid Trend lines

 

[THT]

Thanks @1invest

Here's the 48 mth cycle placed into a composite Index (RED line is the comp Index / BLUE line is price) - As you can see of late its been accurate, in the 2000's it was out of whack - 2008 INVERTED into a high or came early, I prefer the Inversion as when these cycles Invert they act (price wise) as its done in the spike to a high then collapse

Ideally we'd like much more price data to snip out the longer term cycles - For trading purposes, we can move down to the WEEKLY time-frame and run cycles off that chart to find the weekly cycles that move the Monthly charts etc

Although I personally prefer to wait for bullish confirmation of a low around the expected low date, employing a strategy of just buying long around these low dates is fine, as long as the trader/Investor is happy to have some heat on further price falls before it turns

I tend to do a set and partially forget with these types of trades and buy into an ETF for my SIPP and its usually an ALL IN purchase once the bullish confirmation comes in and its a min doubling expectation - Once I've double checked the cycles valid etc - there's structure to the purchase that I've not mentioned, so its not a blindly buy in and hope

This works great with stocks too if a stock displays cyclic patterns on these long term cycles - you buy in near the long term cycle lows and quickly double your money, obviously its not guaranteed, but it happens a lot over the decades - when I say quickly it takes months, but in the grand scheme of things etc

I doubt most Investors/Traders look out decades to see if any patterns exist, because they believe that markets are random and what's the point in wasting an hour scrolling through monthly charts etc, but it is very much worth while

I'm writing a journal on the Journal forum and at some point this year I had planned to write a few posts on the subject - but before I do I'm laying the ground work from easy to complicated and we're not there yet

 

[THT]

Are there types of stocks that, in your opinion, are better suited to cyclical analysis?

Hi Mattk,

Commodities cycle best - Here's a 4 year H-H and L-L / you can work out the years of INVERSION - when it Inverts you'll be able to know as its happening and a trading strategy is to follow the Inversion and then reverse once it tops/bottoms out

For stocks - They all do, BUT, at differing stages and times - I personally just find that running through the monthly/weekly charts and EYEBALLING lows that look "regular" in time

Below is HSBC, which moved into a perfect cycle, but its now "shifted" away from perfection - it would be great if we could set and forget cycles, but it doesn't work quite that easy in the real world

I've found that stocks aren't static, they are more dynamic in terms of swings, but you can def find stocks that cycle from all sectors

You could look at weekly data and find a nice long bull/bear market, use that as a HALF cycle length and then when it reverses, locate the smaller cycles within that prev trend and see if they repeat on the opposite side up/down - All you need to do is find a few decent cycles per year on the weekly time-frame and you can catch some great reversals - on my Journal thread I'm just starting writing about cycles and plan to include some past examples on there

You need to find the right market though, you can't just pick a random stock and force the cycle onto - its much much easier to find the stock in cycle mode and then find the cycles its working to and trade those

I should update the comp Index for this one, but as it's diverged from the big cycles, I'd have to spend time that I don't have to find those smaller cycles etc

 

[zatara1]

When you do this cycle analysis do you use the front month? I ask because the contracts in different months are often priced with massive premiums or discounts already built in. Therefore when you call up a historic chart of 'front month' natural gas, the chart would join up each 'front month' and as there would be a jump with the roll you may end up with a cycle? Thats not particularly useful though if its the case as if you went to say trade a move higher when you roll your contract you would find the next contract along accounts for your expected profit. I may be wrong though and the cycles exist even without the effect of a disjointed futures curve.

 

[THT]

When you do this cycle analysis do you use the front month? I ask because the contracts in different months are often priced with massive premiums or discounts already built in. Therefore when you call up a historic chart of 'front month' natural gas, the chart would join up each 'front month' and as there would be a jump with the roll you may end up with a cycle? Thats not particularly useful though if its the case as if you went to say trade a move higher when you roll your contract you would find the next contract along accounts for your expected profit. I may be wrong though and the cycles exist even without the effect of a disjointed futures curve.

Hi Zatara1,

I just use the data provided by my charting software which is most traded contracts and is continuous - The Important thing with cycles is time, if you can pinpoint a time date, then you can look to trade/Invest in the most appropriate vehicle to meet the expectation and following price action and to do that the look back needs to be years of data, to get the patterns

If rolling contracts then you need to factor all that into the trade decision - I usually just buy the ETF for my pension fund and hold

 

[THT]

This one came in bang on as expected - it typically rallies for 2 years, then collapses back down for 2 years into the next 4 yr cycle low

so you can benefit from the uphill section and the downhill

and they [the printed "experts"] tell you, you can't TIME the markets!

EDIT - Remember, there's a WAR cycle in play - anything major happens and we'll have one hell of a spike up