TA question often asked... "Do trendlines work"?

[longandshort]

@themickey

Ugh this is a microcap AND complex share structure (ADR, aussie, etc.). Not a fan of analyzing these types but here's how I'd approach it:

Step 1: understand the biz. Read the MRY 10-k to learn this.

Key items:

• Company is in "development" so they don't have any actual sales as of their MRFY and don't expect to generate revenue until 3FQ2023. It'll have additional product starting in late 2024-2025, and will be able to produce lithium hydroxide in 2025-2026.
• The company will generate revenues through a 25% stake in SYA (Quebec based lithium projects), 9.4% ownership of Atlantis Lithium (ASX: ALL) Ghana lithium portfolio (including off take of ~50% of life of mine production at market prices), two lithium hydroxide plants (Tennessee and Carolina).
• Note: offtake price is $500-900/ton.

Step 2: How are they doing and what's guidance?

• In Q3 they made their first customer shipment and generated sales of $47M on sales of 29k dmt of lithium concentrate; on track to deliver full year guidance of 56.5k dmt
• results were impacted materially by 45% decline in spot lithium prices
• Company is aussie based so only reports 2x year

Here's guidance:

Step 3: Figure out the critical factors

• Performance is on-track given what they guided, but the price of lithium has fallen
• There's a bidding war over ALL mines which may impact PLLs position given its off-take deal
• The stock roughly tracks lithium prices

Step 4: Examine estimates for the range of outlooks
From bloomberg -- here's the trajectory on growth, given the pace of projects and when they'll add to revenues:

• Evercore ISI sees price of $65 on 6x FY25 EBITDA
• DA Davidson sees $60 on 9.4x FY25 EBITDA
• Other analysts are mixing EV/EBITDA with DCF and other methods (BTIG references 2x FY24 EBITDA)
• Range of estimates for FY25 EBITDA is 96 - 484, with consensus at 204; note: revisions have moved lower since august

Step 5 -- Valuation
Looking at EV/ FY 25 EBITDA .. can see that it has ranged from 1.3x to 2.5x in the past 6 months, with an average this year of ~1.8x. It's currently at ~2x.

Step 6 -- What's your View?
Scenario 1: On track and improving macro dynamics
* Stock grows faster than consensus estimates on the back of improving lithium prices in late 24-2025. FY 25 EBITDA of ~300MM.
* Worth about ~$46 on 2.5x FY25 EBITDA -- company needs to stay on track AND lithium prices improve.

Scenario 2: Project headwinds and weak macro
* Company remains on track but lithium prices remain subdued, plus capex and project drift rises as Ten + Car plants build out. FY 25 EBITDA of 185MM, about -7% below current consensus.
* Stock is worth about ~$23 on 2x FY25 EBITDA.

Scenario 3: Lithium prices do not recover as global supply picks up
* Company remains on track but lithium prices fall to the ~200 level (BBG index) which is -50% where spot is. FY 2025 EBITDA ~120MM.
* Stock is worth about ~13.5 on 1.8x FY 2025 EBITDA

Your prob. of each scenario is:
scen 1: 60% (you are bullish and think lithium will rebound)
Scen 2:10% (you don't think this has a high chance to occur)
Sen 3: 30% (you think there is a real risk lithium prices move lower, not because of a crash in demand, but a pickup in supply)
Your prob. adjusted price is: ~$34. Current stock price is ~$28, so in this example, you would see the stock as currently undervalued.

In this example, because you are bullish on the stock, you expect the price of lithium to rise and that would lift estimates higher (consensus FY25 EBITDA would move from ~200mm to ~300mm). You can see this in "real time" with estimate revisions. If the stock begins to rally w/o estimate revisions or higher lithium prices, then it's EV/EBITDA would be bid higher. In an environment where EV/EBITDA goes up to 2.5x without (1) higher lithium prices or (2) +ve estimate revisions, then you would fade the move. Same thing for if the stock extends below 1.8x.

(Stock price = white line, lithium price = blue line)

How would you trade it?
Well, the signal needs to be worth at least 2x the volatility of the asset. Historical vol of PLL is ~60% on a 252d basis. The next earnings will be in Feb 2024. If I thought that lithium was going to start going up by the end of this year and will "show up" in Febs guidance, then my $46 px target is in play and the ~64% expected return (from $28 to 46) is over ~120% annualized which is 2x the historical vol and therefore in-line with a "good trade".

 

[NoahA]

Ugh this is a microcap AND complex share structure (ADR, aussie, etc.). Not a fan of analyzing these types but here's how I'd approach it:

Wow... I can't speak to the viability of such an approach since its not what I do, but you get an A++ for your detailed example. I have honestly never seen such a comprehensive reply here at ET with regards to how they approach a trade. (Ok... Al Brooks may be even more comprehensive in his books, but he would never actually call out a trade)... LOL

If I thought that lithium was going to start going up by the end of this year

This is the only part that I question. It seems like your entire analysis is dependant on the price of lithium going forward. I really liked how you gave weights to all the different scenarios, but now we are at a crossroad. How do you decide if you think lithium will go up? This decision still needs to be made since it seems like this is the trigger for the trade.

Its really no different than a guy using a trend line. He maybe waits for it to break, and then sees where it closes. Then he maybe waits to see if there is a recovery the next day. If there is a pause, he might choose to hold, but if there is follow through on the downside, this might be the reason to sell. Its kind of elegant in that there isn't much of a decision that needs to be made because the price action will dictate the action. But in your example, you need to figure out if lithium will go up, so how do you do that? And what if the lithium price is slow to react? What if the price of the stock starts going higher before lithium goes higher?

Just to illustrate my example (sorry to hijack the thread @deaddog ), but here we see a support like broken hard. (a horizontal line is kind of a trendline, but even better in my opinion than a diagonal trendline since its at a specific price level that everyone sees) We see the clean break, and a close on the low, but wait. The next day is a solid green bar, not as big at the red bar that broke the support, but still, an attempt to regain support, so maybe you hold.

 

[longandshort]

This is the only part that I question. It seems like your entire analysis is dependant on the price of lithium going forward. I really liked how you gave weights to all the different scenarios, but now we are at a crossroad. How do you decide if you think lithium will go up? This decision still needs to be made since it seems like this is the trigger for the trade.

View attachment 328091

This is a great question and you're on the right path (but using trendlines is still garbage). Generally speaking, as a trader I want to focus on areas (companies and sectors) I can develop an edge in. If I can't develop an edge then I need a great source of info. Personally, I do not believe in using macro to guide an individual equity trade. I would rather trade long PLL vs. short ALB or something along those lines to capture the idiosyncratic returns of PLLs performance (if I thought it was tradeable).

In this example, the analysis is contingent upon a view of lithium prices and that's what needs to be resolved first.

This is a Wood Mackenzie analysis of supply & demand outlook (based upon actual mine production and stated EV and battery demand outlooks):

So it looks like we won't see lithium prices stabilize until 2025 and move higher starting in ~2026 (note: market always looks ahead, so would expect the moves in price to happy 6-12 months before). If we take this as our base case scenario, we can update our estimates of EBITDA. If Lithium prices keep going down through 2025, our EBITDA outlook is at risk right? So in this case, we then overweight scenario 3 and trade that.

 

[themickey]

If I were to reply to @longandshort's reasonings in any detail, I would probably need to write several pages as I'm pretty sure it will blow up into an argument if I left out any detail.
Anyhow, thankyou for the detailed reply.

But briefly, there are two groups, technicians and fundamentalist. The stronger of the two groups will prevail imo. The stronger having the most financial firepower.

PLL or any stock will rise and fall on market sentiment.
~95% of stocks will rise on bullish sentiment and ~5% will fall (the outliers) and vice versa bearish mkt.

Stock rise and fall for numerous reasons besides mkt sentiment, stocks can plummet on good news and rise on bad news, but most follow a herding pattern.
Fundamentals for the most part on stocks are inaccurate, you're told what the directors want to tell, half the story is told, the other half is hidden / obscured.

Going into fundamental analysis to determine stock price is no different than a technician analysing a stock using scores of indicators and lines drawn all over the chart, imo it's bullshit.

A trader wants to trade, it's very simple. Buy when the market's (sectors / Indexes) are bullish and sell when it isn't.

How for example does one trade Bitcoin using fundamentals?
You can't.
But Bitcoin is still relatively easy to trade using technicals.
But not all technicals are created equal and that's what traders need to sort out.

But there is also another consideration, every trader has a different gift.
Some can trade fundamentals while others can't,
Some can trade trend lines while others can't....

 

[Laissez Faire]

Do trendlines work?
Do S-R lines work?
Do MA crosses work?
Do Overbought / Oversold things work?
...

ANSWER :
Sometimes YES
Sometimes NO


WHY ?
The market moves in various
fashion, manner, ways, speed, pattern, characteristics, behaviour .....

Exactly. 100 % spot on.

The OP comes across as a newbie.

First of all, the question asked is "Does trendline work?"

My counter-question: "Work for what, exactly?"

And if it works for that, how often does it work? And how does it work when it actually works? The OP suggests to 'play it by ear'.

The OP is posting ONE single cherry picked example of price history. More like fooled by randomness.

No statistics. No rules. No hard data.

Even on the chart he posted there's plenty of occurences where a so called trend line failed.

Bottom line: Pretty much useless.

 

[themickey]

The stronger having the most financial firepower.

Imo, technicians, short term (intraday intraweek) using algos have the greatest firepower and continually win most of the battles.

 

[maxinger]

Exactly. 100 % spot on.

The OP comes across as a newbie.

First of all, the question asked is "Does trendline work?"

My counter-question: "Work for what, exactly?"

And if it works for that, how often does it work? And how does it work when it actually works? The OP suggests to 'play it by ear'.

The OP is posting ONE single cherry picked example of price history. More like fooled by randomness.

No statistics. No rules. No hard data.

Even on the chart he posted there's plenty of occurences where a so called trend line failed.

Bottom line: Pretty much useless.

After I wrote that, he blocked me.

 

[themickey]

Imo, technicians, short term (intraday intraweek) using algos have the greatest firepower and continually win most of the battles.

Just my opinion again, sloping trendlines are not a major factor in algo trading. They are difficult to code, they rely on different time frames, a bit like MA's.
There's a large combination of other factors besides sloping trendlines one can consider and easier.

 

[Scataphagos]

Continuing with the chart from yesterday.....

This 2nd chart is a zoom-in of the quick flips @ ~4554.. a Bull Pivot.

Such setups occur on all time frames of chart and they always mean the same thing... though not with the same "degree" of course. Then again, you never know when some setup "feels like it ain't much" and it turns out it was the best shot at the play going forward.

 

[Scataphagos]

Just my opinion again, sloping trendlines are not a major factor in algo trading. They are difficult to code, they rely on different time frames, a bit like MA's.
There's a large combination of other factors besides sloping trendlines one can consider and easier.

IMV it doesn't seem that algos can work off of Price AT. For a major reason... you need a "point/price" as a starting point for calculation. But unfortunately, the starting points can occur anywhere on the chart and can't be known in advance. Fortunately the "starting points" can be easily seen on the chart as or just shortly after they occur with Price TA.