OK ... grab a coffee or a bottle and take some time to go over the information being posted.
Read it carefully ... mostly because I may have made a mistake (if something doesn't look obvious ... call it out ... because if I can't explain it there may be an error). Please feel free to question the assumptions ... some of which may be aggressive or flat out unachievable. There may be other constraints that I haven't accounted for that maybe I should. If you think something is missing .... call it out and if I can add/change it I will try and do so.
The end answer first ... as it will be a long long post.
The correct way the read it is the
Intrinsic Value (IV) ranges between $9.79 - $12.94 with 95% confidenceWhat I've given you is the Average ($11.37) and Standard Deviation ($0.79) that is the result of a Monte Carlo simulation of 5,000 iterations (using a randomizer) of how the Book Value of the Equity in the business has grown (and discounted back to a present day value) for a 10 year Net Income Model using a standard normal distribution when the risk adjusted cost of capital is fixed at 10%.
That's a lot of mumbo jumbo, but the goal of the exercise is to use a defined probability distribution, do it 5,000 times and come up with an average.
Models are based on a series of constraints and assumptions. The FIRST assumption is of course the balance sheet as the model starts with that. With TODAY'S CAPITAL RAISING INCLUDED is looks as follows
Everyone should still be with me ...effectively just adding ~$35M of cash to CSE and then increasing share count. You see this here
Before I go any further, this starting point is as if we have jumped forward in time to June 30, 2021.
That is why is reads as $659.742M and not $59.742M
because why?
BIG ASSUMPTION-1:TLG has sold 20% of what I call "Stage-1" ... 10% to LKAB and 10% to Mitsui for $300M each.
Same (corrected) picture as before .. TLG ... green is good ... 80% of Stage-1
That's the easy part right.
Everyone agrees with me so far? For any company, we start off by putting in some equity cash ... which at that point BVE happens to equal the cash contributed and then we put the cash to work buying stuff, selling stuff, hopefully profitably (after paying off all creditors and paying taxes) and with what's left over we might choose to pay ourselves (a dividend) or just keep reinvesting into the business. Doesn't really matter what we do because if we contributed $100, bought stuff and sold it all for $200 then our BVE has gone from $100 to $200. If we only sold half of it ($50 bought and then sold for $100) our BVE of equity is $150 ($100 Cash and $50 inventory)
So it doesn't matter to me whether the $600M is cash or spent in building a mine/plant worth $600M. This BVE is the all important metric I'm tracking. It represent our (equity shareholder) actually cash invested into the business. If we do it wisely the business grows OVER TIME. Do it poorly and it shrinks (i.e. capital is destroyed).
BIG ASSUMPTION-2: Mitsui agrees to offtake of 40,000 tpa Talnode-C for 5 years AND makes payments of $300M on Jul 1, 2021, 2022, ... 2025... This is important as it prepays 2 years in advance when we have no production. They get a good deal as they pay a fixed price of $7,500/tonne for 5 years.
What this does is provide TLG which predictable net income (so actual earnings not EBITDA).
Now look closely under the column FY'221st thing to notice is the earnings of $226,200,001 which is in US$. All US$ converted to AUD$ at AUDUSA = 0.75
2nd thing is the Future BVE which is simply the BVE going into the model ($659M) + AUD Earnings ($302M) = $961M
This is good result for TLG as the ROE is at a very high 34%
Don't start "high 5'ing" just yet as that is just effectively the "current" year. What see down the column is the "cost of capital" of ~$66M ... this is reminding us that there in no free lunch. Having that $660M in CSE does nothing if you don't earn above your cost of capital (which we said was 10%). Deduct that from your AUD earnings and now you have only AUD $235M (so its $302M - $66M). But those earnings were over 1 year .... so we must then discount that back to the present time of Jun 30, 2021 ... so our residual earnings is just $214M
That is the essence of the residual earnings model
Now look to the right of the column FY'22 and you see "Mean" and "Std Deviation" - this is where the earnings estimates are going and for FY'22 and FY'23 it was simple ... we had revenue of $300M ...so pay taxes and that's what's left over
Now we are at FY'24 and things begin to take shape ... actual production starts Jul 1 2023 - Jun 30 2024In the residual earnings model it looks like this
But where did the Mean and SD numbers come from? And why have they dropped?Sitting behind the Residual Earnings model are 10 years of Net Income Monte Carlo models. Same principle of 5,000 iterations to simulate what "might" happen based on assumptions.
This is what the FY'24 Net Income model looks like
Costs first
The above is looking at ASSUMED COSTS
1. TLG G&A costs ... which started at $10M in FY21 and escalates at 5% pa
2. Assumed no debt on TLG balance sheet
3. Assumed Cash cost ... this again uses a normal distribution that is based on the PFS cash cost avg of $2,380 which is the escalated by 2% each year and with standard deviation of 5%.
This is 1 iteration of 5,0004. Total tonnes produced ... this is at 50% of 104,000 tonnes and 80% of that with is TLG. No Talphene
Are we OK on those cost assumptions?
Moving to the sales revenue side
Throughout the model (i.e. for 10 years) these assumptions are made every year1. Talphene price is $15,000/tonne
2. Assuming a ramp of just 25% ... so 2,125 tonnes produced ... out of the target 8,500
3. Talnode-C price is "randomized" along a Trigen distribution where prices range from a low of $7,500 to a high of $11,250 with the mode (half the time above and half the time below) being $9,375. The 0.9 means basically that there is only 10% uncertainty (i,e. possible that 10% of the time it will be outside of those ranges)
What you are seeing IS 1 ITERATION OF 5,000 ... in this iteration Talnode-C sold for $9,673/tonne4. TLG has rights to 41,600 tonnes of Talnode-C ... of which we already sold 40,000 to Mitsui (the prepay column ... now year 3 and this is Mitsui 1st delivery from TLG).... which leaves us with the remaining 1,600 to sell.
5. Allowing for Interest Payment of $20M (cost to TLG's 80% ownership) ... assumed we've issued a ~3% Bond of ~$600M
6. Pay tax and arrive at our Net Income for this iteration of $160M ... which happens to be above average.
7. For FY'24 the overall avg & std dev from those 5,000 iterations is $157M and $4.47M which goes into the Residual Earnings model shown earlier.
By now
@pabs is on his 2nd bottle (1L) of vodka and the screen is blurry
Hoping everyone still following. I wont go into further details as I'm sure its becoming eye glazing but
FY'25 I ramp up Talnode-C to full 104,000 tpa (with TLG getting 80%) and Talphene at 4,250
FY'26 Talphene is now 8,500
and on we go accumulating cash
But the RE model is unforgiving. The more earnings we have the more "CSE" has grown (assuming we aren't destroying capital in failing projects) which means our cost of capital keeps rising (in $$$).
So I did say this is stage one ... sometime in FY25 (or before) I would suspect that cash pile is put to good use with Stage 2. While I did the model through to FY 2030 ... that's eternity in the tech business. Batteries may be obsolete by then, who knows.
IF THE ASSUMPTIONS are reasonable and valid, the project (Vittangi + Niska) has an IV far above the current share price. I would stress again and again that IV is NOT a share price prediction. The Market Price (MP) of a stock is dictated by the "Speculative Price" (either +ve or -ve) to the assumptions modeled and risk adjusted.
Hope you all enjoyed the read.
(20min delay)