It doesn't help that one of the main lithium consultancy firms...

It doesn't help that one of the main lithium consultancy firms (BMI) appears to be confirming they don't actually know quite what's happening in the lithium supply/demand space but as experts they are feeding their incorrect information to clients like PLS who then represent it in their presentations. This circulates through the rest of the lithium system.

Basically there was close to balanced supply and demand in 2016 and 2017 at small scale.
BMI models a surplus of supply in 2018 by ~50kt LCE. I'll accept that, new operations started and Lithium Carbonate (LC) fell from 165k to 80k during the year.

BMI models another supply surplus in 2019 by ~60kt LCE so the cumulative surplus is 110kt LCE. By the end of 2019 LC was 50k so its still plausible but doubling the size of the surplus overhang for a fall from 80k to 50k seems unlikely. A bigger price fall would have occurred with that sized surplus.

BMI models another surplus in 2020 by ~30kt LCE so the cumulative surplus is 140kt LCE. By the end of 2020 LC was 47k so this increased deficit is probably wrong and the market was much closer to balanced for supply/demand than they have modelled. If the overhang was really 140kt LCE the price fall would have been larger. The LC price was increasing in the last month of 2020 so by that stage the cumulative surplus is decreasing not increasing.

BMI models a deficit in 2021 of ~40 LCE so the cumulative surplus is back to 100kt LCE. By the end of 2021 LC was storming to just below 300k CYN. I'd suggest there's no chance of price action like that if there was still a cumulative surplus of supply. If anything there is a cumulative deficit as the supply chain scrunches up shorter than it would like to be so as to provide a little bit of extra demand reduction. The model has a cumulative error of 100k LCE by this point, possibly a little more.

BMI models a balanced supply in 2022 so there's still a cumulative surplus of ~100kt LCE. This is again wrong because by the end of 2022 LC had stormed even higher and was over 500k/t CYN. Prices are at all-time records. Everyone's screaming there's no supply around. One of the sources of supply will be companies operating inventories at or below just-in-time and not where they want for a critical supply. I'd suggest the true position at this stage is customers operating perhaps 50k+ below where they want to be. The cumulative model error is perhaps 150k LCE.

BMI models a 2023 surplus of ~100kt LCE so their cumulative model is now ~200kt LCE sloshing around as surplus. By now prices have returned back to something nearer a non-incentive "normal" of 96.5k CYN. The collapse of prices to this point means any supply deficit of perhaps 50k has evaporated. There is enough supply but there are still good prices and most operations were profitable except recent startup's that haven't made nameplate (Core!!). The market's probably close to balanced but without a huge shortfall looking for a home. The cumulative error is now ~200kt LCE.

BMI models a 2024 surplus of ~80kt LCE so their model has ~280kt LCE sloshing around and various commentors receiving BMI reports have noted values like 100kt of supply needs to come out of the market.

BUT, the run-rate error above is about 50kt/yr so its reasonable to consider that a similar error continues to exist in the 2024 numbers. The incremental surplus may actually only be 30kt and as noted above most of that carry-forward previous surplus of supply figures appear questioanble. If 2023 was back to near balanced, 2024 is indeed over supplied so prices fall. They aren't in free-fall which is what would happen if ~280kt LCE was sloshing around the markets looking for a home. Overall all demand being met and a small surplus would continue to push prices down. Perhaps that 30kt reached 50kt mid-September 2024. Since then prices rallied a little then came back off. This both-way price action is consistent with a finely balanced supply/demand situation. If there was anything close to 280kt of supply overhang existing, prices would still be collapsing and it would need supply well below demand to start chewing through that legacy overhang. I don't think that's the situation.

So while the chart below shows large supply surplus's, the pricing changes that have occurred indicate its overstating things. Its contributing to the market generally thinking there is a lot larger supply surplus than may actually exist. This inturn means the tightness leading to higher prices could be a lot sooner than the market is anticipating (Which would be good for GT1).