It's quite difficult to get a good handle of the supply and demand narratives being tossed around.
I don't agree with the approach of taking each company's projected production and summing them up. I would take away all of the projects which are not yet fully funded (yes, including SDV and JB for now). Apply estimates for time delivery multiplied by a contingency factor. Multiply projected output by a reduction factor based on previous performances across the industry and reducing the final by an overall factor to represent the difference in the amount of lithium mined to the amount of lithium sold. The resulting supply estimate is significantly less on these considerations alone.
Obviously in a market where there is less product than demand the price goes up, but what happens in the world if the supply balances out or goes into surplus? It does not reverse as you may think it might. Lithium producers don't pool all their compounds into one big bucket and distribute it evenly on the spot market, they cut deals with supply chains to buy their product. A compound processor will only use DSO if it turns out to be cheaper than getting a usable compound directly from a supplier which means that all of the DSO could potentially never be used.
But this philosophy can be extended to lithium concentrates themselves. In a hypothetical world where there is 1 tonne of spodumene for every tonne of converter demand I think it would be very unlikely that all of the tonnage finds itself distributed squarely. The converters would still seek the best deals to them which may be a handful of producers while the more expensive producers will just never get looked at. I realise that the argument is that the prices at that point may be adjusted to try and balance out but would it really play out like that?
For example:
A supplier of 100kT per year of spodumene and a converter of 100kT demand
another supplier wanting in on the act says they will supply 20kT of spodumene at a cheaper price and the overall system is in oversupply by 20%. What happens? Do prices crash in a price war? I 'd argue that they don't necessarily. It is still plausible that the converter would change absolutely nothing about their current arrangement as they want to secure the supply as much as anyone and may not want to "bite the hand that's feeding them" or "rock the boat" too much.
In order to make the deals of supply now these arrangements pretty much have to take place before anything is signed to go ahead. Now you might see what some of the complexities of the SDV supply deals may be. In order to secure the deal(s), you need to ensure delivery of supply, which would be really helpful if you had an operational plant to prove you could meet supply, which you can't build without the deal...
Except you could do small scale production and tests funded through company cashflow if you had any. Hmmm...