Why does Goldman Sachs forecast a lithium price of $8900 $/t while the actaul price is around 20000 $/t? This is quite gap.

You’re right to flag that gap—it *is* big. If spot lithium is around **$20,000/t** (China carbonate is recently quoted around 160,000–162,000 CNY/t, roughly $22k/t) and Goldman is talking about **$8,900/t in 2026**, that implies **~55–60% downside** from here. There are three main reasons a bank like Goldman can publish a number that low: --- ## 1. They’re forecasting the *cycle*, not today’s tightness Lithium is a cyclical commodity. Banks don’t usually say “price will stay near today’s level”; they build *cycle* models that tend to pull prices back towards: - **Long‑run marginal cost** (the price needed to keep enough supply in the market), and - **Incentive price** for new projects (the price at which future supply gets built). If their cost curve says: - A lot of planned supply can operate profitably at, say, **$6–8k/t**, and - You don’t need very high prices to bring *enough* new projects online, then a forecast around **$8.9k/t** is basically a statement that: > “By 2026, lithium will likely be oversupplied, and prices will have mean‑reverted towards the cost of the cheaper producers.” So the forecast isn’t about today’s spot tightness—it’s about where the **next down‑cycle equilibrium** might sit. --- ## 2. They expect a *big* supply wave to hit Most bearish lithium forecasts are driven by **supply**, not by demand. Analysts looking out to 2026 see: - **New hard‑rock mines** in Australia, Africa, and Canada ramping up. - **New brine projects** in Argentina/Chile and other parts of Latin America. - **Brownfield expansions** from incumbents (Albemarle, SQM, Ganfeng, Tianqi, etc.). - **Chinese converters** adding refining capacity and pushing more material into the market. If you assume: 1. A large share of those projects actually deliver, and 2. EV demand growth slows from “hype phase” to a more normal curve, you can easily get a model where **supply grows faster than demand** through 2025–2026. In that world, prices don’t need to stay at $20k+; they fall until enough marginal supply becomes uneconomic. Goldman’s $8.9k/t is basically saying: > “We think the project pipeline *actually gets built*, and the market swings into oversupply.” --- ## 3. They’re using a different *benchmark* than what you’re seeing “Lithium price” can mean several different things: - Battery‑grade vs industrial‑grade - Carbonate vs hydroxide - China domestic spot vs global contract prices - CIF (import) vs ex‑works prices You’re likely looking at a **spot** price (or a headline number) near $20,000/t. Goldman’s $8,900/t could be: - A **different quality/spec benchmark**, **and/or** - A **contract/realized price** proxy rather than a spot quote, **and/or** - A **long‑run “real” price** (inflation‑adjusted) expressed in a specific base year. All of those can widen the apparent gap between “what I see on screen today” and “what’s in a bank’s model.” --- ## 4. Their demand assumptions are deliberately conservative Most banks don’t like to build forecasts on the rosiest EV scenarios. Common conservative assumptions include: - **EV adoption still growing, but slower** than 2020–2022 expectations. - **Chemistry mix shifting**: more LFP (less lithium per kWh) and early sodium‑ion (no lithium). - **Efficiency gains**: battery makers using fewer kg of lithium per vehicle over time. - **Policy risk**: subsidies roll off, or economic cycles slow auto sales. Even if EV units rise, “lithium per vehicle” can fall, so **demand growth in tonnes** looks more modest. Combine *modest demand* with *aggressive supply growth* → **pressure on prices**. --- ## 5. Models are simplified—and often wrong at turning points One more important point: these forecasts are **not** a guarantee. They’re an output of a model that: - Extrapolates project ramp‑ups. - Assumes certain unit costs and FX. - Simplifies how quickly high‑cost producers shut down if prices fall. - Doesn’t fully capture political risk (royalties, export bans, community opposition). In reality: - If prices actually moved towards **$8–9k/t**, some high‑cost operations would likely **shut or delay**, tightening the market again. - Governments can **intervene** (e.g., Zimbabwe export bans, higher royalties) and restrict supply. - Demand could surprise **up**, especially if grid storage takes off faster than expected. So the **gap** you see—$20k spot vs $8.9k 2026—is partly: > “Spot is still reflecting today’s supply/demand tensions and some geopolitical risk, > while the model is projecting a ‘normalised’, oversupplied market two years out.” --- ### How to interpret this as an investor 🧠 - Treat Goldman’s $8,900/t as a **stress test**, not a certainty: “What happens to my lithium equities if prices *do* fall another 50–60%?” - Focus on **cost position**: - Low‑cost producers survive even if the bank is right. - High‑cost juniors are effectively a leveraged bet against that forecast. - Remember history: lithium has already shown it can go from **sub‑$10k** to **$70k** and back to **~$20k** within a few years. A forecast of $8.9k is not crazy in that context—it’s essentially saying, “we might end up back near the old ‘normal’.” --- Do you mainly care about what this implies for **lithium miners** (like Albemarle, SQM, juniors), or are you thinking more about **EV/battery names** that are *buyers* of lithium rather than producers?