It’s no secret that solar supply chain costs have been at rock-bottom levels for the last 18 months. CRU Group’s Alex Barrows and Molly Morgan discuss how the market got to the point of the imbalance that drove the drop in PV prices, what this has meant for innovation, and how it may impact future technological transitions.
The Roots of Oversupply
Overproduction at levels that greatly exceed end demand is unsustainable for any business. The PV industry overproduced considerably from 2022 to 2024, resulting in huge inventory accumulation and putting manufacturers in a corner.
CRU Group estimates the worldwide module inventory might have reached a level to accommodate around 50% of installations in 2024 by the end of 2024.
It is worth noting that this overstock did not happen due to a reduction in end demand. PV sector installations were robust in 2022, 2023 and 2024, with outstanding year-on-year growth rates of 36%, 78% and 29% respectively. Since then, the increase in installations has slowed, and CRU Group projects single-digit growth rates in 2025 and beyond, suggesting that the “wiggle room” for extra production in anticipation of future market expansion has decreased even more.
Manufacturers faced a fiercely competitive market as supply far outstripped demand. If costs are greater than selling prices, then module producers have frequently had to accept negative profit margins, or risk being pulled out of the market completely.
This meant that in extreme circumstances some manufacturers used shortcuts to decrease production costs wherever they could to retain profit margins. In 2024 and 2025, quality problems provided another challenge for the sector. As CRU subscribers and others who have read the jointly produced “Solar Technology and Cost (STAC)” papers from module and reliability testing center Kiwa PVEL are aware, PV modules failing quality tests have been noticed at alarming rates.
Innovative Modules
But that’s not the whole story. However, innovation and module performance gains have not halted, even in a difficult market for manufacturers. Commercial maximum tunnel oxide passivated contact (TOPCon) module efficiency increased by 1.3% (absolute) from the start of 2023 to the second quarter of 2025, growing from 22.76% to 24.06%. In the same period, commercial back-contact maximum module efficiency grew enormously by 3.0% (abs.), from 22.53% to 25.54%, led mostly by Aiko Solar and Longi Solar, demonstrating that innovation has not stopped in any way.
If we take the average efficiency for both technology groups, the increase has been relatively moderate. During the same period, average efficiency for TOPCon rose by roughly 1.0% (abs.) and for back-contact by 1.2% (abs.).
When Will Pricing Pressure Ease?
The question we are asked most often by customers and industry contacts today is: when will the pressure on pricing come off and when will supply-demand dynamics balance?
At a high level, there are two situations that would bring the sector back into supply-demand equilibrium. You’d have to see end-demand improve to the point where extra inventory was no longer “excess.” But it is an improbable road back to balance as the increase in global demand slows over the next several years. A more plausible scenario is one where manufacturers scale down output so that surplus inventories are worked off slowly.
There was a widespread expectation that prices may start to recover around the middle to the end of 2025. This was predicated on the forecast that a considerable number of smaller firms would go out of business via bankruptcy in 2024. But bankruptcies and withdrawals from the PV sector have been slower than manufacturers and experts had predicted.
Now, in mid-2025, a corner has been turned on unsustainable overproduction. “Self-discipline” agreements among the main Chinese manufacturers have helped to lower production numbers since the beginning of 2025 — polysilicon production volumes were down by over 45% year-on-year in January to April, and wafer production volumes were down by over 20%. This has relieved the over-production situation, although significant stocks of polysilicon and, in many places, modules have persisted.
Manufacturing giants are signaling an opportunity to work through the inventory overhang by announcing more cautious plans for both capacity expansions and production estimates in 2025, at least by PV industry standards. Rumors of even larger production cutbacks are being planned by top players. If manufacturers do curb their output growth in 2025, then the likelihood of downward pressure on pricing throughout the PV supply chain abating in early-mid 2026 increases. If manufacturers don’t decrease output, then a return to broad profitability may be “one year away” forever, as it has been for the last 18 months.
After 2030
The PV sector may see fast technological transitions if a competing technology can provide superior performance at the same cost as the dominant technology. We have seen this with the fast move from back surface field to passivated emitter rear cell (PERC) and then from PERC to TOPCon. TOPCon surpassed 20% market share in 2023 and is expected to reach over 80% in 2025 in two years. TOPCon is likely to sustain its dominance in the present market share for the next two years. But on this front, new challengers are already rising, but it’s not yet apparent which is most likely to dethrone TOPCon in the future.
TOPCon vs. Heterojunction (HJT)
TOPCon and heterojunction (HJT) both have comparable theoretical limitations for efficiency, and neither seems to be in a position to create a lasting and substantial efficiency advantage over the other at this stage in the game. Today, the two technologies exhibit roughly identical peak and average efficiency commercially, with a narrow advantage for HJT in recent quarters. However, TOPCon and HJT are competitive in terms of efficiency, but not in terms of manufacturing cost at the moment.
The cheap manufacturing cost of TOPCon has been the driving force behind the technology leading the way in terms of market share and production capabilities. Therefore, CRU Group thinks HJT will have a hard time gaining considerable market share from TOPCon with higher production costs unless the technology can develop a meaningful efficiency advantage. To attain cost-per-watt parity, HJT requires an efficiency advantage of at least 2.5% (abs.) and above, even when accounting for the reductions in manufacturing costs we expect to be plausible from our bottom-up cost modeling. At this size, it does not look plausible that HJT efficiency is going to exceed TOPCon.
The Shift Toward Back-Contact
A more probable technological transfer is from TOPCon to back-contact, with back-contact likely to acquire market share over the rest of the decade since it offers better performance, greater efficiency, and competitive pricing. Currently, back-contact – in particular TOPCon-back-contact (TBC) cell design – has achieved efficiency gains of around 1.5% (abs.) over TOPCon and HJT. But this is only the case for the best-performing back-contact modules in Q2 2025, when the gap between the average efficiency of the two technologies is substantially lower at 0.5% (abs.). But these improved efficiencies come with a higher production cost for back-contact technology than for TOPCon.
As with all of the primary technologies on the market today, back-contact has a large spectrum of efficiencies within its own technological category. This implies that a prolonged efficiency lead would also need to be obtained more broadly among manufacturers in order to facilitate a broader technological shift, at least, assuming manufacturing costs stay higher than TOPCon.
With our internal cost modeling, we presently project TBC COGS to approach conventional (both sides contacted) TOPCon COGS by 2028, and then perhaps drop just below regular TOPCon by 2030. All this suggests that we may witness a substantial shift to TOPCon back contact around the end of the decade. Our estimate, however, does not show a clear crossover of prices as we have seen in the past with our modeling for TOPCon vs. mono PERC, making it unclear if any transition would be industry-wide or whether the two technologies may coexist. In any case, we anticipate seeing back-contact become a more frequent product option across module makers in the future years.
Beyond Silicon: The Tandem Technology Horizon
The PV sector is usually driven by a want to increase performance or decrease cost. Since 2017, the industry has been in an “era of performance improvement,” meaning that absolute cost reductions were tougher to accomplish but module efficiencies grew fast. Crystalline silicon technologies will probably reach actual efficiency limitations by the end of the decade. How will the industry continue to push performance improvements?
The next technological transformation is looming. CRU Group predicts that in the early 2030s, tandem technologies, such as silicon-perovskite designs, might gain momentum at scale, since they have far more efficiency potential than single-junction silicon devices. Some manufacturers are already commercially manufacturing perovskite-tandem modules today, although at very tiny sizes; therefore, the early beginnings of a tandem technological shift are already underway.
