The Question
"Solar" is a sector, but it's really a value chain: polysilicon producers feed wafer manufacturers, who supply cell makers, who feed module assemblers, who serve downstream installers and power plant developers. At any given time, only one or two segments capture most of the profit—and that allocation shifts rapidly with technology changes, capacity additions, and commodity price swings.
Buying a "solar index" or "solar ETF" exposes you to all segments equally, regardless of where profitability is concentrated. When polysilicon prices spike, upstream material suppliers make windfall margins while downstream module makers face compression. When panel prices collapse, installers and project developers benefit while manufacturers bleed.
Traditional top-down sector calls ("solar is bullish") or bottom-up stock picks ("this module maker looks cheap") miss the dynamic interplay between segments. Can you model prosperity at the segment level and rotate tactically within the solar value chain to capture profit migration before the market reprices it?
The Approach
We build a three-dimensional solar prosperity model using SAM product chain data, pricing data, and company segment revenues:
Dimension 1: Volume/Capacity
- Track new project announcements, installation forecasts, capacity additions across polysilicon/wafer/cell/module segments
- Measure participant entry and exit—when are companies adding capacity vs. idling plants?
- Identify bottlenecks: If cell capacity grows but polysilicon supply is constrained, upstream margins expand
Dimension 2: Price Dynamics
- Monitor spot and contract prices for key products: polysilicon (per kg), wafers (per piece), cells (per watt), modules (per watt), installation rates (per kW)
- Calculate per-unit profitability for each segment: (selling price - input costs) × volume
- Track price transmission: How long does it take for upstream price changes to flow downstream? (Answer: 1-3 months typically, creating exploitable lags)
Dimension 3: Technology Evolution
- Map efficiency improvements: PERC→TOPCon→HJT cell technology transitions
- Identify which segments benefit from new tech: When TOPCon ramps, silver paste suppliers and specialized wafer makers gain; legacy PERC cell makers lose
- Monitor R&D spending and patent filings (from SAM product classification) to predict next waves
Using SAM's product supply chain taxonomy, we map every listed solar company to specific value chain nodes—not just "solar company" but "polysilicon producer" or "TOPCon cell manufacturer" or "module frame supplier." We weight each company's exposure by segment revenue from Financial Notes (a company making 60% modules, 40% cells gets 60/40 split).
Prosperity signal construction:
- Calculate per-segment profit-per-unit weekly (using pricing data + cost structure estimates)
- Rank segments by week-over-week change in unit profitability
- Overweight top 2 segments, underweight bottom 2 segments
- Within each segment, overweight companies with highest market share (from SAM revenue data)
The strategy rebalances weekly—solar supply-demand dynamics shift fast, and segment profitability can flip within a quarter as prices adjust.
The Finding
The segment rotation strategy delivered 71.84% annualized returns from January 2021 to October 2022 with a Sharpe ratio of 1.74—massively outperforming a naive equal-weight solar portfolio (20.70% annualized, Sharpe 0.57).
The alpha came from catching profitability whipsaws:
Example 1: Polysilicon shortage (Q1 2021 → Q3 2022)
Polysilicon spot prices surged from $8/kg to $35/kg as wafer/cell capacity outran material supply. The model correctly overweighted polysilicon producers (Daqo, GCL-Poly) and underweighted downstream module makers facing margin compression (JinkoSolar, LONGi at that point). When polysilicon supply eventually expanded and prices crashed back to $10/kg in late 2022, the model flipped—overweighting module makers and underweighting material suppliers. This whipsaw rotation captured gains in both directions.
Example 2: TOPCon technology transition (2022)
TOPCon cells offer higher efficiency than legacy PERC but require different wafer specs (thinner, higher purity). As TOPCon ramped, the model identified that specialized wafer producers (companies investing in TOPCon-compatible capacity) were gaining unit profitability faster than the market recognized, since they could charge premiums while legacy wafer makers faced commoditization. Overweighting these early TOPCon enablers ahead of their earnings beats generated significant alpha.
Example 3: Installation surge in China (H2 2023)
Domestic installation rates in China spiked unexpectedly, creating a scramble for modules and inverters. The model detected rising per-unit profitability in inverter manufacturers and downstream engineering/procurement/construction (EPC) firms. While most investors were still focused on module oversupply concerns, the prosperity model recognized that installation bottlenecks were shifting margins downstream—and positioned accordingly.
Market share matters: Within each segment, the strategy overweighted companies with dominant market share (measured via SAM revenue data). High-share players often have better pricing power or cost advantages, so they capture disproportionate segment profits. For example, among polysilicon producers, the top 3 captured 80% of the margin expansion during the shortage period.
Turnover and implementation: Weekly rebalancing obviously creates high turnover—the strategy had ~40% monthly turnover, which is impractical for large institutional portfolios. But the signals persist at lower frequencies: rebalancing biweekly or monthly still delivered 50%+ annualized returns with Sharpe >1.0, making it viable for real-world implementation.
Try It Yourself
This strategy requires integrating three data streams: segment-level pricing (from industry data providers), company product mapping (SAM taxonomy), and segment revenue splits (Financial Notes or company disclosures).
Implementation considerations:
Price data sourcing: Polysilicon, wafer, cell, module prices are tracked by specialized solar industry consultancies (InfoLink, PVInsights, etc.). These datasets aren't cheap, but for sector specialists, the alpha justifies the cost. Alternatively, public companies often disclose average selling prices in earnings calls.
Cost structure modeling: To estimate per-unit profitability, you need input cost assumptions (e.g., polysilicon cost per wafer = spot price + transport + conversion fees). These can be built from industry reports or reverse-engineered from company gross margins.
Segment rotation vs. stock selection: This strategy is primarily about segment timing—which part of the value chain is hot right now. Within each segment, simple heuristics (market share, historical profitability) work surprisingly well. You don't need complex alpha models; the segment call drives most of the performance.
Applicability to other chains: The same framework applies to any vertically integrated supply chain: EV batteries (lithium→cathodes→cells→packs), semiconductors (silicon→wafers→fab→packaging), even agriculture (seeds→fertilizer→crops→processing). Map the chain, track segment-level profitability, rotate accordingly.
Capacity lag exploitation: Segment profitability shifts often stem from capacity mismatches. If you can model capacity additions with 6-12 month lead time (using SAM new participant entries, capex announcements, construction data), you can predict profitability shifts before they show up in prices.
Ready to build supply chain prosperity models for your sector coverage? Book a call to discuss product chain mapping, profitability estimation frameworks, and portfolio implementation.