| Force | What it asks | What strong looks like |
|---|---|---|
| Bargaining power of buyers | Can customers force the company to cut prices or accept worse terms? | Fragmented customer base with no single concentrated buyer; the product is small relative to the buyer's total spend (a switch is not worth their time) |
| Bargaining power of suppliers | Can suppliers raise prices or capture more of the value? | Many suppliers competing for the firm's business, or supplier inputs are commoditized; the firm can vertically integrate or substitute upstream |
| Threat of new entrants | How easy is it for someone with capital to start a competitor? | High up-front capex, regulatory licenses, decade-plus brand-building, or network effects -- all costly for new entrants to replicate |
| Threat of substitutes | Could a different product solve the same job for the customer? | Strong switching costs (data, training, integration), product is irreplaceable for the use case, or substitutes exist but are meaningfully worse |
| Industry rivalry | How intense is the competition between existing players? | Few players, differentiated products, growing industry, capacity discipline; the opposite is the airline industry circa most of its history |
Worked example -- NVIDIA's five forces in 2020 vs 2024. (1) BUYER POWER: shifted from price-sensitive gaming OEMs and crypto miners to capacity-constrained cloud hyperscalers competing for allocation. (2) SUPPLIER POWER: TSMC is the chokepoint and grew stronger as the only credible leading-edge foundry; modestly negative for NVIDIA. (3) NEW ENTRANTS: AMD's MI300 launched 2023 but the CUDA software gap proved harder to close than the silicon gap; Intel exited training-class AI. Net: lower threat. (4) SUBSTITUTES: cloud TPUs (Google) and custom silicon (AWS Trainium, Meta MTIA) emerged but at smaller scale; the substitution risk is real but currently bounded. (5) RIVALRY: with 80%+ share in training-class GPU and an effectively-uncontested software stack, rivalry was DOWN despite the price points being UP. Four of five forces strengthened structurally between 2020 and 2024. That is what the higher absolute share price was paying for.
The Five Forces is a frame, not a forecast. A 2010 analysis of Nokia using these forces would have looked strong: brand, scale, supplier power, distribution. The framework didn't predict that smartphones would redefine the substitute set within three years. The lesson is to weight 'threat of substitutes' heavily for any company where a technology shift could re-categorize the product entirely. The framework also says nothing about MANAGEMENT QUALITY or CAPITAL ALLOCATION, both of which can destroy a structurally favored business (Sears, Kodak) or rescue a structurally unfavored one (American Airlines, post-bankruptcy).
Industry structure sets the long-run ceiling on the returns a business can earn. Run the Five Forces before you build a DCF or pick a peer multiple, and run it on the same company at two points in time to see whether the franchise is widening or eroding. Multiple expansion on top of widening forces is the compound-bagger pattern; multiple expansion on top of weakening forces is the value-trap pattern.
Sit with the ideas.
NVIDIA in 2020 had a P/E around 60. NVIDIA in 2024 had a P/E around 50 despite the share price being roughly 10x higher. What changed about the BUSINESS that justified the market accepting the higher absolute price at a similar earnings multiple?