Isn't Rapidus's 2nm effort at least partially funded in this manner?
(note, I'm not advocating this 'as a good method').
They are different.
Rapidus is funded by the Japanese government (through grants and share purchases) and by many Japanese companies such as SoftBank, Sony, Toyota, NTT, Kioxia, NEC, and MUFG Bank. These are conventional equity investors with no guaranteed investment return, dividends, or exit clauses. This is no different from you or me buying 100 shares of Amazon or 1,000 shares of Google. As typical shareholders, we have no guaranteed returns or dividends, and aside from selling our shares on the stock market, we cannot force Intel to buy our shares back.
The Intel–Apollo and Intel–Brookfield joint ventures are essentially high‑interest loans in disguise. Brookfield and Apollo are guaranteed to receive a certain level of investment return and are provided with exit options, privileges that private equity firms typically demand.
Private equity firms are not white knights. Their business model depends on securing high returns and well protected exit mechanisms. Without those, they would not function as private equity at all.
Major credit rating agencies Fitch and S&P treated the 2022 Intel-Brookfield JV as debt‑like (rather than equity investment) because the transaction included contractual, fixed minimum dividend guarantees to the partner.
Then in 2024, Intel and Apollo learned from that experience and structured their deal so it would be treated as an equity investment instead of debt. But based on the high price Intel is now paying to buy back Apollo’s 49% stake, I am almost certain that Apollo received protections similar to those Brookfield received, even if they don’t call it a “high interest loan.”
One thing is certain: even if Intel Ireland paid the same price as TSMC or Samsung for an EUV machine, Intel’s additional costs (such as this Intel–Apollo JV and the buyback) will make Intel’s true production cost much higher than its peers.
