---
title:
slug:
date: "2026-07-04"
summary:
categories:
tags:
  - 
featured: false
published: false
---

[[IRR Math]] is interesting in this business - if a project has levered IRR of 20%, the unlevered IRR is higher because the cost of capital is much lower than the return, and the excess all accrues to equity investors (unlevered) 

IRR is never the same the return that is earned by the investors so IRR is often seen as the project's return. because it basically is the cash that is returned to the project but the moment that cash is returned from the project to the investors it is on the investor to reinvest it so there is always a reinvestment risk a project assumes that it would be able to reinvest it within the same project at the same rate of return. However that's not how things generally work in real life because investors would have to reinvest it. So if a project IRR is 20%, And if we assume that there are infinite possibilities to obtain more such projects, then maybe theoretically the project's 20% return makes sense. However, when the 20% of the euro, when the euro on cash flow is paid back to the investors, right? The investor needs to find another opportunity to invest. that money at a 20% return only then is the IRR is equal to the return to the investor. That's one thing that needs to be kept in mind. 

--- 

Unit Economics
$12B / GW - Coreweave Esimate
$14B / GW - SpaceXAI Estimate
$22-23B / GW - Anthropic SpaceXAI Deal
$50B / GW - Google SpaceXAI Deal
