Capex Outrunning FCF

It means the company is spending more on long-lived assets than its operations generate after working capital and maintenance investment. The pattern fires when the trailing 12-month capex-to-FCF ratio exceeds 1.5× and management characterizes the spend as growth investment with multi-year payback. The diagnostic is the gap between the spend and the verifiable return on prior capex cycles. Companies whose prior capex cycles produced documented returns above cost of capital can extend the gap; companies whose prior cycles did not are firing the pattern at elevated magnitude. Tesla's recent cycles are the framework's most-discussed contemporary case.

High capex is not bad in itself. The framework reads capex against three reference points: the company's own historical return on tangible capital, the company's stated growth thesis, and the trajectory of free cash flow conversion. Capex above 1.5× FCF that produces documented returns above cost of capital is the framework's compounder thesis. Capex above 1.5× FCF that has not produced these returns across prior cycles is the firing pattern. The discriminator is not the spending level; it is whether the spending has demonstrated return on capital across enough prior cycles to support the thesis. Compounder thesis depends on this measurement.

The framework's diagnostic conditions read three signals across the trailing 5-year window: return on tangible capital trajectory, capex efficiency (revenue generated per dollar of capex three years forward), and management capital-allocation track record on prior major investments. Companies passing all three read as legitimate compounders justifying high capex. Companies failing any one read as the firing pattern. The framework's compounder composite explicitly requires reinvestment yields above cost of capital — companies that fire high capex without this confirmation are the most-documented category of failed compounder thesis in the framework's case library.

The framework's case library shows heavy-investment payoff windows ranging from 5 to 12 years from initial capex acceleration to operational return materialization. Cisco 1995-2002 produced returns at the upper end of this range during the infrastructure buildout; Corning 1998-2003 produced returns over a similar window during fiber buildout. The current AI infrastructure cycle 2022-active is at the early stage of the equivalent window. The framework reads where each company is in its capex-vs-FCF cycle and tracks the operational signals that distinguish productive heavy investment from compounding capex burden. Free registration shows per-ticker reads on the live engine.

The framework reads AI infrastructure capex differentially by ticker. Companies producing AI infrastructure (NVIDIA, certain hyperscalers) have firing patterns that differ from companies consuming AI infrastructure (other hyperscalers, AI-application companies). The capex-to-FCF ratio reads acceptable for companies whose prior cycles produced returns above cost of capital and whose AI-specific revenue trajectory matches the spend trajectory. The pattern fires for companies whose AI capex acceleration outpaces verifiable AI revenue contribution. The framework's per-ticker reads in the live engine show which AI-cycle exposures are firing the pattern at what magnitude.