Posts  / #POST-249354
REDDIT

A practical way to evaluate capital allocation beyond headline ROIC

Disclosure: I am the developer/operator of DDUBUCK Financial Data. The service is free and does not require an account. I am sharing the framework itself here and intentionally leaving out a link.

When I look at a company, I try to separate three questions that often get mixed together:

1. Is the existing business economically good?

2. Did management reinvest new capital at attractive returns?

3. Did the value created actually reach each share, or was it offset by dilution and stock-based compensation?

Headline ROIC mostly helps with the first question. For the second, I find three-year incremental ROIC more useful: change in after-tax operating profit divided by change in invested capital. It is imperfect when acquisitions or working-capital swings dominate, but it forces the discussion toward what the latest dollars of capital produced.

For the third question, I check owner-oriented free cash flow per diluted share, the change in diluted shares, stock-based compensation, repurchases, and dividends together. A large buyback number can look shareholder-friendly while the diluted share count barely falls. Likewise, high free cash flow growth at the company level may not translate into comparable per-share growth.

I also avoid applying this template mechanically to every industry. Banks, insurers, REITs, mortgage REITs, and investment companies need different measures of capital and distributable earnings. Forcing a conventional industrial-company ROIC calculation onto them can produce a precise-looking but meaningless result.

I built the tool because checking these items across up to twenty years of filings was tedious. It keeps the reported filing source beside the calculated metrics and leaves a metric blank when the required facts are missing instead of inventing a value.

What would you change in this framework? In particular, would you weight incremental ROIC, per-share owner earnings, dilution, and cash returns equally, or would you make the weights industry-specific?

This is a data-analysis discussion, not investment advice.