How do you actually evaluate whether a stock is "fairly valued" when the metrics all say different things?

This is the problem that's been frustrating me the most lately and I don't think there's a clean answer.

Take almost any large-cap stock right now. Look at P/E: it looks expensive. Look at PEG: it looks reasonable. Look at forward P/E based on analyst estimates: it looks cheap. Look at free cash flow yield: it tells a different story than earnings yield. Look at EV/EBITDA: different again.

And it's not just the metrics disagreeing with each other. It's the benchmarks. Is a stock expensive compared to its own historical average? Compared to its sector? Compared to the S&P? Compared to the 10-year Treasury yield? Depending on which comparison you pick, the same stock can look both overvalued and undervalued simultaneously.

Then you layer on qualitative factors. Is the moat widening? Is the TAM expanding? Is management executing? These are subjective assessments that can justify almost any multiple if you're creative enough with the narrative.

The cascade: if I can't confidently answer "is this cheap or expensive," I can't size positions properly. If I can't size properly, my portfolio construction is arbitrary. If my portfolio construction is arbitrary, I'm not really investing, I'm just buying things I like and hoping.

I've settled on using a combination of forward P/E relative to sector average plus free cash flow growth rate as my primary valuation frame. But I know it's incomplete, and I know smart people would use completely different metrics and reach different conclusions.

What's your actual process? Not "what metrics do you look at" but how do you reconcile conflicting signals and make a final decision?