Because without profit the denominator is zero or negative. The P/E then says nothing.

Is a high sales multiple always expensive?

Not automatically. With high growth and rising margins it can fall without the price falling (SpaceX: from 92x to 18x when you use 2027e instead of TTM). What matters is whether the forecast comes true.

When do you go back to valuing on profit?

Once earnings and FCF are stably positive and the margin has settled. Until then it remains a range with wide ends.

Knowledge · Growth stocks

Valuation without profits: it still works

When earnings or free cash flow are negative, P/E and a simple DCF do not work. Then you value via sales multiples, normalised future figures or individual divisions.

At a glance

Tools: EV/sales, EV/EBITDA, valuation on future years (2027e instead of TTM), SOTP and a DCF with a path to a target margin.
Negative FCF can come from investment (Amazon, SpaceX data centres) or from losses. You have to separate the two.
The further profits lie in the future, the wider the range.
For growth companies there is the “Rule of 40” rule of thumb: growth plus margin should be at least 40%. It comes from software and is only a guide.

Formula

EV/sales = (market cap + net debt) ÷ sales
Value    = sales(e) × multiple + net cash

Worked example: SpaceX (as of Oct 16, 2026)

Metric	Value
Free cash flow H1 2026	−$26bn
EBIT Q2	−$0.14bn
Sales Q2	$7.81bn (+92%)
Gross margin	52%
EV/sales TTM	approx. 92x
EV/sales on 2027e (consensus $115bn)	approx. 18x
Peers: Rocket Lab / Redwire	approx. 50x / approx. 4.6x

EV/sales of SpaceX and peers

SpaceX: run rate versus 2027e

Plus the SOTP calculation with 10x / 15x / 20x sales (bear / base / bull):

Basis	Bear	Base	Bull
Run rate (Q2 × 4)	$27	$43	$70
Sales 2027e	$92	$135	$179

The share price is $146.03. The difference between the two rows lies solely in whether you believe the growth forecast.

More examples

Amazon: negative FCF → DCF $104 → SOTP $224 instead.
BYD: EV/EBITDA 5.7x, no fair value stated.

What the number does

On run-rate figures the price is more than three times the base, on 2027e figures roughly at the base.
So you are buying the forecast along with the stock.

Pitfalls

The multiple comes from a hype phase of the peers.
The consensus forecast rises with the hype.
Dilution from stock compensation and capital raises is easily overlooked.
The terminal share is very high (SpaceX DCF: 77%, see terminal value).

FAQ

Why not P/E? Because without profit the denominator is zero or negative. The P/E then says nothing.
Is a high sales multiple always expensive? Not automatically. With high growth and rising margins it can fall without the price falling (SpaceX: from 92x to 18x when you use 2027e instead of TTM). What matters is whether the forecast comes true.
When do you go back to valuing on profit? Once earnings and FCF are stably positive and the margin has settled. Until then it remains a range with wide ends.