Price Multiples: P/E, P/B and EV/EBITDA

"Trading at 15 times earnings" packs a whole valuation opinion into three words. A multiple is nothing more than price divided by some measure of fundamental value — but the choice of which price and which fundamental changes what the number is actually comparing, and mixing them up is the most common error in a quant or banking interview.

P/E divides share price by earnings per share (equivalently, equity value by net income):

$$P/E = \frac{\text{Price per share}}{\text{EPS}}$$

It is an equity-value multiple — net income already sits after interest payments, so it reflects a claim only on what is left for shareholders. That makes P/E sensitive to leverage: two operationally identical companies with different debt loads will show different P/E ratios purely from financing.

P/B divides price by book value per share, comparing the market's opinion to the accountants' — useful for banks and insurers where book value is close to the assets actually being valued (loans, securities), less useful for a company whose real assets are people and code.

EV/EBITDA divides enterprise value by EBITDA:

$$EV/EBITDA = \frac{EV}{EBITDA}$$

This is a firm-value multiple: the numerator (EV) already includes debt, and the denominator (EBITDA) is measured before interest, so the leverage effect cancels out on both sides. That is why bankers use EV/EBITDA to compare companies with different capital structures, and P/E to compare what a levered shareholder actually earns.

A worked example

Company A: share price $40, EPS $2.50 → $P/E = 40/2.50 = 16.0\text{x}$. Its enterprise value is $3,200m and EBITDA is $400m → $EV/EBITDA = 3{,}200/400 = 8.0\text{x}$.

Now compare to Company B, operationally near-identical but funded with much more debt: same EV/EBITDA of 8.0x, but heavier interest expense drags net income down, pushing P/E to 22.0x on the same underlying business. EV/EBITDA says the operations are priced the same; P/E alone would have wrongly suggested Company B is "more expensive."