@tjt - It totally depends on the business. In this case, no, earnings are not very relevant. Here, something like the current ratio and when liabilities mature would be important from the balance sheet. I don't consider the income statement to be very useful here.
A growth business would be evaluated based on P/E. I think of these as having P/E be the leverage on the earnings. Hence, when earnings increase, P goes up accordingly. The higher the P/E the more money can be made or lost.
Stodgy old companies that make saws and chisels and comprise buildings, trucks, ... would be evaluated on P/B. If there's nothing special about them, then their value = the sum of their parts and nothing more. So you're trying to buy assets-liabilities = book value for less.
Cash cows (big telecom, conglomerates, etc.) would trade on EBITDA/EV which is essentially the value of the company if it was bought up for cash (EV) and the money you'd get if the company wasn't a public entity, loosely speaking. EBITDA/EV are comparable to P/E but they're typically much lower, like 5-10.
Then you have REITs and other types where the tax law gives you big write offs even though your assets aren't really decaying. They should be evaluated based on FFO or something similar.
For the dividend payers, you have growing dividend payers and nongrowing. Lots of people like the former. I find them dangerous because their evaluation is based on a guess for the growth rate of the dividnd which acts as a significant leverage. If they "Fail a growth target" it results in _significant_ capital losses as the stock price drops. See what happened to GE. OTOH, growing dividends means that the company is growing, so it's a trade-off. Nongrowing companies are more predictable ... here you gotta worry whether the business model is still viable in 50 years... like tobacco or tool companies. These companies have higher pay out rates because there's no point in retaining dividends if say your customer base is declining by 1% annually.