On the credit side you never start with what a company is worth. You start with what kills it. It's a different question, and it puts you in a completely different set of numbers.
Margin of safety on this sub means price. Your estimate says 20, it trades at 10, you have room to be wrong about the 20. Thats fine as far as it goes and its real protection against your own model.
But it only protects the estimate. It doesn't protect the holding period.
If the business has to survive five years for that gap to close, then whether it survives five years is part of the investment, and the discount does nothing for you there. Cheap and gone is still gone.
Graham had both halves. The defensive criteria weren't only about paying below value, there were balance sheet requirements too, current ratio, debt levels, earnings that held up in bad years. The price half gets quoted here constantly. I can't remember the last time I saw anyone apply the rest of it.
Things the credit side asks that I rarely see in a writeup here:
How long can this run on what it has right now, with no new financing.
What does a 30 percent revenue drop for two years do to it. Not to earnings. To whether it can pay what it owes.
When does the debt come due, and what has to be true in the market for it to get refinanced.
Are there covenants, and how close is it to them.
None of that shows up in a discounted cash flow. You can have a beautiful model and own something that gets diluted at the bottom or forced into selling the good assets to cover the bad, and your fair value was right the entire way down. Right and useless.
That's the thing that turns a drawdown into a permanent loss. Not the price falling. The company needing money at the exact moment nobody wants to give it any.
Where it gets hard is that the two halves work against each other. Companies with fortress balance sheets are rarely cheap, that's most of why they're expensive. Insist on both and your universe shrinks fast and you end up paying for the safety.
I've come down on the side of shrinking the universe. A company has to clear a set of financial strength tests before I'll spend any time on what it's worth. One example is a combination of interest coverage, debt measured against what the business actually earns rather than against equity, and how both of those held up the last time conditions were bad. Fail that and I don't open the valuation engine, doesn't matter what the price looks like.
It costs me. There are companies that worked out fine that I never even opened. I don't know how many, and I'm ok with it.