longtermvalue

Amazon: Is 'Free Cash Flow' More Important Than Net Income?

1) “Company makes very little profit, a total of $1.5bn in the past 6 years P/E 67- very high Market Valuation $33bn “
I have already described why I believed GAAP earnings to be relatively unimportant. Besides, I already conceded that it’s a bit pricey for my blood. My point in writing was to delve into this issue of why FCF would not be better than earnings.

2) “Their Total Debt has been increasing every year”
Debt load appears to be very manageable at more than 10X covered.

3) “FCF seems to be a way to pump up or support the stock price”
I think it is much harder to manipulate FCF than it is to manipulate earnings. Furthermore, price manipulations can work in the short-run, but in the long-run, the economics always win. Bezos, as owner ¼ of the equity, would appear to be in it for the long haul.

4) “You said Berkshire recognized the valued that float, he used a more meaningful definition to measure it, as was mentioned in the article.”
I think the definition you offered is sufficient in most circumstances of low/steady growth companies where you don’t have a negative CCC. In those cases, it just doesn’t move the needle much. I don’t know of Buffett offering the precise definition you cite (I don’t doubt it either), but I do know the ironclad one he has discussed is the present value of the cash that the owner can extract over the life of the business. From there, I use my judgment as to what can be extracted. A structural negative CCC represents extractable cash, and I have seen no argument why it isn't.

5) “The way some analysts, Amazon, FCF it can be easily manupulated. For example in the December quarter, if they obtained better terms on $100ml of A/Ps and delayed payment until January their "FCF" (their definition) would have been $100ml higher, $1.281bn rather than $1.181bn. Would that add'l $100ml be meaningful? (except for some investment income)?”
This is why I previously indicated that it must be structural issue and not just shifting between time periods. If it is a one-off thing where they were able to push out payables a bit longer this once and we don’t expect that to happen again, then that $100mm is not meaningful from a valuation standpoint. If, however, the business model is such that they are able to consistently generate cash from working capital, then an appropriate valuation should reflect that. When you talk about manipulation, I would also just say that getting better terms from suppliers seems more like a sound business practice that adds value than manipulation. And contrary to your claim, it is not easily done – by extending terms (i.e., Amazon wants 60 days to pay, rather than 30, or whatever the numbers are), Amazon would be extracting cash from its suppliers, who probably wouldn’t like it and would push back as much as possible.

Consider two fictional companies – AverageCo and RichCo:
AverageCo RichCo RichCo*
NI $100 $100 $100
D&A $40 $40 $40
Change in WC ($20) $20 $15*
CFFO $120 $160 $155
CFFI (incl Capex) ($40) ($40) ($40)
FCF $80 $120 $115
Value of firm at 10X $800 $1,200 $1,150
* Adjusted for $5mm of unsustainably extended terms

NI, D&A, CFFI are identical. The only point of difference is that AverageCo (like most companies) has a positive CCC (and WC is therefore a use of $20mm cash), while RichCo has a negative CCC (and WC is therefore a source of $20mm cash). The situation you present, a presumably unsustainable extension of terms with suppliers, would make RichCo’s change in WC of $20mm unsustainably high. So maybe it’s only a $15mm source of cash – that doesn’t make the $15mm irrelevant. I just slapped a simple multiple on it to make things easy, but if you DCF it, you get to the same place. If you accept the premise that FCF is a superior measure of value than GAAP profits (for reasons I discussed in my previous post), the valuation difference between AverageCo and RichCo is substantial (50% greater). Adjusting for the unsustainably extended terms decreases the valuation from $1,200 to $1,150, but it does not make them $800 (AverageCo's value) or $1,000 (10X the $100 accounting profit).