This series, brought to you by Yahoo! Finance, looks at which upgrades and downgrades make sense, and which ones investors should act on. Today, our headlines feature a better price target for CSX (CSX -3.02%), and an actual buy rating for Garmin (GRMN 0.20%) -- but...

It's curtains for Cree
The trading week dawned bleakly for investors in LED lighting specialist Cree (WOLF 8.46%) this morning, when analysts at Stifel Nicolaus announced they were pulling their buy recommendation from the stock, and downgrading to hold.

Why? For one thing, the analyst seems doubtful about Cree's decision to sell its LED light bulbs primarily  through Home Depot (HD -1.77%), and at subsidized prices. But Stifel also cites high valuation, and high expectations for future earnings (perhaps too high to meet) as primary drivers of its downgrade. And for me, these are the simplest reasons to shy away from this stock.

Priced at 74 times earnings today, and 44 times free cash flow, Cree shares cost a lot relative to the 15.5% long-term-growth rate that most analysts assign the stock. If Stifel is right, and this growth rate turns out to be too optimistic -- if it's right, too, to worry that subsidies may be a temporary phenomenon driving Cree's sales growth, and won't last forever -- then there are even more reasons to worry about the stock. But for now, simple overvaluation will do: 74 times earnings is too much to pay for a 15% grower. Full stop.

Putting the brakes on CSX
If 15% growth is "too slow" for an investment in Cree, you might think that the 10% projected growth rate at CSX would be a good reason to avoid that stock, as well. But in fact, I'm going to veto today's target price improvement to $35 (from FBR Capital) for CSX for another reason entirely.

As a matter of fact, if you value CSX purely according to PEG valuation, the stock doesn't look like all that bad of a deal. CSX shares currently cost just 15.5 times earnings. Between the company's 2.1% dividend payout, and its projected 10% growth rate, that's not a bargain, but it doesn't seem too egregiously overpriced.

The big problem at CSX, though, isn't growth, but cash. Specifically, the fact that CSX isn't making enough of the stuff. Free cash flow at the railroad amounted to only $943 million over the past 12 months. That's less than half the $1.9 billion in profits that CSX claims to have "earned" under GAAP accounting standards over the same period. Weak free cash flow at the company also isn't doing any wonders for the company's balance sheet, which currently sports about $8.3 billion in net debt.

Consequently, the stock carries an enterprise value to free cash flow ratio of about 40 -- very close to Cree's EV/FCF, as it turns out. So I guess when you get right down to it... yes, if Cree's 15% growth rate is too slow to support its 44 times EV/FCF, then CSX's even lower growth rate is a problem for this, stock, too.

Strike three, and Garmin's out as well
Last and least, we come to Garmin -- once the powerhouse company in the market for GPS navigation devices, but now, in the era of Google Maps, a bit of an historical artifact. This morning, Garmin won one of the few out-and-out upgrades of the day, when analysts at Oppenheimer assigned the stock a $58 price target and an outperform rating. But while the prospect of 27% profits (from today's price of $45 and change) may sound enticing, there's just one problem with Oppenheimer's recommendation:

It's dead wrong.

Priced at 15.5 times earnings (coincidentally, identical to CSX), Garmin's projected earnings growth rate of only 5.6% is much worse than CSX's (which is, in turn, worse than Cree's).

Granted, there are factors in Garmin's favor which could argue for the stock being at least a hold. For example, unlike CSX, Garmin is sitting on a nice pile of cash -- $1.2 billion, and with no debt. Garmin also generates very respectable annual free cash flow of $602 million, superior even to its reported earnings. Put these two numbers in a blender, and spin well, and you come up with an enterprise value to free cash flow ratio of only 12.6 on the stock. Between Garmin's 4% dividend yield, and its 5.6% growth rate, that's almost enough to justify owning the stock.

But even under this most favorable reading of the numbers, the stock's still not cheap enough to call it a bargain -- and so Oppenheimer is wrong to recommend it.

Fool contributor Rich Smith has no position in any stocks mentioned. The Motley Fool recommends Home Depot. The Motley Fool owns shares of CSX.