At The Motley Fool, we poke plenty of fun at Wall Street analysts and their endless cycle of upgrades, downgrades, and "initiating coverage at neutral." the pinstripe-and-wingtip crowd is entitled to its opinions, but we have some pretty sharp stock pickers down here on Main Street, too. And we're not always impressed with how Wall Street does its job.

So perhaps we shouldn't be giving virtual ink to "news" of analyst upgrades and downgrades. And we wouldn't -- if that were all we were doing. Fortunately, in "This Just In," we don't simply tell you what the analysts said. We also show you whether they know what they're talking about. To help, we've enlisted Motley Fool CAPS, our tool for rating stocks and analysts alike. With CAPS, we track the long-term performance of Wall Street's best and brightest -- and its worst and sorriest, too.

And speaking of the worst …
Semiconductor investors got a great big shot in the arm when one of Wall Street's name-brand analysts upgraded a raft of semiconductor stocks and reiterated its highest possible ratings on two more:

  • RF Micro Devices (Nasdaq: RFMD) and TriQuint (Nasdaq: TQNT) -- both upgraded to "buy."
  • Spreadtrum and MaxLinear -- both now "strong buys."
  • Cavium and Microsemi -- reiterated at "strong buy."

There's just one problem: the analyst doing the upgrading was Needham & Co. And when it comes to picking winners in the stock market, Needham's record is, shall we say, spotty.

You see, we've been tracking Needham's performance here at Motley Fool CAPS for going on six years now. What we've discovered, though, may not exactly thrill you: Technically classified an "All-Star" investor based on its outperforming more than 80% of the other investors we track, Needham is within just a few points of losing that rank. Its major malfunction: Needham just plain guesses wrong a lot. In fact, out of the 600-odd recommendations it's made, a sizeable majority -- 54% -- fail to beat the market at all.

The details
Call me a pessimist, call me a Fool -- but I fear Needham's not going to do much better with this batch of recs, either. Let's focus in on the two stocks where Needham's had the most change of heart this week -- TriQuint and RF Micro, both upgraded from "neutral" to "buy."

I went into some detail on TriQuint last month, on the occasion of Charter Equity's downgrade, and basically, my opinion of the stock hasn't changed. TriQuint is a fine company, with good leadership and a logical plan for developing its business. But the company simply isn't churning out cash the way I want to see it doing, and it's been forced to change direction on the products it's building to accommodate an industry shift toward 3G and 4G technology -- imperiling free cash flow even further. For that reason, I view any upgrade of the stock with skepticism.

RF Micro, in contrast, I haven't looked at in a while, so let's do that. According to Needham, now's the time to buy RF because (1) it has its dependence on Nokia (NYSE: NOK) sales down to about 10% to 15% of its revenue stream, (2) it's gaining chip share in Samsung's smartphones and "accelerating momentum … at other OEMs including" Research In Motion (Nasdaq: RIMM), HTC, and LG. For these reasons, Needham thinks RF Micro "has reached a revenue inflection point" and that its shares offer "an attractive long term entry point."

I disagree, and for several reasons.

Reason No. 1
I don't mean to hoist Needham by its own petard here, but … Nokia and Research In Motion? Really? I have to say that recommending a stock based on its doing business with these two industry laggards is immediately suspect. Nokia's on the ropes and has been forced to turn to Microsoft (Nasdaq: MSFT) for a rescue. RIMM … well, you know how badly RIMM has been doing, too. If RF is doing more business with LG, Samsung, and HTC, that's probably good news, but I fear it's been infected by groundless optimism about the also-rans.

Reason No. 2
Whether or not RF Micro has reached a "revenue inflection point" is open to debate, and time will answer that question. One fact that's not in question, though, is that RF's trajectory right now is following a downward path, with no sign of an upwards "inflection." On the surface, I admit, RF looks attractive at 16 times earnings, and with a price-to-free cash flow ratio that's even cheaper -- just 11.9.  My worry here is that while the stock looks cheap valued on its free cash, that number took a big tumble in the most recent quarter, falling from positive $46 million in last year's July quarter to negative $0.9 million in the most recent quarter.

Reason No. 3
That's a particularly disturbing trend in light of comments from JPMorgan Chase. Yesterday, JP took the opposite side of Needham's trade, arguing that far from being poised for "inflection," high inventory levels at RF Micro (along with peers Texas Instruments (NYSE: TXN), On Semiconductor, and Cypress Semi (NYSE: CY)), makes the stock especially vulnerable to the risk of a renewed global recession. Specifically, JP tagged RF Micro as a stock where "inventory levels are [among] the highest above normal among stocks in our coverage universe." JP worries that in the event of a downturn, RF  would be forced to reduce production, lowering capacity utilization and, as a result, squeezing margins -- and converting even fewer revenues into real cash profit.

Foolish takeaway
JPMorgan's reservations aside -- and my own -- I admit it's possible that this time, Needham is right about RF Micro and the other semi stocks it's backing this week. Before buying into the analyst's thesis that RF has hit an inflection point, however, I'd suggest you look for numbers to back up that point of view -- specifically, numbers on the company's cash-flow statement showing that free cash flow is back on the rise, and not continuing to slide.

As for me, until I see the evidence, I won't be buying RF Micro.