"Got a quarter? Can you flip it? Congratulations, you're a stock guru."

That's basically the thesis of my recurring "Beat the Street with 25 Cents" column. We Fools know that when a Wall Street firm upgrades a stock, investors are likely to bid the stock price up in response. Conversely, when a big-name analyst pans a company, the stock often drops.

But should it rise or drop? If "80% of mutual funds underperform the market," and if the vast majority of Wall Street analysts get more of their picks wrong than right, it seems odd for us to buy or sell stocks based on their say-so. That's why I use "25 Cents" to clue you in on the most clueless stock analysts.

The dirty half-dozen
However, some analysts hog all the attention in "25 Cents." Historically, firms like Maxim Group and Next Generation have occupied two of the bottom seven slots of the Wall Street roll call (the "unlucky seven"). That leaves only five spaces open for other lousy stock pickers to fight over -- and that just ain't fair.

To spread the wealth, as it were, we've introduced "Subpar Analyst Showdown" with the intention of pointing the spotlight a bit higher up the ladder. Here, we illuminate a few better-known Wall Street names that consistently score below 50% accuracy in their picks. Not as bad as their less-famous brethren, these names will be better known to the investing public.

By deflating the prestige-bubble these firms have undeservingly acquired, I hope to help you sleep better, secure in knowing which big-name firms are more often wrong than right. Here's this month's batch of not-yet-ready-for-prime-time players:

Firm

Accuracy

CAPS Rating (out of 5)

One Really Bad Pick

How Bad?*

Caris & Co.

49%

81.14

Nokia (NYSE:NOK)

23 points

Wachovia Securities

48%

79.93

Ryland (NYSE:RYL)

50 points

Roth Capital Partners

45%

67.42

Netflix (NASDAQ:NFLX)

49 points

JMP Securities

45%

"Under 20"

Lennar (NYSE:LEN)

71 points

Thomas Weisel

42%

"Under 20"

Force Protection (NASDAQ:FRPT)

70 points

Northland Securities

40%

"Under 20"

Smith & Wesson (NASDAQ:SWHC)

43 points

*How badly is this active pick underperforming the S&P 500? (Note that with the exception of Roth's rating on Netflix ("underperform"), all ratings were the equivalent of "outperform.")

Lies, damned lies, and statistics (Part 2)
Now, the caveats I've expressed before about CAPS hold true for the above firms as well:

  • We do not count ratings on "half-penny" stocks with market caps of less than $100 million or stock prices below $1.50 per share. Counting such picks could help (or hurt) the accuracy of the numbers reflected above.
  • CAPS is still in "beta." Glitches will surface that could affect our numbers. We'll do our best to squash the bugs as we find them, though, and we invite the named analysts to help us improve our product. If you have a gripe about your rating and the facts to back it up, we'll work with you to fix the problem. Drop our CAPS feedback board a note, and we'll give your arguments a fair hearing.

Plus, there's one other factor to consider in weighing this column's findings. CAPS' scoring system consists of two parts; we rank firms based on their accuracy, and on how right or wrong the analyst's picks are. In theory, an analyst can get two picks wrong for every one right, yet still rank very highly among investors.

For example, you'll see that despite sharing the odds of a flipped coin for accuracy, Caris & Co. nonetheless manages to place within the top 20% of investors. Superb picks like gene-tester Illumina (NASDAQ:ILMN) -- which nearly tripled over the two years that Caris has recommended it -- have more than offset losing bets like Nokia.

Incidentally, being right half the time is not an investment strategy without worth. Our hypergrowth stock team at Motley Fool Rule Breakers follows similar principles: swinging at fast pitches in hopes of hitting home runs. It's worked for us (see our scorecard when you sign up for a free trial here). It works for Caris. Just recognize that you'll probably strike out a few times along the way.