Just two weeks ago, I argued that MercadoLibre (Nasdaq: MELI) can't keep rising. I made a pretty compelling case, if you don't mind my saying so. At nearly $70 per share, the auction site's current valuation bakes in more growth over the next 10 years than the entire Latin American region, from a demographic perspective, seems able to support.

But while I'm not buying MercadoLibre, I'm not shorting it, either. That's because hypergrowth stocks -- and MercadoLibre occupies that category, whether it achieves its valuation or not -- are notoriously difficult to predict. Couple that fact with the reality that shorting has an unlimited downside, and there's no sane reason to short MercadoLibre stock, even at today's lofty prices.

Good thing, too: MercadoLibre stock is up almost 10% since my article just two weeks ago. (Perhaps that will put an end to the always-fun conspiracy theories on the Yahoo! discussion boards, claiming that anyone who ever says anything negative about a stock is a nefarious short.)

Another example
For additional context on why shorting hypergrowth stocks is a loser's game, let's take a look at the three most expensive stocks in the Nasdaq 100 -- a tech-heavy, notoriously overvalued index:

Company

EV/EBITDA Ratio (TTM)

P/E (TTM)

Baidu.com (Nasdaq: BIDU)

67.3

87.6

Amazon.com (Nasdaq: AMZN)

31.1

53.7

Illumina (Nasdaq: ILMN)

27.0

79.8

Data from Capital IQ, a division of Standard & Poor's, as of Aug. 18, 2010.

For context, the P/E of the broader market today is approximately 15. It should go without saying that these stocks look extremely expensive. But are they shorts? Some data from three years ago might prove enlightening:

Company

EV/EBITDA Ratio (2007)

P/E (2007)

3-year return

Baidu

94.3

106.7

370%

Amazon.com

41.2

103.3

73%

Illumina

38.7

Neg.

97%

Data from Capital IQ, a division of Standard & Poor's, Aug. 18, 2007 to Aug. 18, 2010.

While I acknowledge that the EV/EBITDA and P/E ratios are often crude measures of value, each of these expensive stocks was nonetheless even more expensive three years ago. Had you shorted then, you would have had a disaster on your hands. These companies' shares have risen anywhere from 70% to almost 400%.

This is the problem with shorting hypergrowth stocks. Even though they may look expensive, and even though they may be facing challenges -- as Baidu is today, amid increasing competition from Google (Nasdaq: GOOG) and China Mobile (NYSE: CHL) -- they're also fast-growing companies with wide market opportunities and clear competitive advantages. If the companies succeed in executing on their business plans, however unlikely that possibility may be, any shorts run the risk of getting pounded.

That's why I'm not shorting MercadoLibre, and why I wouldn't short Baidu, Amazon, or Illumina today, either.

But there are stocks you can short
All the same, shorting can be a very profitable practice, and a potent way to even out the returns in your portfolio during bear markets. So how can you find stocks you should short? The key is to look for overvalued companies that also seem to be overstating the success of their operations, through practices such as channel-stuffing. That combination in particular makes for a promising short opportunity.

If you'd like to learn more about identifying potential manipulators, sign up below to get your free copy of "5 Red Flags -- How to Find the Big Short" – a new report from John Del Vecchio, CFA, leading forensic accountant and lead advisor to Motley Fool Big Short.