這篇文章可以做為上篇文章「一個投資分析師的告白──華爾街牛肉場」的補充資料。

文章來源是華爾街日報,它建議大家購買分析師評為賣出的股票。

Analysts Keep Misfiring With 'Sell' Ratings

By E.S. BROWNING
Staff Reporter of THE WALL STREET JOURNAL
April 11, 2005; Page C1

After all the post-bubble scandals involving analysts who recommended stocks they knew were troubled, you would think that Wall Street analysts would be doing a better job of picking stocks.

You would be wrong.

As has been the case in past years, stocks with large proportions of "sell" recommendations from Wall Street analysts have lately performed better than those with plenty of "buy" or "hold" ratings and no sell ratings at all, according to an analysis by Zacks Investment Research in Chicago, done for The Wall Street Journal. In fact, the stocks with sell recommendations have widened their lead since the stock bubble burst in 2000.

It might seem surprising that the stocks that Wall Street hates would do better than the ones it loves, but it makes a certain amount of sense.

Contrarian investors actually look around for stocks with a lot of sell recommendations, and, if the company looks likely to survive, they buy. The reason is simple: Wall Street analysts hate to tell clients to sell. They avoid lowering their recommendation on a stock until after something bad has happened, when the stock already has fallen.

"I have noticed it and I frequently buy stocks when they have sells on them," says Robert Marcin, who runs a hedge fund called Defiance Asset Management in Conshohocken, Pa. "By the time analysts get around to putting sells on stocks, the stocks are probably primed to perform well."

Mr. Marcin recalls speaking with an analyst last year about home builder KB Home, on which the analyst had an "underperform" rating -- a nice name for a sell. "It is up 100% since I spoke" with the analyst, he says.

Consider computer maker Gateway. At the start of last year, eight of the 14 analysts following it rated it a sell, and no one considered it a buy (the rest called it a "hold.") Gateway rose 31% last year, eclipsing the 9% gain of the Standard & Poor's 500-stock index. By the end of the year, its gains had turned some analysts into bulls -- just in time for it to fall 33% in this year's first quarter.

"Clearly no one has 100% accuracy with recommendations," says Rod Lache, auto-industry analyst at Deutsche Bank Securities in New York, who on Friday added his voice to the analysts who have put sell ratings on General Motors. "If the consensus is negative, of course you are going to look for the contrarian thesis. You ask: Could the market be wrong?"

But in the case of GM, despite all the negative views, he thinks analysts still aren't bearish enough. He actually had rated GM a sell until last fall, at which time a seemingly improved outlook led him to boost it to a hold. Now he figures he should have stayed negative on the stock, which continued to decline.

Despite some analysts' awareness of the problem, they clearly have trouble with their ratings. The problem has gotten worse lately, even after the Wall Street stock-research scandals. After the market's bubble, e-mail messages showed that some analysts had put buy ratings on stocks they privately ridiculed. Brokerage firms had to pay millions of dollars in settlements and promised to change their ways. Analysts were pressured to issue more sells. Wall Street firms assured clients that, henceforth, a buy would be a real buy, and when an analyst lost faith in a stock, he or she would issue a sell.

On the face of it, things did change. For a brief period in 2000, just as the stock bubble was bursting, 95% of the stocks in the S&P 500 had no sells at all, according to Zacks. No stock had more than one sell rating. That soon changed. Today, only 38% are without sell recommendations. Of the 62% with at least one sell, 9% have five sells or more.

But human nature doesn't seem to have changed. Even after this groundswell of Wall Street sincerity, the stocks with the highest percentage of sell ratings still are doing better than those without any sells.

In fact, the period when buy-rated stocks have recently performed best was during the stock mania of the late 1990s, when out-of-favor stocks were being abandoned and it was hard to find sell-rated stocks at all. From 1991 through 1996, the stocks with the most sell ratings outgained those without any sells, Zacks found. But from 1996 through 2000, those with no sells outpaced those with the most sells -- possibly in part because there were so few sell-rated stocks to measure. Since 2000, even though Wall Street supposedly has become more discriminating, the sells are leading again. In 2003-04, for example, the sells rose 36% on average, while the buys rose just over 25%.

To measure the trend, Zacks divided stocks in the S&P 500 into 16 sectors, from auto makers to computer makers to utilities. It included only stocks followed by at least five analysts. Then, to avoid focusing on a few out-of-favor sectors with heavy sell ratings, Zacks separated the stocks in each sector based on the proportion of sell ratings they had, and measured the different groups' performance over the following year. Because analysts' ratings are constantly changing, it renewed the process each month and averaged the results to get yearly performance.

Of course, not all stocks with heavy sell ratings are good stocks to own, and not all stocks with many buys are bad stocks. Enron, for example, was widely hated by analysts shortly before its demise, and they were right.

"You do have to do a fair amount of research" if you invest this way, to be sure you aren't getting an Enron or a WorldCom, says David Dreman, chairman of Dreman Value Management in Jersey City, N.J. He likes to buy stocks such as retailers, which can rise and fall sharply based on the economy, and tobacco makers, whose fortunes can depend on litigation. He finds analysts tend to abandon them at the bottom and begin recommending them at the top.

But the experience can be scary. He bought scandal-plagued Tyco International in 2002 in a belief that its business was solid, but the stock continued plunging, losing another 50% of its value. He needed courage to hold on, but today the price is almost twice what he paid.

Mr. Dreman believes analysts are often wrong. A study he has done shows that analyst forecasts for corporate profits in a wide group of large and small companies have been off over the past 30 years by an average of 40%, either above or below the actual result.

Scott Black, president of Boston money-management firm Delphi Management, ignores analyst reports and does his own research, saying, "We've always thought that they were promotional literature."


Friday's Market Activity

Constellation Brands rose 6.4%, or 3.48, to 57.95. Strong sales at the Fairport, N.Y., company's wine business helped it beat expectations for its fiscal fourth quarter. The company, which reported results after regular trading Thursday, said it would split its stock 2-for-1.

U.S. Steel dropped 4.2%, or 2.13, to 48.22. CIBC lowered its recommendations on the Pittsburgh company to underperform from "sector perform."

-- Gaston F. Ceron

Write to E.S. Browning at jim.browning@wsj.com

Lizzy 發表在 痞客邦 PIXNET 留言(0) 人氣()