What is the purpose of Inside Wall Street? To report the latest information and market talk--usually not yet widely known--that could affect the fortunes of companies and, therefore, the price of their stocks. Sometimes this information is about early discussions of possible mergers or takeovers. Sometimes it's about new products. Sometimes it's about corporate developments, such as restructurings, spin-offs, and asset sales. This information can come from investors, money managers, analysts, or executives themselves. It's in this sense that Inside Wall Street is ''inside.''

The column has performed pretty well indeed, according to an analysis of the 172 stocks featured in 1997. On average, the picks have measured up favorably against the market's main yardsticks: the Standard & Poor's 500-stock index, the Dow Jones industrial average, and the Russell 3000. BUSINESS WEEK compared Inside Wall Street's picks with those indexes over one day, one month, three months, and six months.

Each issue of BUSINESS WEEK is available online on Thursday evening and begins to reach subscribers and newsstands starting on Friday. So we measured the column's impact on stocks starting from Thursday's closing prices. The advantage is most pronounced in the short term (table). The one-day impact, on average, comes to a gain of 4.7%. Of the 172 stocks, 139 produced a gain on Friday, 101 were ahead after one month, 103 after three months, and 104 after six months.

The strong impact on Friday reflects to some degree the ''announcement effect.'' For example, when a well-known analyst recommends a stock, it often rises immediately. If the views of any widely followed market forecaster or investor become known, there is an immediate impact. When Warren Buffett, for instance, let it be known on Feb. 3 that he had bought 111.2 million ounces of silver in 1997, the metal's price shot up. The better the record of the forecaster, the greater the announcement effect.

For the stocks featured in the column, the announcement effect often hits at the opening on Friday. Traders enter their orders before the opening, and if there are lots of buy orders, the specialists or market makers will open the stock at a level above the Thursday close. Sometimes, the price goes higher during the day, but in general, many of the gains from the announcement effect show up early in the trading day.

Over time, the fundamentals assert themselves. If they're sound, then the stock may continue to rise in price well after the recommendation becomes public knowledge. Sometimes, the forecast is off the mark. No one will dispute that Buffett is an astute investor. But the price of silver, after surging initially, fell because of oversupply and weakening demand.

BIG MOVERS. In the case of stocks in Inside Wall Street, the data show that the gain for the average stock continues--but at a sharply reduced rate. After one month, the picks had gained an average of 5.4%. That handily beat the 1.9% gain by the S&P 500, the 1.7% garnered by the Russell 3000, and the Dow's 1.3%. The column's stocks also achieved decent returns vs. the indexes over the three-month and six-month periods. The column's picks posted a gain of 7.8% in the three-month period, outscoring the S&P's 7%, the Russell's 6.7%, and the Dow's 5.2%. And over six months, the column's stocks advanced 15.1%, a bit behind the S&P's 15.4%, but ahead of the Russell's 14.8%, and the Dow's 11.6%.

But averages are just that. The column has produced some sparkling winners as well as crushing losers--and those can affect the average. So we also examined the median return, which softens the impact of those big movers. Of the column's 172 stocks, the median return for the six-month period was 9%, a good deal less than the average return of 15.1%. That means half the stocks had returns in excess of 9%, and half were below. The median returns for the shorter periods varied: 3.2% for the first day, 2.4% for the first month, and 6.7% for the three-month period.

But such a gap between average and median returns is found in the performance of most portfolios. For instance, if you give equal weight to all the stocks in the S&P 500, the year-to-date return would be an average of 8.9%, compared with 13.9% for the capitalization-weighted S&P 500 (through June 19). And equal weights were used in measuring the performance of the Inside Wall Street stocks. But even the 8.9% is influenced by some of the bigger winners. The median return for the equal-weight S&P is just 6.7%.

Takeovers provided a big slice of the column's spice in 1997. More often than not, however, stocks mentioned as takeover targets do not get taken over. That is to be expected. Suitors looking for companies--or investment bankers trying to do deals--talk to many prospective targets before a match can become a reality. For example, Sanford Weill talked to Goldman Sachs and several other big companies before his Travelers Group zeroed in on Citicorp.

The job of the column is to report the news that company X is talking to company Y or that investment bank Z is trying to get companies X and Y together. The column also reports which ''takeover suspects'' are being bought by big players. Regardless of the outcome, stocks of these targets start to move as they begin to build up takeover premiums.

Mellon Bank, for example, was tagged as a takeover target in the issue of June 30, 1997. Among the suitors mentioned was Bank of New York. Mellon was then selling at 46. Six months later, the stock had risen to 60, even though no deal had materialized. (This year, Bank of New York made a hostile bid at 87 a share, which Mellon rejected. The stock traded on June 24 at 70 7/8.)

DARK HORSE. Sometimes, the successful suitor turns out not to be the one identified in the column. Take Salomon Brothers. It was trading at 63 when highlighted in the column as a target in the issue of Aug. 11, 1997. The likely suitor mentioned was a major New York bank. On Sept. 24, 1997, Salomon agreed to be acquired--but by Travelers--for $9 billion in stock, or about 78 per share at the time. By Nov. 26, 1997, the last day that it traded, Salomon stock closed at 83 3/8.

Also featured as a takeover bet was USLife: It agreed to be acquired by American General for $2.4 billion on Feb. 13, 1997--about a month after USLife was discussed as a target in Inside Wall Street in the issue of Jan. 20, 1997. The stock was trading then at 34 7/8; it rose a tiny 0.4% on the first day, and climbed 17.9% after a month, 30.5% in three months, and 51.6% by the time it stopped trading on June 17, 1997.

In some cases, companies highlighted in the column as takeover targets failed to end up in a deal during the six-month period. But a number of them did find themselves being taken over after that. One of them was Mercantile Stores, featured in the issue of July 7, 1997, when the stock was at 56 a share. The column reported that Federated Department Stores had been rebuffed earlier in its informal bid at 75 a share but that it might come back with an offer of 80 to 90 a share. Rumors persisted, but nothing happened--until May 18, 1998, when Dillard's Department Stores, another chain, made a bid to acquire Mercantile for $3.14 billion, or 80 a share.

Some of the column's biggest winners weren't even involved in takeovers. For instance, Premisys Communications was highlighted in the issue of Apr. 14, 1997. Inside Wall Street reported that technology guru Michael Murphy saw Premisys, then trading at 8 7/8 a share, as a ''screaming buy.'' In six months, it had soared 177.5%. An even bigger performer was EarthLink Network, whose stock, trading at 13 1/2 when mentioned in the column on Sept. 8, 1997, jumped 38% in one month, 41.7% in three months, and 263% after six months. Other big winners over the longer term: Protection One, up 90.8% after six months; Egghead.com, 91.7% in six months; Gencor Industries, 187% after six; and Lernout & Hauspie Speech Products, up 212.7%.

DELISTED. Inside Wall Street also featured some stocks that turned out to be big losers: The column of Jan. 13, 1997, reported that Jeffrey Logsdon of Cruttenden Roth, an investment firm in Irvine, Calif., was predicting that Graphix Zone, then selling at 3, would become one of the largest music Web sites on the Internet. That didn't happen. Instead, the company remained in the red and got delisted from the NASDAQ. The stock had gained 16.7% in one day, dropped 12.5% in a month, had fallen 58.3% in three months, and at the end of six months was down a staggering 88.3%.

Other big losers: CardioThoracic Systems, which fell 1.2% after one day, gained 7.5% in a month, lost 17.9% after three months, and fell 42.2% after six; Pacific Chemical Group, which gained 2.4% in one day but lost 26.7% in three months and dropped 44.8% in six months; and American Pad & Paper, which gained 3.8% after the first day but fell 5.2% in one month, 35.1% in three months, and 47.4% in six months. All told, about 40% of the stocks featured in the column had negative returns from the first month on.