It's that time of year again. "Six sizzling stocks for 2008"; "What's hot and what's not for the coming year"; "Seven stocks to avoid": there's something about a new year that brings out the prognosticator in people. The question is, who do you believe? The media pundit? The analyst? The newsletter writer?

Answer: none of them. That's the unequivocal conclusion of a myriad of academic researchers who have painstakingly researched the subject of stock tips.

Media tips have long been scoffed at by professionals in the investment business. After all, "sell when BusinessWeek'scover says buy" has long been a Wall Street maxim. That magazine's comically long litany of ill-timed covers led to the popularisation of the Magazine Cover Indicator - the notion that one should run a mile from stocks being feverishly tipped on the cover of mainstream business magazines.

Silly as it sounds, the theory received academic approval this year after academics, poring over 549 covers over a 20-year period, concluded that "positive stories generally indicate the end of superior performance and negative stories generally indicate the end of negative performance". Stocks receiving the most negative coverage went on to trounce the stocks most favoured.

That paper focused on three of America's most prominent magazines - BusinessWeek, Fortuneand Forbes. The same publications were analysed in a separate paper which looked at every stock tip generated by magazine columnists from 2000 to 2003.

Looking at different timeframes (weeks, months or a year), it found that the buy recommendations of columnists invariably underperformed the overall market. BusinessWeek'stips were the most eagerly awaited, with high trading volumes and an average price rise of 2.5 per cent occurring from the day before public availability through to the day after.

Unfortunately for investors who bought at these elevated levels, the stocks quickly gave up their gains, typically reversing over the following 10 days. The authors go on to say that, "even an investor who was able to buy on publication date from the previous day's closing could not generate abnormal returns by investing on the stocks recommended by BusinessWeek".

Other papers confirm media ineptitude. Prof Thomas Schuster of the University of Leipzig says many stocks pop after being tipped, a product of "price pressure from 'naive' investors hoping to profit from the experts". The rise is shortlived, however, as "most media lack any real information that is not yet reflected in stock prices", ensuring that "investors who follow such advice will lose in the long run".

Today, people increasingly get their tips from the internet or financial television. Particularly popular is Mad Money, the CNBC show characterised by the madcap antics of its host, former hedge fund manager Jim Cramer.

It's not your typical financial programme. In fact, it's positively bizarre, with Cramer's yelling and screaming typically accompanied by a host of peculiar sound effects (pigs squealing, toilets flushing, trains crashing) or the sight of plastic bulls or bears being roasted in an oven.

Despite this, Cramer purports to be an earnest stock-picker. "Picking the right stocks is one of the hardest points of investing, and every night on Mad MoneyI try to take some of that burden off your shoulders," he writes in his book, Mad Money: Watch TV, Get Rich.

Unfortunately, research suggests that following his tips does anything but. Cramer's picks tend to pop overnight - on average, by almost 3 per cent, although smaller stocks see an average gain of almost 7 per cent, according to an academic paper published last summer.

These gains tend to disappear within days. Trading volumes and short sales activity appreciate substantially on the day following Cramer's recommendations, suggesting that "smart" money is benefiting from the impulsive actions of the Mad Moneyviewers. Nor do his tips fare well on a longer timeframe.

Madmoneymachine.com, a website that tracks Mad Moneytips, calculated that a portfolio of Cramer picks would have just failed to break even in 2006 - a poor record, given US markets rose by 16 per cent that year.

Another study went back further, looking at every tip given by Cramer since 2000. This included more sober commentaries in his subscription-based columns at thestreet.com, the financial website he founded.

Conclusions? His predictions "sometimes swing dramatically from optimistic to pessimistic, and back again, over short periods" and his "stock market calls since May 2000 have low consistency and average accuracy".

Cramer's popularity is proof that the most widely followed gurus are not necessarily the most accurate. US-based research firm CXO

Advisory Group found that the most searched-for gurus on their site achieved "only average to well below average accuracies" but had "frequent and systematic national media exposure" and that "media exposure, rather than forecasting accuracy, probably drives investor/trader attention to stock-market gurus".

Still, Cramer and his ilk are an easy target. Professional analysts are no better. Research from David Dreman, one of the few fund managers to beat the market averages over the past few decades, shows that analyst earnings forecasts are hopelessly wide of the mark - typically, they miss out by 30 to 60 per cent.

Not only that, but Dreman calculated that the odds of analysts' consensus forecasts being within 5 per cent of the actual results for any four consecutive quarters are slim indeed - there's a one in 170 chance of it happening. "Large earnings surprises are the norm, not the exception," says Dreman.

"Analysts, money managers and investors appear to ignore the industry's poor forecasting record, although it questions the viability of many important stock valuation methods."

Dreman looked at forecasts between 1973 and 1993. More recent research shows that things haven't improved in the analyst community. A study in 2006 looked at 100,000 12-month stock-price targets made by analysts between 1997 and 2002.

Conclusion? "Expected returns exceed actual returns by 35 per cent and only 24 per cent of target price forecasts are met at the end of the 12-month period."

What about investment newsletters? One study which looked at 237 newsletters from 1980 to 1992 concluded that fewer than 35 per cent achieved market-beating returns. Another study found that newsletters underperformed the market in 11 out of 16 years between 1980 and 1995.

Why then is there such a huge appetite for tipsters and forecasters and price targets and other such malarkey?

According to legendary trader Jesse Livermore, it's a form of optimistic laziness. "The average man . . . [ wants] to be told specifically which particular stock to buy or sell. He wants to get something for nothing."

Cynical? Perhaps. It may be that the people who jump on tips are simply innocent souls who believe "expert" commentators are "in the know". Indeed, the very fact that most people choose to put their money into expensive actively managed funds rather than cheap index trackers, despite a mountain of research showing that most fund managers underperform the market, suggests that investors have too much faith in the so-called pros.

Irrespective of the reason, there's little doubting the conclusion of the professors: tips, quite simply, should be left to waiters, not investors.

This journalist was not the originator of that waiter quip, by the way. Ironically, Cramer came up with it.