Oliver Kamm, the Head of Strategic Research at Commerzbank Global Equities in London sorts out the investment chaff, in an exclusive article for BBC News Online.

"You can't do serious economics unless you're willing to be playful." So says MIT economist Paul Krugman in his new book "The Accidental Theorist".

Are hectic market movements rational or accidental?

The best protection against bad economics is to test an idea using an apparently trivial thought experiment. If it makes no sense in a simple example, it will still be nonsense when applied to the real world.

Krugman's test helps separate investment sense from nonsense. And it tells us that much conventional investment wisdom is not strategy but superstition.

Here are four common examples.

1. Last week markets suffered a "wave of selling"

If so, there was a simultaneous "wave of buying". If you sell shares, there has to be another investor on the other side of the transaction whom you sell them to.

And of course, investors didn't sell their equity portfolios on Brazil's initial devaluation and buy them back the next day. Prices were marked down, to reflect a perceived higher risk of global recession, and then readjusted to take account of new information.

A wild ride on the Footsie - speculation-driven or a reflection of profit expectations?

2. The bull market is being driven by a "weight of money" - mutual fund inflows, PEP purchases, etc.

The equivalent mistake: there can't be "more buyers than sellers".

Moreover, the level of the stock market is independent of how much money is invested in it. Since 1982, the correlation between net equity investment by US mutual funds and the annual returns on the US stock market has been zero.

To see why, try this thought experiment.

Suppose tomorrow every quoted company in the UK buys back all of its shares, except for ICI. Cash available for investment in equities rockets, while the supply of equities available for investment contracts massively. What will ICI's share price be? Several million pounds?

Of course it won't. ICI is worth only the profits it can generate for its shareholders in the future. If its share price rises above the market's estimate of what those profits are worth now, astute investors will sell the stock and put the money in an investment with a higher expected return - like a building society account.

3. Share "experts" can tell where markets are headed in 1999

They can't. And if they could, they wouldn't be telling.

Internet share frenzy or e-commerce promise?

Markets do make errors in valuing individual stocks, especially those that are less well-known than ICI. But "timing the market" doesn't work; it just loses money.

Why? Because you have to get two decisions right, not just one: when to get into the market and when to get out. So you have to be right a lot more than half the time if you're going to show a profit.

To do that, you need to know something the market hasn't "discounted" yet. But stock markets incorporate new information rapidly - consider Brazil's devaluation. There is no reason to suppose that a market guru knows something about business conditions that the market doesn't.

4. The Internet share price 'bubble' shows the irrational, speculative nature of the market

All stocks are priced according to the market's estimates of their future earnings. We don't have reliable knowledge of what the earnings of, say, ICI will be. But we do know what ICI's earnings have been in the past and can make a reasonable estimate of the likely range of possible future outcomes.

"Successful investing is about controlling risk."

Internet stocks have no historic earnings. Information about their future earnings is minimal. That is why their stock prices are so volatile.

But we do know that the Internet has huge growth potential. Those Internet companies that survive the next five years are likely to have achieved a commanding position - perhaps as dominant as the one Microsoft has built in computer software.

We don't know which these companies will be. But there is nothing irrational about paying an apparently high price for a company that might be a new Microsoft.

The market is aware that the risk of failure for these companies is high and that the probability of any particular company's becoming a dominant industry player is low. But the potential reward of getting it right is huge.

Investing is not ultimately about picking the right stock 'tips'. It is about effective management of risk. The Internet share phenomenon does not contradict that principle.

Strategy, not superstition

These notions are investment superstitions. All lack empirical support; two are logically impossible.

Do you sell if you have seen a black cat?

So why do they survive? Because we hunger to know the future, and desire quick and painless profits.

Neither course is open. Investment strategists do typically pretend to be able to forecast the market, and investors occasionally pretend to believe them - but it's a fiction. The real task of investment strategy is to test ideas and preserve us from the worst of them.

Risk management

Knowing that stock markets aren't driven by supply and demand makes us look at the things that do matter (corporate profits and discount rates), not at things that don't (mutual fund flows).

Knowing that successful investing is about controlling risk makes us prefer an efficiently diversified portfolio to a collection of "stock tips". And it ensures we avoid the course beloved of personal finance editors, of dribbling money into the market after regular intervals - a high-risk and irrational strategy, sold as the opposite.

Above all, it makes us ask, before judging any investment decision, not "what will happen, when?" but "what would happen, if?" Nobody knows what the market will do. But we can estimate, using reason tested against empirical evidence, what it might do under a range of scenarios.