The future did not look bright for the American Stock Exchange (Amex) in the early 1990s. The once-venerable institution was losing the listing of some of its large, most widely traded companies to rival markets such as the New York Stock Exchange and the upstart Nasdaq.

The number of companies trading dropped from 896 in 1988 to 816 in 1992. Critics muttered it was in danger of becoming just like any other regional exchange - a far fall from its coveted place at the centre of the action.

In 1998, it was gobbled up by the National Association of Securities Dealers, owner of the Nasdaq. Since then, Amex officials say, business has picked up thanks to a refocus on mid-cap and small-cap stocks. But many market watchers believe the real turnaround came earlier, in 1993, with the advent of a single new instrument: exchange traded funds.

ETFs track indexes like a mutual fund, and can be traded like stocks. Average volumes are about 30m shares a day, accounting for more than half of the exchange's total volume, although Amex does not break out index-tracking funds separately.

Now, there are around 30 different products with more on the way. Assets total $37bn, up from $461m in the first ETF at the end of 1993. It's been a long journey since the concept of a tradable basket of stocks was first dreamed up in the late 1980s.

After one false start, four years of gestation through development and approval by regulators, another year or so to catch on with institutional investors, and even longer to take hold of the investing public's imagination, ETFs now reign as some of the hottest instruments around.

Part of the allure of ETFs is the coat-tail effect of the wild success of the S&P 500 index, and mutual funds that track it, over the late 1990s.

Hefty 20 per cent-plus returns blended with today's passion for trading daily - or even hourly or by the minute - has made ETFs a force the mutual fund industry is trying to figure out how to compete with - or replicate.

There is lots of competition. The grand-daddy of all the ETFs, the Spider which tracks the S&P 500, is still the most popular, with more than half the assets - $19.8bn at the end of 1999 - in the instruments. And it has been copied prodigiously.

Subsets of the S&P 500 as well as indices the world over have tracked using the same structure. As the toolbox of instruments grew, the assets under management grew rapidly. By the end of 1996, when World Equity Benchmarks (WEBs) arrived, which tracked global indices, there was $2.4bn in ETFs.

In the year from 1998 to 1999 assets more than doubled from $15.6bn to $33.9bn at the end of 1999. Predictably, with this kind of success, the blueprint is being further replicated.

Some of the new ETFs in the registration queue at the Securities & Exchange Commission will be traded on the New York Stock Exchange, bringing the Amex's lock on the ETF craze to a close. The smaller exchange hopes its head start will help.

Ironically, the ETF was an idea the Big Board originally turned down. The concept of an instrument that would offer investors the diversification of owning an entire index, with the ability to trade all day, did not have a lot of takers at first.

Dow Jones refused permission to peg the proposed product to its well-known Industrial Average index. Standard & Poor's, however, was willing to license the name of its large-cap index for the project.

But even with a relatively well-known index to peg the products to, difficulties overcoming resistance to the new vehicle did not stop.

Doug Holmes, an index specialist at State Street Global Advisors who worked with the Amex to design the first tradable basket of stocks mirroring an index, says it was slow going at first to drum up interest in the better mousetrap among brokers and institutional investors.

"It wasn't so much an objection as a lack of understanding," says Mr Holmes. Brokers and institutions had to be taught how the instrument could help in asset allocation.

But once institutional money managers understood how ETFs could be useful when they had put a lot of money to work quickly, they warmed to the products.

ETFs are also good for institutions that are unwilling to use futures contracts to deploy assets, or unable to because of charter restraints.

As institutional investors began to use the products more, brokers thawed, telling their customers about them, and the gospel was spread to individuals. Favourable press reports explaining ETFs' tax efficiency helped, says Mr Holmes.

Steven Bloom, one of the team at Amex that created the better mousetrap, says the struggle called for stamina from the developers, but was well worth it. "You really have to have a champion on it. You have to have conviction and exuberance from the outset."

r Bloom, now a consultant, notes that before ETFs, it was options in the 1970s, another product with a lengthy, difficult birth, which helped the market.

"If you go through the history of Amex, new products have breathed new life into the exchange," he says. "Its institutional survival has long relied on stock trading, but also new innovation and product development."

Some enthusiasts, including Mr Bloom, predict ETFs will relegate traditional, actively managed mutual funds to the dust heap of history. The notion of a mechanical index retiring all the stock pickers - as well as overcoming that tantalising possibility of beating the market - is premature.

But, the ravenous investor demand that ETFs have inspired has made the industry sit up and try to puzzle out how to replicate the better mousetrap with actively managed funds.

If they do concoct a way to trade actively managed mutual funds, Amex will fight to be the home of the first listing. It would have history behind it.