Last week I wrote about some academic research that suggests that Internetcommerce might, indeed, be too much of a good thing. Half of it is the oldWal-Mart argument that big, cheap stores will kill little stores that can never bequite as cheap. The other part was the result of simulations that showed onlineretailers entering a price-cutting frenzy that could end only in bankruptcy for thefinancially weaker player. If this were to happen, of course, prices could then goback up until the next competitor and the next price war erupted. Eventuallyretailers would stop entering the business with such glee and electronic commercewould slim down to just another way of buying stuff.

The general reader reaction to these ideas ranged from yawns to argument. Thearguers tended to see a greater value in buying from a local vendor either becauseof instant gratification or quality of service. In response to this I hold up the PCsoftware business, which is hurtling toward total electronic distribution (or at leastelectronic sales and UPS distribution) at an amazing rate. Egghead went electronicfor a good reason, though it's my understanding that the urge to sell direct has asmuch to do with the fact that distributors are paying an average of six months lateas anything else. By the time Ingram pays for the software you've sold throughthem, your little company could be out of business.

So I believe this stuff, that we'll find a negative side to this rush to Internetcommerce and ultimately it will be important, but not as important as some peoplethink it should be. Perfect markets, it seems, can become perfect messes. But Ialluded last week to an answer to all this, a possible alternative that would makethe imperfection of the perfect market seem insignificant. That's this week'scolumn. Hold on, because this is not going to be an easy ride for the next fewparagraphs.

Let's for the moment put aside any worries about how electronic commerce won'twork and instead look at its strengths. As illustrated in last week's column, it can bea very good way of finding the lowest price when you know exactly what you wantto buy. A lot of people are already using local stores as showrooms for Internetretailers, something the local retailers don't like at all. The essence of finding thelowest price is the Internet's elimination of time, distance and, to a certain extent,effort. These are, after all, computers, and they can be used to crunch data as wellas download dirty pictures. So the Internet is well suited to any task where thecomputer can be used to eliminate human effort. That's part of the reason whyInternet stock trading is growing like crazy (that and the predicted price war: haveyou noticed the downward trend in sales commissions?).

Buying and selling stocks, applying for mortgages, anyplace where clerical laborcan be eliminated and the savings passed on to the consumer is perfect for theInternet. E-Loan, the Internet mortgage broker, claims for example that it can savethose seeking a home mortgage an average of a quarter to half a percent on thebase interest charge (that's the part where we use the Net to find the best price)and 75 percent on closing costs (that's using automation to eliminate labor and cutcosts).

But these computers can do so much more. What if we used the automationpotential to buy and sell goods that have never before existed, that can be made toexist only through the application of the very technology that to this point we'veonly considered as a new retailing channel? This is my big idea, an idea that's notonly bigger than the whole Internet itself, but hasn't, to my knowledge, yet beenthe basis of an Internet startup. Want to found the Internet's first $100 billioncompany? Read on.

I know this looks like a digression, but it's not. Remember the energy crisis? OPECstarted an oil embargo and the price of crude oil rose over seven years from aroundfour dollars to $41 per barrel. It was the biggest transfer of wealth in history. Butwho, today, gives a second thought to OPEC? The price of oil has dropped slowlyfor the last 14 years. Correcting for inflation, the price of oil is today about where itwas back in 1973, before the oil embargo. What happened?

OPEC's fall from grace can be traced to two influences, with one being much moreimportant than the other. Dusting off the law of supply and demand, that $41 perbarrel price stimulated a lot of oil exploration and made it economically sensible toapply enhanced extraction methods to recover more oil from known reserves.Under this influence, U.S. oil production rose from seven million to 11 millionbarrels per day by 1984 - just about the time when prices again began to drop. Butthe reason for OPEC's decline goes far beyond supply and demand because U.S.oil production is now back where it was, around seven million barrels per day andwe are importing more oil than ever before. Yet still the price of crude continues itsdecline. Why?

OPEC's real nemesis is the Chicago Board Options Exchange and the market in oilfutures. Before the oil embargo, there was very little volatility in oil prices - they'dbeen dropping slowly since around 1960 - so there was little incentive to create afutures market. But OPEC changed that, giving to the CBOE the reason - hugevolatility and hence huge trading profits — to invent an anti-OPEC weapon. The oilfutures market that appeared in the early 1980s made it possible for the first timeto effectively hedge against price volatility. No matter what OPEC did, the CBOEcould absorb the economic impact, making money along the way.

At the heart of this is the concept of the derivative security, the ultimate product ofwhat Wall Street calls "Rocket Science." Sure, derivatives are a bad word to some,but they don't have to be. While nobody on the street seems to know where thename "rocket science" came from, when it came is very clear. The financial worldchanged after 1973 when Fischer Black and Myron Scholes, a pair of academics,published their landmark article "The Pricing of Options and Corporate Liabilities."What Black and Scholes had come up with was not a way of making money infinancial markets, but a mathematical way to structure deals to avoid losing moneyby buying a mix of securities to cancel out the risk in virtually any deal. Ironically,most rocket science is used not to make money, but to keep from losing it. Thisis by far the most important role of rocket science today.

"Rocket science is primarily a form of insurance for organizations that want tocontrol their exposure to risk," explains Darrell Duffie, professor of finance atStanford University's Graduate School of Business. "Black and Scholes showedhow to compute the value of an option or how to synthetically create an optionusing other securities and thereby hedge almost any risk. The result has been anexplosion of activity in the foreign currency and bond markets in the last 20 years,making the volume of equity exchanges like the New York Stock Exchange seemlike a drop in the bucket."

What Black and Scholes did was to literally make possible much of the marketexpansion of the 1970s and '80s. Remember those were times of high inflation andprice volatility. If you were trying to run an airline in the middle of the OPEC oilembargo, protecting your business from a catastrophic rise in jet fuel prices wasvitally important. The work of Black and Scholes showed precisely how to gainthat safety by hedging with heating oil futures or some other instrument that wouldrise in value along with jet fuel, providing the profits needed to cover higher fuelcosts. And since the cost of futures or options is always less than actually buyingthe underlying commodity, this maneuver both provided insurance and left moremoney for running the business.

The same effect held for the commodities firms that sold those heating oil futures:by following Black and Scholes and hedging with some specific combination ofoptions or futures, the firms' risk could be cut nearly to nothing, while still makinga profit on the deal. "Black and Scholes made it possible to do more business withthe same underlying capital," said Duffie. "Or for firms that are willing to takesome risk, it provides a longer limb to climb out on."

And climb out on that limb we have, with swaps alone running at an annual volumemeasured in TRILLIONS of dollars.

Where computers come into this is in the bewildering array of synthetic orderivative securities that became progressively available after 1973. Warrants(options sold or granted by a company on its own stock), swaps (privateagreements between two companies to exchange future cash flows), swaptions(options on swaps), caps (a maximum loan interest rate supported by a portfolio offixed-price options to buy securities), floors (a lower limit on loan interest ratessupported by options to buy discount bonds), collars (a combined maximum andminimum interest rate, supported by a long position in a cap and a short position ina floor). It takes a computer just to keep track of all the complex obligations andtimeframes, and even then it sometimes doesn't work.

During 1987, according to an account in the June, 1991 issue of Moody's SpecialComment, Merrill Lynch & Co., Inc. created a synthetic, or derivative security bysplitting the principal and interest portions of a portfolio of mortgages. The planwas to hold one part of the package and sell the other, but some trader at Merrill --definitely not a computer — SOLD THE WRONG PART, incurring a trading lossfor the company of $377 million. Oops.

Bringing the story more into the present, the big foreign exchange trading lossesincurred by Wall Street firms in the last week because of the Russian financialcrisis came not from derivative securities, themselves, but from traders efforts touse those securities to take higher risks and make bigger profits, rather than to limitrisk.

The most advanced investment firms use their computer systems to coordinate theportfolios of all their traders in a common database, because trades are sometimesoffsetting, especially in large firms with thousands of customers. That is, onecustomer may want to buy a synthetic security that is hedged by selling a certainsecurity, while another customer may want a deal that requires buying that samesecurity. Offsetting the trades means that no securities actually have to be boughtor sold and the profit to the firm is even higher.

Enter the Internet. Wall Street was already in a computer arms race long beforethere was an E-Trade, but now we have an instantaneous global network linkingmost of the active stock, bond, and commodity traders in the world. On your deskand mine sits a computer equivalent to the mainframe used by Black & Scholes 25years ago. Within the Internet, itself, there is more server capacity than can bemeasured linking firms holding either in fact or in trust most of the world'ssecurities. This is the perfect market and the perfect opportunity to create anarmada of Internet startups offering baskets of securities to hedge against ANYrisk. Whether you are a wheat farmer, a little league coach, or Boris Yeltsin, theInternet ought to be offering you a rocket science avatar, a hedge instrument that,for a small fee, effectively eliminates speculation and emotion and allows theunderlying fundamentals to take the market either up or down. The people whomake it happen will make Bill Gates look like a Fuller Brush salesman.

Until this idea came to me, it was my hope that the biggest casualties of theInternet would be the Harvard MBA's. With a spreadsheet they trashed the workerbees. With the Internet I figured, companies would need more worker bees andfewer corporate suits. But this rocket science idea kind of trashes that. Oh well.