When it comes to creating new and amazing products, the guys who concoct new securities on Wall Street are a lot like the guys in laboratories who splice genes.

The idea of gene splicing, of course, is to combine all the good things you want in a single organism and leave all the bad stuff behind. The new-product boys on Wall Street do the same thing, but they call it "finance" instead of "biotechnology." And it pays a lot better.

The latest product to emerge from Wall Street's gene splicing vats is something called the SFP Pipeline Holdings variable rate, exchangeable debenture of 2010. These little beauties, about $210 million of them, are being peddled by Goldman, Sachs & Co. for Santa Fe Pacific Corp. of Chicago, which owns pipelines, the Santa Fe railroad and various other things.

These debentures are a marvelously inventive piece of paper, showing how you can take an investment that has a limited market because of tax problems, slice out the problems, splice in a couple of bells and whistles, and voila! You have a whole new security that appeals to a whole new universe of buyers.

In the process, Santa Fe stands to take in about $200 million in a tax-advantaged way, and Goldman stands to make a couple of million bucks in underwriting fees, not to mention the profit it stands to make over the years by making a trading market in this security. Pretty slick.

The game goes like this. Santa Fe wanted to get some money out of its 44 percent ownership of Santa Fe Pacific Pipeline Partners L.P., which owns pipelines that carry petroleum products to parts of the West and the Southwest. Santa Fe had sold the other 56 percent of the pipeline company to public investors in late 1988.

So why didn't Santa Fe just sell its 44 percent of the pipeline company and go home? Because, for starters, it would produce a huge capital gain, and thus a big tax bill for Santa Fe.

And, more important, it couldn't get a decent price for Santa Fe's 44 percent, for reasons that have to do with the fancy footwork that enabled Santa Fe to score big selling the first 56 percent. This whole thing is an example of why tax lawyers get rich. Rather than selling stock equal to 56 percent of the pipeline company, Santa Fe sold 56 percent of the company by peddling units of a master limited partnership. These units trade on the New York Stock Exchange like a stock, but they are really a partnership.

Why bother with these contortions? Because making the pipeline company a partnership rather than a corporation eliminated the pipeline's corporate income taxes and allowed Santa Fe to sell investors a security with a very high, tax-advantaged payout. The distribution currently is almost 10 percent of the units' selling price. Only about 20 percent of the partnership's distribution is taxable, though that percentage will rise in a few years. (You may have to pay tax on some of the tax-advantaged money when you sell, but meanwhile, it's like having an interest-free loan from Uncle Sam.)

The tax advantages and high payout made partnership units a much more attractive security to individual investors than regular old boring common stock would have been. Thus, Santa Fe got a higher price selling partnership units than it would have gotten selling stock.

But one investor's meat is another investor's poison. Even though partnership units offer tax advantages to individual investors, they offer tax problems for tax-exempt investors such as pension funds. The reason: Partnership income is taxable to tax-exempt investors. A tax-exempt investor paying corporate income tax? Yech. It violates God's law.

Not only would institutional investors not buy Santa Fe's 44 percent of the pipeline company, individual investors wouldn't do so, either. For reasons that are incomprehensible to mortal man -- only tax lawyers understand them -- the particular partnership units that Santa Fe owns produce more taxable income than the cash they actually pay out. Sort of like a reverse tax shelter.

Enter the Goldman Sachs gene splicers.

What they have created is a debenture designed to pay the same interest that partnership units would pay. But because it's a debenture, the security produces interest income, not partnership income.

I'll spare you the technicalities and oversimplify. If everything goes as planned, anyone plunking down $1,000 for a debenture when Goldman peddles them in August has the right to exchange that debenture for partnership units worth $1,000 on the day the debentures are first sold.

In other words, if Goldman sells these things on Aug. 15 and units are selling for $25 that day, each $1,000 debenture will be convertible into 40 units on Aug. 15, 2010. The interest the debenture pays will be equal to the distributions the 40 units would pay. If the units crater, you can get your $1,000 back in the year 2010.

Where does Santa Fe get the money to pay the interest on these debentures? Easy. It still owns 44 percent of the units, even though it has committed to give them to debenture holders when the conversion day arrives. Santa Fe collects the distributions from its 44 percent of the pipeline, then turns around and pays the same number of dollars in interest. Meanwhile, Santa Fe has the use of the $200 million or so raised from the debenture sale, and it won't have to pay capital gains tax on the sale of its units until debenture holders exchange their debentures for partnership units 20 years from now.

So what you have is a security that is a bond, that pays like a partnership and that lets Santa Fe peddle its tax-disadvantaged partnership units as if they were tax-advantaged units. Talk about financial gene splicing. Maybe business schools should teach biology as well as finance. Allan Sloan is a columnist for Newsday in New York.