High interest rates have inspired innovative and often ingenious new twists not just in the home and mortgage markets, but recently in traditional business debt instruments as well. Companies have found that dressing up debentures with various trend-sensitive mechanisms can create a readier reception, may reduce borrowing costs, and can sometimes take the edge off long-term fixed-rate commitments.

Predictably, one of the first new turns taken by corporate debt was toward floating rates. Then came original deep-discount bonds, which paid modest interest but sold substantially below par. More recently, zero coupon bonds, sprung on the market in March 1981 by J.C. Penney Co. and quickly copied by a number of other borrowers and bond-fund syndicators, have become the objects of investment analysis. As its name suggests, a zero coupon bond, or "zero" for short, pays no interest at all. Instead, the issuer sells it at a large discount so that when it is redeemed at par a few years later, the investor's profit is the equivalent of having received compounded interest while the bond was maturing. The buyer of a zero coupon bond thus locks in interest that in effect is reinvested at the same constant rate; with a real interest-payer, the interest can be reinvested only at prevailing rates.

The concept of issuing debt without paying interest is in itself nothing new; short-term Treasury bills have been sold that way for decades. But zeroes are a novel idea for corporations and corporate-bond investors. Because zeroes have been accorded a surprisingly warm welcome in what otherwise has been an ice-cold bond climate, businesses are finding that they can borrow over longer terms at rates that are lower than those of conventional debt instruments. One drawback, however, is that the net amount a borrower receives is far less with a zero than through customary financing. To get the same amount of money up front, a company often has to take on three or four times as much debt, thus affecting sinking-fund provisions.

There are disadvantages to a buyer, too. For example, one of the main differences between a zero coupon and a fixed-interest bond selling at a "natural" deep discount in a depressed market -- such as, say, an AT&T 2 7/8 of '87 at $71 -- is the tax treatment. The Internal Revenue Service rightly takes the position that yearly interest must be reported even though none is tangibly paid. That phantom income is taxed at ordinary rates, whereas the profit made when the AT&T bond matures is a capital gain and therefore is taxed at a lower rate. (The popularity of zeroes has taken the Treasury by surprise, too; it recently proposed new tax rules for the corporation "paying" the interest that reduce early-year deductions.) Zeroes do make sensible holdings in IRA or Keough accounts, however, since the interest does not have to be reported.

All else being equal, anyone who wants to cement a high interest rate would probably be better off shopping in the open market. For instance, there's no guarantee that interest rates won't go higher, rather than lower, during the life of a zero. If that happens, a bondholder would rather have a conventional bond that pays full interest, since he or she could reinvest the periodic payments at the higher rates. With a zero coupon, of course, there are no periodic payments. By the same token, a zero has no return-of-investment point, either. Where a fixed rate of 15% compounded annually on $100 will yield $100 in five years, there is no such insurance with a zero -- it's all or nothing many years away, and what if the issuer goes belly up in the meantime? Thus bond rating is a particularly important concern with zeroes. Merrill Lynch, Pierce, Fenner & Smith Inc., one of the chief underwriters of publicly placed debt, is seeing many zeroes salted away in retirement accounts. "That's serious investing," warns Edwin H. Hall Jr., Merrill Lynch's director of investment products marketing. "You have to make sure they're quality issues. There's starting to be a lot of junk out there."

Because they are considered relatively safe, such zeroes as Aaa-rated IBM Credit Corp. can yield less than straight-interest bonds upon issue, and they often move quickly to premiums in the aftermarket. An IBM Credit Corp. 7-year zero floated last year, for example, was priced at $39.16 to return 13.81%; within a few months it was bid at $49.75 and paid 11.75%. A J. C. Penney 10-year zero that came out at $25 to yield 12.98% was bid up in the aftermarket to $26.75.

Against this possibility of somewhat lower return should be weighed the favorable circumstance that, for all practical purposes, a zero coupon bond is noncallable, compared with conventional debt instruments, which may be called (unless there are provisions against doing so) whenever it is advantageous for the issuer to refinance at lower rates. It is unlikely that a company would call a zero, inasmuch as it would have to pay off the entire note, future interest and all. On the other hand, it is also unlikely that a deep-discount bond like the AT&T '87 would be called; any corporation would be delighted to pay only 2 7/8% until the end of time.

Since the advent of zero coupons only last year there have been even more fanciful devices aimed at currying public favor. One is a put, an option to sell at a given price, that allows a debenture holder at certain intervals either to cash the bond in -- to "put" it to the issuer -- or to keep it. A holder would choose to cash in the bond if interest rates had risen and better yields were then available. The put offers protection against a rise in interest rates, and therefore a decline in the market price of the bond, while granting the investor the opportunity to refinance.

With this type of aggressive marketing taking place, financial institutions have entered the capital arena with a vengeance. Given the competition, banks had to come up with something dramatically different. They borrow frequently, but their issues generally are not that popular. Chase Manhattan, Chemical New York, and Republic New York have each introduced a preferred stock with a floating rate. At Chase Manhattan the rate is adjusted every three months to 50 basis points above the highest of three current rates: 3-month Treasury bills, 10-year Treasuries, or 20-year Treasuries. The stock is nonredeemable and carries a floor of 7 1/2% and a ceiling of 16 1/4%. Demand for such issues has been strong. Observes Merrill Lynch's Hall: "A floater is the best of all worlds."

Maybe so, but new territory is still being carved. This May, Chase Manhattan sold a complicated package that included a conventional 10 1/2-year note at 15 1/2%, a 10-year zero priced at $25.293, and a put. The innovative part was that the put contract allowed the company to put the stock to the buyer -- a revolutionary man-bites-dog gambit that reportedly caused Bache Halsey Stuart Shields Inc. to drop out of the underwriting group.

Essentially, under the plan an investor buys a unit of the fixed-coupon 10 1/2-year note plus the put contract, which obligates him to purchase Chase common prior to maturity in May 1992, at a minimum of $83.34; he must keep sufficient funds in his account to be able to do so, but such funds can be in the form of the zero bond. He then can sell the fixed note and get his money back, ending up holding the put obligation; for about $2,500 he can buy the zero for collateral. If the stock rises markedly, he can make money on the put; if not, well, he owns expensive stock. At the time of issue, Chase was around $55.

The permutations of such a play are complex, but the idea at least suggests the kinds of clever games that will emerge in financing instruments in the future. Some of these, the more imaginative experts predict, will include such debt repayment offers as the return of enough dollars to buy a certain amount of oil, and a bond indexed to a large choice of rate-setting statistics such as the Consumer Price Index. An investor alert to how and where the good hands are being dealt will be able to play these games profitably.