This article appeared in the July 1999 issue of the
© 1999 Massachusetts Bar Association
Philippe G. Koenig, a patent and trial attorney, practices in the Winchester firm of Pittas Koenig. He represented theprevailing party, Signature Financial Group, in the litigation giving rise tothe decision in State Street Bank and Trust Co. v. Signature Financial Group,discussed in this article. The opinions expressed are those of the author anddo not necessarily represent those of Pittas Koenig or its clients.
In State Street Bank & Trust Co. v. SignatureFinancial Group, Inc.,
149 F.3d 1368, cert. denied, 119 S. Ct. 851(Fed. Cir.1998), the Federal Circuit Court of Appeals vaporized the cloud that hung overthe patentability of pure software patents, and endorsed the patenting ofmethods of doing business, computerized or not. Its opinion has beencharacterized in the ABA Journal as "being to patent law what RachelCarson's Silent Spring was to the environmental movement," (Aquino,J., Patently permissive: PTO filings pile up after ruling expands protectionfor business and Net software, ABA Law Journal,
May 1999, p. 30). Widelydiscussed in patent law circles and in software publications, the State Streetdecision is hailed by some as sensible and overdue, and condemned by others asputting the public and businesses at risk of attack by unscrupulous holders ofdubious business method patents, covering "obvious" computerizationof ordinary business methods. It is accordingly a good vehicle for a discussionof the application of patent law to business software and business methods, andof the patent system as a whole. It is a case the Massachusetts bar needs toknow, and will often need to bring to the attention of clients.
THE SIGNATURE PATENT Signature Financial Group is a Massachusetts firm engagedin providing consulting and management services to the financial servicesindustry. In the early 1990s it developed a novel form of mutual fundmanagement system, which it called "Hub and Spoke," and it obtainedU.S. patent No. 5,193,056 to protect the software it developed to implementthis form of mutual fund management structure on a personal computer.
In a Hub and Spoke mutual fund structure, the assets ofseveral mutual funds, the "Spokes," are pooled into a commoninvestment portfolio, the "Hub," that is legally a partnership ofthe several funds. This structure makes possible the sharing of administrativeand management expenses in a manner that creates economies of scale, and italso results in tax advantages. The administration of a Hub and Spoke mutualfund system requires several accounting operations to be performed on a dailybasis in order to satisfy regulatory requirements. The '056 patentdiscloses and claims software routines for making the necessary calculations,such as allocating daily changes in portfolio value, and Hub asset managementexpenses, to the several funds on a daily basis, thereby enabling the dailycalculations of share values for each fund. As a result, investors are able totrade in shares of each of the participating funds, assured that the price paidfor its shares correctly represents their aliquot value of the pooled assetsheld by the Hub, notwithstanding daily changes both in the value of theportfolio and in the shareholdings of each participating fund. Absent thecomputerized accounting capability provided by the '056 software, itwould not be practically feasible to set up and operate a Hub and Spokemulti-fund system.
Signature entered into negotiations with The State StreetBank & Trust Co., a giant in the mutual fund management industry, tolicense the '056 patent to State Street, but these negotiations brokedown. State Street then filed a declaratory judgment action in the federaldistrict court in Boston, seeking to have the Signature Financial patentdeclared invalid as directed to non-statutory subject matter. That is, itsought a ruling that the '056 patent, in effect a pure software patent,related to a kind of innovation which falls outside the categories of subjectmatter for which patents may be granted. Signature, for its part,counterclaimed and charged State Street with infringing the '056 patent.
By motion for summary judgment of patent invalidity, earlyin the case, State Street argued that a software patent like the '056,which is capable of implementation on any personal computer, discloses andclaims only "mathematical algorithms," which the U.S. Supreme Courthad ruled unpatentable a generation ago, and that Signature was attempting toobtain a patent monopoly on a business method, long recognized as anunpatentable category of subject matter. Whatever its merits, State Streetcontended, the '056 patent could not be valid as a matter of law. Thetrial court agreed. It ruled that the '056 patent fell outside the categoriesof patentable subject matter set forth in 35 U.S.C. 101, and it stated thatupholding the patent would enable Signature to monopolize an entire economicsector. 927 F. Supp. 502 (D. Mass. 1996).
On appeal by Signature, the Court of Appeals for theFederal Circuit reversed, in a unanimous panel opinion authored by legendaryjudge, Giles S. Rich, the eldest and longest serving jurist in federal courthistory. The court ruled that essentially all software-related inventionsconstitute "patentable subject-matter," meaning thatsoftware-related subject matter is entitled to the same consideration forpatentability as more traditional subject matters. The Supreme Court deniedcertiorari in January, and the case has been remanded to the trial court forfurther proceedings, still in the discovery stage.
The proceedings to date in the State Street case have notrequired consideration by the courts of the particular merits ofSignature's invention, and the debate occasioned by the decision istherefore independent of the specific merits of any particular innovation. Thebroad question at issue has been, simply, should the patent system entertain,and reward with patents, significant inventions in pure software and in ways ofdoing business. The affirmative ruling of the Federal Circuit in State Street,and the Supreme Court's subsequent denial of certiorari, appear to haveput the legal debate to rest, but left the business community facing importantpractical questions: does State Steet impact our business? Should we considerpatent protection for our business innovations? Is there a risk that ourcompetitors will develop, and patent, business methods that we will need inorder to compete effectively? The answer to all three questions is,emphatically, yes.
SOFTWARE PATENTS' CHECKERED HISTORY The State Street decision has fueled a good deal ofcontroversy, but primarily among commentators outside the patent community.Among patent specialists, a minority view State Street as disregardingappropriate limitations which the Supreme Court had established on thepatentability of software, but most see a refreshing landmark in the slowprogress of the courts towards an understanding of computer technology and acorrect application of classic principles of patent law to computer-relatedinventions.
One wholly new to the issue of software patents might wellask, what's the problem, why shouldn't software inventions betreated by the Patent Office like all other inventions? Why indeed? The answer,simply, is that in the 1970s the Supreme Court, then wary of all patent"monopolies" and unfamiliar with the nature or value of software,created a seemingly blanket exclusion from patentability for pure softwareinventions, and it has taken the lower courts 20 years to find their way out ofthe box.
Software patents are in fact quite old. Candidates forearliest honors would include U.S. Patent No. 2,552,629, issued May 15, 1951,to Bell Labs for its Error-Detecting and Correcting System, only a few yearsafter the construction of the first electrical computing machines. One couldeven argue that Samuel B. Morse was granted the first software patent in 1840,for the coding system he claimed in these words: "I claim, as myinvention, the system of signs, consisting of dots and spaces, and of dots,spaces and horizontal lines, for numerals, letters, words or sentences,substantially as herein set forth and illustrated, for telegraphicpurposes."
The Supreme Court upheld Morse's patent in 1853, inO'Reilly v. Morse, 56 U.S. 62 (1853). It did not revisit software patentsuntil 1972, when, in Gottshalk v. Benson, 409 U.S. 63 (1972), the court ruledinvalid an issued patent that claimed a computerized method for convertingdecimal numbers to binary numbers. The court stated, correctly, that the patentappeared to "wholly pre-empt the mathematical formula and in practicaleffect would be a patent on the algorithm itself," and held that any suchpre-emption rendered the invention unpatentable, as in effect an attempt topatent a mathematical truth, like the Pythagorean triangle.
To distinguish unpatentable computer-related inventionsfrom those which might be patentable, the court in Benson chose to draw abright-line distinction based on physical factors: "Transformation andreduction of an article ‘to a different state or thing' is the clueto the patentability of a process claim that does not include particularmachines." (Id. at 70). Pure software, under this test, is unpatentable,even when it is what we now call "application software," which hasbeen developed to solve a real-world, practical problem, as opposed to"algorithms" that are intellectual exercises in pure math.
The Benson approach was reaffirmed by the Supreme Court afew years later, in Parker v. Flook, 437 U.S. 584 (1978), an equallyrestrictive decision that appeared to seal the fate of patents directed to puresoftware. These Supreme Court decisions caused a good deal of consternationamong patent professionals, including the judges of the then Court of Claimsand Patent Appeals, a specialized appellate court that heard all appeals fromactions of the Patent Office and was accordingly well versed in classic patentlaw principles.
Following Benson and Flook, the CCPA made valiant effortsto develop distinctions that would salvage the patentability of worthy softwareinventions. These efforts resulted in the development of an arcane and unwieldyjurisprudence, culminating in the so-called Freeman-Walter-Abele test, namedafter three cases decided by the CCPA. Under the FWA test, one first determinedwhether any element of the patent claim at issue constituted the recitation ofa "mathematical algorithm." If so, the analysis proceeded to adetermination whether the mathematical algorithm was "applied to or limitedby any physical elements or process steps." In practice, efforts toimplement the FWA test generally confused courts, commentators and litigantsalike. It also became increasingly obvious to observers that, to pay obeisanceto questionable Supreme Court precedents, the courts had developed apatentability standard that was divorced from the practical utility of theinventions at issue, which is the classic touchstone of patentability.
PATENT OFFICE ADVOCATES FOR SOFTWARE PATENTS The U.S. Patent and Trademark Office, which is chargedwith protecting the interests of the public by guarding against the issuance ofunworthy patents, has consistently favored including software within thecategories of patentable subject matter. The Patent Office began issuingsoftware patents in the 1950s, and issuing patents on software processesdirected to business systems in the late 1970s. One such patent, issued toMerrill Lynch, Pierce, Fenner & Smith in the late '70s, claimed acomputerized system for combining several types of accounts, including moneymarket funds and charge/checking accounts. Long before the State Streetdecision, that Merrill Lynch patent was upheld in a 1983 ruling by the FederalDistrict Court of Delaware, in Paine, Webber, Jackson & Curtis, Inc. v.Merrill Lynch, Pierce, Fenner & Smith, Inc., 564 F. Supp. 1358 (D. Del.1983), but the case settled before appeal. Still the Patent Office, includingthe patent examiner who allowed Signature's '056 patent, relied onthe Merrill Lynch decision to support the grant of software patents.
The Patent and Trademark Office increased its advocacy ofsoftware patents under the direction of Commissioner Bruce Lehman, who wasappointed by President Clinton in 1993 (and who left the office in 1998). Acopyright lawyer by trade, and the only modern patent commissioner not a patentlawyer, Lehman brought to the Patent Office a clear understanding of theenormous significance of computer software to the world economy, and acommitment to maintaining U.S. supremacy in this critically important economicsector. Aware, as a copyright attorney, of the shortcomings of copyright lawfor protecting software, Lehman led the Patent Office to issue"Examination Guidelines for Computer-Related Inventions," 61 Fed.Reg. 7,478 (March 29, 1996). Though acknowledging the need to reconcile PatentOffice policies with the guidance of the courts, including those difficultSupreme Court precedents, the guidelines expressed a strongly favorable view ofsoftware inventions as patentable subject matter.
FEDERAL CIRCUIT TAKES CHARGE The CCPA was eliminated in 1982, upon the creation of theCourt of Appeals for the Federal Circuit, which succeeded to the CCPA'sjurisdiction over appeals from the Patent and Trademark Office and appeals fromthe U.S. Court of Claims. In addition, the Federal Circuit was grantedexclusive jurisdiction over patent appeals from all federal district courts,with a mandate to create uniformity in patent decisions. Many CCPA judges,including Giles Rich, moved onto the new court.
A way out of the box in which Benson and Flook hadseemingly trapped software patents was suggested, in 1992, in a concurringopinion by Federal Circuit Judge Randall Rader (then a new arrival on thebench), in Arrhythmia v. Corazonix, 958 F.2d 1053 (Fed. Cir. 1992). Rader notedthat the Supreme Court itself, in 1981 in Diamond v. Diehr, 450 U.S. 175(1981), had used broad language in characterizing the scope of the statutoryprovision, 35 U.S.C. 101, that limns the categories of patentable subjectmatter, and had thus cast a shadow on its earlier and more restrictiveopinions. Others noted that in Diamond v. Chakrabarty, 447 U.S. 303 (1980), nota software but a biotech case, the Supreme Court had likewise employed verybroad language in characterizing the categories of patentable subject matterunder Section 101.
Chakrabarty made two trips to the Supreme Court. In thefirst, a cursory lower court ruling of patentability had been reversed and thematter remanded to the CCPA. On remand, then CCPA Judge Giles Rich proceeded towrite a lengthy opinion, akin to a lawyer's brief, vindicating hiscourt's initial ruling of patentability, and expounding on the broadscope that the 1952 patent statute intended for the enumerated categories ofpatentable subject matter. Rich, a principal author of the text of the 1952act, knew whereof he wrote. On renewed appeal to the Supreme Court his viewprevailed, and the Supreme Court set forth a statement of patentable subjectmatter that would determine the result in the State Street case:"everything under the sun made by man."
In the mid-1990s the Federal Circuit handed down a seriesof decisions that used the broad language of the Supreme Court in Chakrabartyto chip away at the restrictive language of the earlier Benson and Flook decisionsof the Court. Three 1994 decisions, notably In re Alappat, 33 F.3d 1526 (Fed.Cir. 1994) (en banc), indicated that a clear majority of the Federal Circuitwas intent on scuttling the "physical transformation" requirementof Benson, if need be by relying on the fact that all software, when run on acomputer, occasions physical changes in the electronic status of computercomponents. All the court needed, to complete the job, was a pure softwarecase.
THE STATE STREET DECISION In considering State Street's motion for summaryjudgment of patent invalidity, in 1996, the U.S. District Court forMassachusetts turned a deaf ear to Signature's arguments that currentFederal Circuit decisions, and notably Alappat, demonstrated a sea change inthe treatment of software patents. The District Court undertook a classicapplication of the Freeman-Walter-Abele test, and ruled that the Signaturepatent failed that test. It concluded that:
[The] change of one set of numbers to another, withoutmore, is insufficient to confer patent protection. The invention does nothingother than present and solve a mathematical algorithm and, therefore, is notpatentable.
927 F. Supp. 502, 516 (D. Mass. 1996).
The District Court ruled also that the Signature claimswere not directed to patentable subject matter because they were directed to amethod of doing business.
Given the trend in the Federal Circuit to liberalize thepatentability of software, most knowledgeable observers were not surprised thatit reversed the District Court's ruling of patent invalidity. What didsurprise was the adamancy of the reversal, including the unconditionalabandonment of the "physical transformation" element as the"key" to software patentability, and the equally unconditionalabandonment of the "business method" exception to patentability.State Street
presented the court with its first "pure software"case, requiring it to confront, head on, the Benson
principle thatpatentability required some transformation of a physical nature. The closestprevious case, Alappat,
permitted the argument that the smoothing of jaggedelectrical signal traces on a computer screen constituted a physicaltransformation. State Street
, dealing with software that manipulated financialdata, allowed no such out. The court bit the bullet, in the most controversiallanguage in the opinion:
[W]ehold that the transformation of data, representing discrete dollar amounts, bya machine through a series of mathematical calculations into a final shareprice, constitutes a practical application of a mathematical algorithm, formulaor calculation, because it produces "a useful, concrete and tangibleresult" — a final share price momentarily fixed for recording andreporting purposes and even accepted and relied upon by regulatory authoritiesand in subsequent trades.
The court then concluded:
[C]laim1 is directed to a machine programmed with the Hub and Spoke software andadmittedly produces a ‘useful, concrete and tangible result.' Thisrenders it statutory subject matter, even if the useful result is expressed innumbers, such as price, profit, percentage, cost, or loss.
With these words the "physical transformation"test of Benson
sank at last into oblivion, and a new standard took pride ofplace: whether the software at issue produces a "useful, concrete andtangible result."
The new focus on the production of "a useful,concrete and tangible result," eliminates any need for the multi-stepFreeman-Walter-Abele
analysis. Rich dispatched the FWA test (and, implicitly,the remnants of Benson
and Flook) with a wave of Chakrabarty
: "AfterDiehr and Chakrabarty
, the Freeman-Walter-Abele
test has little, if any,applicability to determining the presence of statutory subject matter."
BUSINESS METHOD EXCEPTION BURIAL
It will be recalled that, as an independent ground forinvalidating the Signature patent, the District Court had invoked the so-called"business method" exception to patentability. The business methodsexception is usually traced to a 1908 case from the 2nd Circuit Court of Appeals,Hotel Security Check Co. v. Loran Co., 160 F.2d 467 (2d. Cir. 1908), which hadheld invalid a patent claiming a novel method of hotel bookkeeping. In moderntimes this doctrine has been invoked rarely, and usually in dicta; mostrecently it had been lambasted by Judge Pauline Newman, concurring in Allapat,as lacking any discernible merit. In the State Street
appeal the FederalCircuit buried this doctrine, conclusively. Rich described the business methodexception as an anachronism, which had not survived the enactment of the 1952Patent Act (which rewrote patent law in major respects). Since the enactment ofthe 1952 act, the court held, "business methods have been, and shouldhave been, subject to the same legal requirement for patentability as appliedto any other process or method."
STATE STREET IMPACT State Street
has reveived generally favorable reviewswithin the patent bar. In the public media and in tech publications, however,some commentators have condemned the State Street
decision as opening thegates. In some quarters the decision has been condemned as opening the gates toan avalanche of unworthy software patents, likely to hamstring legitimatesoftware development and to hold the public hostage to baseless demands fortribute, notably in the context of Internet commerce. How legitimate are theseconcerns? Does the Federal Circuit decision make good sense, and good patentlaw?
In the judgment of this author, the core of the StateStreet holding accords with the central objective of the patent system, whichis to provide an incentive to invest resources towards the discovery of novelways to produce "useful, concrete and tangible results," that is,inventions of value to the public, whatever their physical form. There is noreason to discriminate against inventions that are directed to software or tobusiness methods, and the prior limitations on the patenting of such inventionslacked any basis reconcilable with classic patent law principles.
Opponents of software patents, and of patents generally,often contend that many patented inventions are trivial and don't deservea patent. Not uncommonly, people who, in fact, desire to use, free of charge,the "trivial" inventions of others, advance this contention. Stillit is true that the great majority of issued patents disclose and claim minorinventions, worthy but narrow in scope. Call them trivial if you will, but thepatenting of such inventions benefits the public, and causes no harm. Thepublic is free to disregard "trivial" patented inventions for theproducts and services of a competitor, who in turn is free to developalternative, non-infringing means to carry out the function to which a"trivial" patent is directed. Indeed, it is an important benefit ofthe patent system that potential users of patented invention are spurred, inorder to avoid royalties, to develop alternative and sometimes better solutionsto the problem addressed by the patent. To that end, every patent grantedresults in a public record, which contains a detailed description of theinvention, an analysis of the problem it addresses, and often a good review ofthe prior art, all of which may be consulted free of charge by anyone with amodem.
As an aside, it is a common misunderstanding that there issomething unfair about people obtaining patents for unimportant inventions.That notion obsessed Justice William O. Douglas, who believed patents should bereserved for inventions that evinced "the spark of genius," therebydemonstrating he hadn't the faintest idea what the patent system is for.MacArthur grants and honorary degrees serve to reward the "spark ofgenius," but the patent system has the more pragmatic purpose ofproviding an incentive to the development of things and schemes that improveour lot, in small ways and large. The logic of that system applies equally topure software and to "business methods" as to mouse traps andinventions that positively glow from the sparks of genius.
What about the risk, however, that broad software patentswill be granted, greatly enriching their inventors and enabling them tomonopolize entire industry sectors? Of course no such patent should issueunless the applicant has satisfied the fundamental patentability tests ofnovelty and non-obviousness that are applied to all inventions. Specifically inthe context of software, and in light of the wondrous capabilities that arealready in use or disclosed in the literature, future inventions that earnbroad patent protection are likely to be such as to open broad new capabilitiesfor computer users. Assume, for example, a pure software invention disclosingan "algorithm" that enables a 500 fold increase in data compressionpower, enabling movies and their soundtracks to be readily transmitted in realtime over any 56k modem. Is it not fair for the inventor of such a compressionscheme, in exchange for disclosing the method in detail, to be provided with ameans to profit from her invention, through patent licensing?
The economics of the patent system serve to weed out weakpatents quite effectively, whatever their subject matter. It is expensive toobtain a patent, and expensive to maintain it. The fees of patent agents andattorneys average about $4,000 for simple patents, and can easily rise morethan $10,000 for complex inventions. The fees payable to the Patent Office arein the order of $1,000 to file for, and obtain the issuance of, any patent;following issuance, keeping the patent in force requires the periodic paymentof maintenance fees, which presently total about $3,000 over the life of thepatent. If foreign patent protection is desired, increasingly the case forsignificant inventions with good prospects of international marketability,these expenses are greatly multiplied. Thus, keeping a single patent in forcein the dozen most industrialized countries will likely result in average yearlycosts, in maintenance fees alone, exceeding $10,000 per patent per year. Fewindividuals or firms will incur the cost of maintaining a portfolio of a dozenpatents, at a yearly cost exceeding $100,000, absent a sincere belief that thepatents cover genuine innovations that are likely to prove their value in thecrucible of commerce. Add the enormous costs of patent litigation, and itbecomes obvious that enforcing bad patents as a business strategy rarely leadsto riches.
The patent system is not foolproof, and unscrupulouspatent holders occasionally do attempt to assert weak patents against largescale activities that don't really infringe. However, there is no reasonto expect such conduct to be more prevalent with software patents than othertypes of patents. Furthermore, the same Federal Circuit judges who haveexpanded the range of patentable activities have also tightened therequirements that must be satisfied in order to establish infringement. A lineof recent Federal Circuit cases, including Chiuminatta Concrete Concepts, Inc.v. Cardinal Industries Inc.,
46 USPQ2d 1752 (Fed. Cir. 1998), and Signtech USALtd. V. Vutek Inc.,
50 USPQ2d 1372, 1999 US App. LEXIS 6341 (Fed. Cir. 1999), hasincreasingly limited the ability of patent holders to assert, successfully,interpretations of their patent claims that go appreciably beyond the specific,articulated teachings of the patent.
IMPLICATIONS FOR THE FUTURE
The State Street
decision, in broadly recognizing thepatentability of software inventions, and of inventions directed to businessmethods, has done no more than bring these important fields of invention withinthe same statutory scheme that applies to all other inventions. Like all other inventions,any application directed to software, or a business method must still, toqualify for a patent, satisfy the statutory requirements of novelty andnon-obviousness. There is no reason to doubt that these standards, whichoperate satisfactorily in other contexts, will not serve their gating functionsequally well for inventions directed to software and business methods.
The State Street
ruling does, however, create newopportunities for inventors of novel business methods, including new forms of insurancecoverage, models for Internet auctions, business valuation methods, and myriadother commercial ventures, to seek patent protection. As has been true for moretraditional subject matter, the availability of patent protection for novelbusiness methods will unquestionably result in the development, and the publicdisclosure, of many more clever and worthwhile inventions than would otherwisebe the case. Gazing into the crystal ball, this aspect of the State Street
decision is likely to have the greatest initial impact, and a substantialincrease in patent filings directed to business systems, notably in the contextof Internet commerce, is already being reported by the Patent Office.
The United States leads the world in developingapplications software, with some 80 percent of the commercial value of softwaresales worldwide reportedly attributable to software developed in this country.Yet the sometimes enormous investment required to develop breakthrough softwareis subject to copying, and misappropriation, with a few clicks of a mouse on a$600 computer. As Commissioner Lehman clearly recognized, it is in our nationalinterest to provide strong legal protection against the uncontrolled copying ofnovel software, both in the United States and in foreign countries. The StateStreet decision serves both to provide software with patent protection in theUnited States, and also assists in the effort to encourage the adoption ofpatent policies in other countries that will further protect U.S. developed softwarefrom misappropriation.
Author's note: Judge Giles Rich, the author ofthe
State Street decision, died June 12 at the age of 95. He was the oldestsitting and longest-tenured federal judge in history.