Citibank Na v The Comptroller General of Patents

This is an appeal from the decision of Mr. Peter Marchant, Deputy Director of the Patent Office, acting for the Comptroller, in which he refused an application by Citibank NA for a patent under application number 0019522.2. The ground of refusal was that it fell foul of section 1(2) of the Act on the footing that the alleged invention related to a scheme, rule, or method, for doing business and/or a mathematical method. He also ruled that it did not involve a technical contribution. I shall deal in due course with the extent to which the law has shifted from the law as it appeared to be to the hearing officer as at the date when he decided the matter, that date being 4^th March 2005.

The background to the application for the patent is as follows. Amongst those who trade in financial derivatives there is a need to manage the risks associated with that trading. Various techniques have been developed in order to measure those risks and various statistical and probabilistic techniques have been used for that purpose. One of the problems which is apparently faced by those devising and utilising such techniques is the integrity of the data to which those techniques are applied, and it is as to the integrity, and in particular the error-free nature or error-prone nature, of that data to which the invention is directed.

The application describes the field of the invention as follows: “The present invention relates to a system and method for measuring the financial risks associated with trading portfolios. Moreover, the present invention relates to a system and method for assuring the integrity and validity of data used to evaluate financial risk or exposure.” The application then goes on to set out the background of the invention, I have sought to describe it generally, and at page 4 of the application there is set out a summary of the invention in the following terms: “In accordance with the invention, a method and system are provided for detecting abnormalities in input data to a financial risk management system. The method includes receiving a set of input data to a financial risk management system; receiving one or more historical values, each historical value representing a calculated content from a previous set of input data; and calculating the likelihood that changes to the set of input data are the result of one or more errors by determining the information core of the input data, and performing a statistical analysis of the calculated information content relative to historical values to determine the likelihood that changes to the set of input data are the result of one or more errors.”

The claims in the application, after amendment, number 16 but it is common ground that there are only two independent claims and they are claims at 1 and 9. Claim 1 reads as follows: “1. A method for detecting abnormalities in input data to a financial risk management system, the method comprising: (a) receiving a set of input data to a financial risk management system; (b) receiving one or more historical values, each historical value representing a previous set of input data; (c) calculating the likelihood that changes to the set of input data and (d) performing a statistical analysis of the calculated information content relative to the one or more historical values to determine the likelihood that changes to the input data are the result of one or more errors.”

Claim 9 introduces a computer into the system and reads as follows: “9. A system for determining abnormalities in input data to a financial risk management system, the system comprising: a data processing server that receives a set of input data; a computer storage device containing one or more historical values, each historical value representing a previous set of input data; and one or more central processing units coupled to the computer storage device, the one or more central processing units being arranged to (i) calculate the information content of the input data, and (ii) perform a statistical analysis of the calculated information content relative to the one or more historical values to determine the likelihood that changes to the input data are the result of one or more errors.”

In due course after various internal procedures the matter came before the Hearing Officer as a result of which he gave the decision to which I have referred. He rejected the application and, as I have indicated, he rejected it on the footing that the application fell within two categories of excluded matter, mathematical method and business method. I should set out very shortly the structure of the decision. In paragraphs 1-7 he sets out the procedural history and in paragraph 8 he starts to embark on the invention. Paragraph 9 sets out claim 1, and paragraph 10 refers to claim 9 as being the other independent claim without setting it out. In paragraph 11 the Hearing Officer correctly identifies the relevant provision of the Patents Act, which is section 1(2) to which I myself will come in due course, and in paragraphs 12-15 he deals with certain matters of law which arose. Paragraph 16 contains the discussion of the points and contains the reasons for the decision at which he arrived.

In paragraph 18 he deals with the question of exclusion in the following terms: “To address the first part of the test for patentability, I think it is undoubtedly the case that the invention falls within the scope of the excluded matter listed in section 1(2). It does not sit very clearly under a single one or other of the exclusions listed; it has aspects of a scheme, rule or method of doing business and aspects of a mathematical method. But I am confident since the invention lies in the statistical manipulation of data relating to a financial transaction system, that it falls within these exclusions.”

In accordance with the law as it was then understood to be, the Hearing Officer did not stop there. He went on in the remaining paragraphs of that section, that is to say paragraphs 19-24, to consider whether despite the fact that the invention fell within those two realms of excluded matter nevertheless it made a technical contribution, and he came to the conclusion that it did not.

Paragraphs 25, 26 and 27 then contain a summary of his conclusions, his final conclusion being that the application failed. From that decision Citibank, the applicant, appeals. The basis on which the appeal was originally run was that the Hearing Officer erred in concluding that the application fell within the two heads of excluded matter to which he attributed it and also erred in finding that the invention had made no technical contribution. However, since that decision and since the commencement of this appeal the legal ground has shifted somewhat. As a result of cases to which I will refer in a moment, the proper approach to these matters can be see to have changed from that adopted by the Hearing Officer. That is not to say that the Hearing Officer is in any way to be criticised for taking the approach that he adopted. It was at the time entirely conventional. However, it does mean that the appeal before me has now been advanced on a different basis, with each counsel, that is to say Mr. Pritchard, who appears for the appellant, and Mr. Birss, who appears for the Patent Office, inviting me to apply the new approach as it appears from the cases. Mr. Pritchard says that applying that new approach and taking the material found by and available to the Hearing Officer the appeal should now be allowed. Mr. Birss takes the position that even applying the new legal basis for approaching excluded matter questions the decision of the Hearing Officer was correct.

The new approach has been emerging in a series of cases and I think it is sufficient for these purposes for me to refer to only two of them. The first is a decision of Pumfrey J in the matter of Research in Motion UK Limited vInpro Licensing SARL [2006] EWHC 70, (the “RIM” case). That case concerned a large number of points which are not germane to the matter before me but at paragraph 186 of his decision Pumfrey J said this in relation to excluded matter:

“It is now settled, at least at this level, that the right approach to the exclusions can be stated as follows. Taking the claims correctly construed, what does the claimed invention contribute to the art outside excluded subject matter?”

A similar approach can be detected in the matter of Crawford’s Application, a decision of Kitchin J, reported in 2006 RPC 11, which related to an invention for putting signs up on buses. Kitchin J, having referred to various cases relating to excluded matter, said in paragraph 11:

“At the heart of all these decisions is the consistent principle that an inventor must make a contribution to the art (that is to say the invention must be new and not obvious) and that contribution must be of a technical nature (susceptible of industrial application and not within one of the areas excluded by Art. 52(2)).”

The reference to Article 52(2) is a reference to the Patents Convention whose terms are not materially different from section 1(2) of the Patents Act 1977.

I myself had to consider the correct approach to these cases in the matter of Macrossan v Comptroller General of Patents, Designs & Trade Marks [2006] EWHC 705, and I adopted the general approach referred to in those two cases and, in particular, the RIM case. That is the approach I propose to adopt for the purposes of dealing with this appeal and indeed neither counsel before me urged me to do otherwise.

The relevant provision of the Patents Act 1977 is as follows, leaving aside immaterial words:

“1(2) It is hereby declared that the following (among other things) are not inventions for the purposes of this Act, that is to say, anything which consists of – (a) a discovery, scientific theory or mathematical method…(c) a scheme, rule or method for performing a mental act…or doing business…”

There is no novelty or obviousness point in the present case. It is to be assumed for the purposes of this appeal that the application will not fail for obviousness or want of novelty. The real focus in this appeal is whether or not the alleged invention falls within the two exclusions to which the Hearing Officer attributed it. The Hearing Officer came to his conclusion in short terms in paragraph 18 of his decision, which is in the following terms:

“To address the first part of the test for patentability, I think it is undoubtedly the case that the invention falls within the scope of the excluded matter listed in section 1(2). It does not sit very clearly under a single one or other of the exclusions listed; it has aspects of a scheme, rule or method of doing business and aspects of a mathematical method. But I am confident since the invention lies in the statistical manipulation of data relating to a financial transaction system, that it falls within these exclusions.”

The question for me is whether or not the Hearing Officer was right in his assessment that the matter fell within those two provisions. I shall take claim 1 first and consider each of the two exclusions in relation to claim 1. The first exclusion within which the Hearing Officer held the matter to fall is the business method exclusion.

In the Macrossan case I commented on the difficulties in applying this exclusion in paragraph 30 of my judgment. That case, of course, involved a different sort of invention but in paragraph 30 I said:

“The exclusion is aimed more at the underlying abstraction of business method - the market making technique in Merrill Lynch’s Application [1989] RPC 561, or the way of doing the auction in Hitachi, T258/03 (EPO Board of Appeal, 21^st April 2004). In CFPH Mr. Prescott identified joint-stock companies themselves as being methods of carrying out business - see paragraph 41 of his judgment. That has the necessary level of abstraction or method (business method). The invention in the current application does not. It is merely a facility which might be used in a business; or, to put it another way, a tool. That cannot be said of the inventions in Merrill Lynch and Hitachi…Again, in Pensions Benefit Business Partnership (T931/95) the Technical Board of Appeal held that the operation of a pensions scheme on a computer amounted to a method of doing business”. In doing so it said that, ‘All the features of this claim are steps of processing and producing information having purely administrative, actuarial and/or financial character. Processing and producing such information are typical steps of business and economic methods.’”

I went on to observe that that citation from the Pensions Benefit Business case did not amount to and could not be treated as a definition but it captured the essence of business method.

In the case before me the invention is clearly to be used in a business. One only has to look at the opening words of claim 1 in order to see that. But it does not follow from that that this is a method of doing business. In fact, claim 1 says nothing at all about how the business is to be done. All that the invention in claim 1 does is to produce some information which may or may not be used in the course of the business; it provides a form of data. That is not, in my view, a business method within the meaning of section 1(2) and it cannot be reasonably described as such. In my view, the Hearing Officer erred in so describing it.

Mr. Birss sought to say that it is within the citation from the Pension case which appears in my judgment in Macrossan and he sought to say that that is a reason for saying that this invention was indeed a method of doing business. However, when that citation is properly understood and when the present invention or alleged invention is properly understood, the alleged invention does not come within those words. The information in the Pensions Benefit case related to the administrative actuarial and financial functions of the actual business itself. That is not the case in relation to the invention purportedly described in claim 1. The purported information in this case relates to the valuation of things which might or might not ultimately be the subject matter of the business and not to the internal business function of the concern doing the business. It is, in my view, qualitatively different for these purposes. What claim 1 therefore describes is a method for acquiring information about a business product. It does not amount to a method of doing business within the meaning of section 1(2). It does not, in my view, come close to the sort of level of abstraction which in my view is necessary in order to bring a case within the business method exclusion. Accordingly, in my view, the Hearing Officer erred in this respect.

However, I still have to deal with the other basis of exclusion, which is mathematical method. Mr. Birss said that this was a mathematical method, relying on the substance of the matter. The substance involved two calculations and a statistical comparison and analysis, and that qualifies as a mathematical method. The way the invention works is as follows. One set of data is put in and then in due course a second set of data is put in. That data is not directly compared but a content analysis technique is applied to each of them. The application makes it clear that the content analysis might be in accordance with one or more of various systems for analysing content and, in particular, analyses made by entropy and in particular Shannon Entropy. When that technique has been applied to the two sets of data a statistical analysis is carried out by reference to the result of that application and comparisons are made with a view to flagging up any factors or anomalies which demonstrate that differences between the sets of data may be attributable not merely to normal and to be expected market movements and other such things, but to error in the data and, in particular, the most recent data, itself, that error perhaps arising from operator input error, from glitches in the computer, from glitches in the power supply, and so on.

That is the essence of the invention and an appropriate description of it, and that is what Mr. Birss says is a mathematical method. He was unable to support his proposition or his position on this by way of authority. At the heart of his submission was, I think, the assertion that the answer was obvious and the obviousness was in his favour. Mr. Pritchard’s approach was more subtle than that. Mr. Pritchard sought to draw a distinction between mathematical methods on the one hand and methods of calculation involving the application of mathematical methods on the other. Thus differential calculus and the decimal counting system were given as examples of the former, that is to say, as examples of true mathematical methods. This invention did not involve a method of that kind; it merely involved calculations.

The meaning of mathematical method within section 1(2) was not the subject of any authority placed before me. If one looks at the company that the head keeps in section 1(2) then Mr. Pritchard’s distinction has some attractions. The concepts of discovery or scientific theory might be thought to occupy a similar high level to the systems which Mr. Pritchard says populate the category of mathematical method. However, I do not think that the distinction drawn by Mr. Pritchard is correct. His category of methods can probably be better described as systems, if one is looking for a word, rather than methods. I do not see why the concept of mathematical method should, as a matter of policy or language, operate at the more rarefied atmosphere of calculus and not at a lower level as well when that lower level deploys elements of the higher level.

Although, as I have said, no authority was cited to me as actually debating the point there is a good indication in authority that Mr. Pritchard’s approach is not correct. In Gale’s Application [1991] RPC 305, the invention, or the alleged invention, in question was a ROM which contained an algorithm for calculating a square root. The question in that case was whether the invention related to a computer programme “as such”, but in the course of their judgments all the judges, both at first instance and in the Court of Appeal, considered that the underlying algorithm was itself a mathematical method. Thus Aldous J said, at page 316:

“The claim of this application opens with the words: ‘Electronic circuitry in the form known as ROM’. These words define a well known product. Thereafter the claim goes on to define that circuitry by the way that it will be operated which in effect is a mathematical method of obtaining the square root of a number. Despite the fact that the circuitry is so defined, it seems to me that this is a claim to a ROM having certain circuitry which could be installed in a device such as a calculator. No doubt the basis behind the claim can be said to be a mathematical method or a method for performing a mental act or even a program for a computer in that the ROM functions as a carrier or a program which will be used in a computer.”

I rely particularly on the reference to “mathematical method” in that citation. If one looks at how the algorithm is described in that case, one can see it is not a matter operating at the higher level at which it might be said that a system like calculus would operate. It is the application of conventional mathematical techniques to numbers, and therefore a matter of calculation.

In the Court of Appeal the three judges each said things to the same effect. At page 327 Nicholls LJ said:

“In the present case Mr. Gale claims to have discovered an algorithm. Clearly that, as such, is not patentable. It is an intellectual discovery which, for good measure, falls squarely within one of the items, mathematical method, listed in section 1(2). ”

Jonathan Parker LJ and Sir Nicholas Browne-Wilkinson, VC, expressed similar view. Those views do not support Mr. Pritchard’s dichotomy. The algorithm in question was not, as I have already said, in the same category as, for example, calculus. The case therefore supports the proposition that mathematical method has a wider meaning and effect than that attributed to it by Mr. Pritchard in his submissions.

Looking at the overall scheme of claim 1 it seems to me that it does indeed amount to a mathematical method. The use of the word “method” in the claim is an apt word to use and its end is to detect abnormalities. It does so by what can fairly be described as a mathematical method calculating the informational content of two sets of input data and then carrying out a statistical analysis to compare those results. I consider that to be a mathematical method both in terms of the normal use of language and in terms of the likely policy underlying the Patents Act. It follows that applying the legal approach referred to above, what is said to be the contribution to the art in this case falls under that particular head of excluded matter and the application fails accordingly.

I should, however, say something about one other area of excluded matter. Section 1(2)(c) excludes a scheme, rule or method for performing a mental act. It seems to me that if the invention was not a mathematical method, it would very arguably, if not very probably, have been a method of performing a mental act falling within that exclusion. The closeness of those concepts was reflected by Laddie J in Fujitsu Limited’s Application [1996] RPC 511, 532.

That conclusion would have been fortified by the expressed views of the appellant’s solicitors in a letter to the Patent Office dated 11^th August 2004, enclosing some amendments to the claims in which they stated:

“Although in theory the method steps could be performed mentally, in practice it would not be possible to do this within any meaningful time frame. The detection of input errors is useful only if it can be done substantially in real time. The invention can only be implemented as part of a computerised system, and the end result is a computerised system which functions more efficiently. In other words, the invention in substance is an improved computerised system, and is not a mental act or a method of doing business.”

It is a bit of a mystery to me why this was not a basis of rejecting claim 1, which says nothing about a computerised system, but the point was not taken in the decision and there was no respondent’s notice on the point. The decision seems to consider that claim 1 itself requires what the Hearing Officer describes as “a physical data channel” which I take to mean a physical data channel in something like a computer: see paragraph 20 of the decision. To my mind, it simply does not. It refers to methods and techniques, not the physical means of carrying them out.

However, Mr. Birss did not advance any arguments on any of this and Mr. Pritchard was understandably not prepared (and indeed bearing in mind the stance of the Patent Office not required) to deal with these sorts of points. It was therefore not part of the substance of this appeal. I confess to finding that somewhat unsatisfactory because the interrelationship between the concepts of a mental act and mathematical calculation might have assisted in arriving at a resolution of this matter, but I am left with the appeal as it was presented to me. Accordingly, I do not base my decision on the invention in claim 1 being a method of performing a mental act.

I turn now to claim 9. I have set out its terms above. That claim adds a computer into the process. It is the means of receiving data and carrying out the statistical analysis referred to in claim 1. In my view, the computer adds nothing material for the purposes of the excluded matter point. The process is still a mathematical method albeit carried out by a computer rather than pen and paper. It is to the substance of the matter to which I have to look: see Fujitsu Limited’s Application (in the Court of Appeal) [1997] RPC 608, and that is indeed the substance of the matter and remains a mathematical method. Accordingly, that exclusion applies to this case and notwithstanding the fact that, in my view, the Hearing Officer erred in applying the business matter exclusion, I am satisfied that the mathematical method by itself covers the entirety of the alleged contribution to the art and, accordingly, the appeal is dismissed.