Software patents in Australia: Court prefers Curves to Asset Index

Software patents in Australia: Court prefers Curves to Asset Index

Software patents in Australia: Court prefers Curves to Asset Index

Research Affiliates LLC v Commissioner of Patents [2013] FCA 71

Software patents have been available in Australia for over 20 years since the Federal Court considered IBM’s curve generation process patentable1. The same Court has now made it clear that schemes or business methods do not become patentable just by implementing them in a computer2. In Research Affiliates3, using a computer to generate a weighted index of assets was found to be not patentable because the steps could have readily been done manually and the computer implementation was nothing more than the use of a computer for a standard purpose4. It seems using software to generate curves is fine, but not to generate financial data.

A computer-implemented method for generating an index

The Court focused on one claim of two patent applications that had been rejected by the Patent Office, and the claim read:

A computer-implemented method for generating an index, the method including steps of:

  • accessing data relating to a plurality of assets;
  • processing the data thereby to identify a selection of the assets for inclusion in the index based on an objective measure of scale other than share price, market capitalization and any combination thereof;
  • accessing a weighting function configured to weight the selected assets;
  • applying the weighting function, thereby to assign to each of the selected assets a respective weighting, wherein the weighting:

(i) is based on an objective measure of scale other than share price, market capitalization and any combination thereof; and

(ii) is not based on market capitalization weighting, equal weighting, share price weighting and any combination thereof, thereby to generate an index5

An example of an index generated using the method, known as the Colonial Index, was given during the trial. This is created by importing into a computer program data relating to entities that offer securities listed on the Australian Securities Exchange (ASX) and processing the data to generate four measures of company size, together with a fundamental score for each security. The data is populated in an Excel file that in one example relates to 496 entities listed on the ASX. That Excel file is further manipulated, so that the top 250 securities are selected by reference to the fundamental score for each security and the weighting of each security within the top 250 is calculated. The file with the selected top 250 securities and the file weightings forms the Colonial Index which Research Affiliates licensed to owners of assets and fund managers6.

Australian patentable subject matter

Australian law simply requires that for a process to be patentable, it needs to offer some advantage that is material, in the sense that the process belongs to a useful art, as distinct from a fine art. Its value to the community must be in the field of economic endeavour7. An artificial state of affairs needs to exist, and a physical effect, in the sense of a concrete effect or phenomenon or manifestation or transformation, is required. It is sufficient if there is a component that is physically affected or a change in state or information that is part of a machine. However, if the claimed invention is a mere scheme, an abstract idea or mere information, it will not be patentable as there is no physical consequence8.

Court relies on deficiencies

Research Affiliates argued9 that the claimed method resulted in the generation of an index and was the same kind of artificially created state of affairs as IBM’s curve, the representation of Chinese characters approved in CCOM10, and the writing of data to a smart card approved in Catuity11. Research Affiliates also argued that it was clear the index produced by the method had utility in the field of economic endeavour, particularly when it had been licensed to Colonial First State for a significant fee.

The Court, however, felt that in the previous decisions there was a component physically affected or a change in state in a part of a machine which made the inventions patentable12. It considered that the index generated was nothing more than a set of data, information, or a set of numbers and was no more patentable than a bank balance13. It said the patent specification appeared to be intended to create the impression of detailed computer implementation, but said almost nothing about how that was to be done14.

The Court said the method did not involve a special effect being generated by a computer, there was no practical application in the method for improved use of computers, nor did the effect of the implementation improve the operation or effect of the use of the computer15. The Court found there was nothing in the specification or claim 1 to indicate: how the data was accessed in the first step of the claim; the nature of the processing undertaken in the second step to identify the assets; how the weighting function is accessed in the third step; how the relevant measure of scale is chosen in the fourth step; or how the weighting function is applied in the fourth step to assign a weighting to each asset16. The reliance placed on the Colonial Index was a good example of what was not in the specification17.

In rejecting Research Affiliate’s applications, the Court felt the implementation of the method was no more than a modern equivalent of writing down the index on pieces of paper and the series of steps that give rise to the generation of index could even have been carried out manually18.

Lessons learned for software patents in Australia

The Federal Court has made it clear there are limits to patentable subject matter in Australia and that merely relying on implementing in a computer a general process that can be done manually, with nothing more, will not be sufficient. The Court was particularly critical of the lack of disclosure in the specification regarding actual data processing. It was clearly looking for more than just a recital of high level process steps that can be carried out manually, but are claimed as being done by a computer. It indicated that while what constitutes a patentable invention is still evolving, and new developments in technology might widen the notion of what is patentable, the modern availability of computers as a standard means of implementing arithmetic or computational processes, which could have been implemented manually in the past, does not carry with it a broadening of the scope of patentable subject matter19.

For large data sets, it is of course impractical for the index generation method to be carried out manually. Perhaps if the specification had included some more detail on the actual processing performed, and this was claimed, then the result may have been different. Also it seems unjust to approve processes that generate curves and display Chinese characters and reject one that generates clearly useful and very valuable financial data. Focussing on what a process produces rather than the steps claimed also seems incorrect.

The claim was limited to a process performed by a machine, not a manual process. Surely the real enquiry should be whether the process is inventive or properly defined if the claim is considered too broad. Yet this would involve more evidence and analysis, and perhaps it is easier for decision makers to apply the blunt instrument of lack of patentable subject matter, regardless of the damage and financial loss this may cause.


  1. International Business Machines v Commissioner of Patents (1991) 33 FCR 218.
  2. Research Affiliates LLC v Commissioner of Patents [2013] FCA 71, 13 February 2013.
  3. Ibid
  4. Ibid, Para 72.
  5. Ibid, Para 50.
  6. Ibid, Para 51.
  7. Ibid, Para 13.
  8. Ibid, Para 22.
  9. Ibid, Para 63.
  10. CCOM Pty Limited v Jiejing Pty Limited (1994) 51 FCR 260
  11. Welcome Real-Time SA v Catuity Inc (2001) 113 FCR 110
  12. Supra no.2, Para 65.
  13. Ibid, Para 67.
  14. Ibid, Para 68.
  15. Ibid, Para 70.
  16. Ibid, Para 70.
  17. Ibid, Para 68.
  18. Ibid, Para 72.
  19. Ibid, Para 73.