Why IP Australia is Clearly Wrong About Research Affiliates

Research Affiliates, LLC. [2010] APO 31 (17 December 2010)
Research Affiliates, LLC [2011] APO 101 (5 December 2011)

Manner of manufacture – whether method, system and computer program product for generating a weighted securities index is patent-eligible

In this continuation of IP Australia’s ‘war on business methods’, a Delegate of the Commissioner of Patents has, for a second time, rejected claims directed to an invention directed to the construction and use of passive portfolios and indexes for securities trading.

A first set of claims was rejected in a decision issued on 17 December 2010, with that decision subsequently being appealed to the Federal Court of Australia.  However, the applicant – originally the inventor Robert D Arnott, and subsequently the assignee, Research Affiliates LLC – had filed a divisional application with slightly different claims, and the appeal was placed on-hold pending a decision on this further application.

The outcome in the second case is unsurprising.  The Delegate (who also issued the first decision) has not even bothered to provide full reasoning for his decision, instead settling for a comparison table to explain why the reasons in the first decision apply also to the revised claims of the divisional application.  Indeed, he has gone a step further, stating that the divisional claims constitute ‘an even greater contravention’ of the requirement to define a manner of manufacture than the claims of the parent application.

Research Affiliates has now appealed the second decision also.  We expect that the two appeals will be joined to a single proceeding, to be heard together.

We are hopeful that the appeals will be successful.  IP Australia’s decisions in these cases are not merely wrong, they are demonstrably inconsistent with very clear statements made by the Full Court of the Federal Court of Australia (by which a single judge of the court is bound) plainly indicating that the subject matter of the Research Affiliates claims, in both the parent and divisional applications, is patent-eligible in Australia.

BACKGROUND

Research Affiliates (‘RA’) describes itself as ‘a global leader in innovative investing and asset allocation strategies’ which is ‘dedicated to solving complex investment issues’ and to ‘creating innovative strategies that respond to the current needs of the market.’

Inventor Robert Arnott is Chairman and CEO of RA, and one of the firm’s lead researchers.  According to the RA website, one of its key competitive advantages is its ‘RAFI® (Research Affiliates Fundamental Index®) methodology’ which:

…involves selecting and weighting securities by fundamental measures of company size, as opposed to market capitalization. The methodology captures many of the benefits of passive investing— such as transparency, objectivity, broad economic representation, and diversification— with less exposure to pricing errors and fads.

The RAFI methodology is designed to work in inefficient markets. We believe that prices contain errors and that they revert back to fair value over time. Using fundamental measures of company size, such as sales and dividends, the RAFI methodology represents a company's economic footprint, not constantly shifting market expectations, bubbles and anti-bubbles reflected in its share price.

So what is so special about this?

Even the most casual observer of share markets, or the finance news, is familiar with the usual concept of an ‘index’.  Examples of indexes (or ‘indices’ depending upon linguistic preference) include the Australian ‘All Ordinaries Index’ and ASX 100, the US ‘Dow Jones’ and ‘S&P 500, the UK ‘FTSE 100’, the Japanese ‘Nikkei 225’ and the Hong Kong ‘Hang Seng’.  Each of these indexes represents a valuation of a market or sector, and is generally some form of weighted sum of a selected set of stocks.

As an individual investor, if you were to buy a portfolio of shares, consisting of all those making up, say, the ASX 100 index, and in exactly the same proportions, then the value of your portfolio would rise and fall exactly in accordance with the index.  Since, over the long term, most of these ‘standard’ indexes tend to rise, this would actually not be a bad investment strategy.  And, of course, it would not require any thought on the part of the investor.  Portfolios built in this way are therefore called ‘passive’.

All of the conventional indexes are based on weighting the investments according to one of market capitalization weighting (i.e. in proportion to total company value – most of the well-known indexes are of this type), equal weighting (self-explanatory – examples are the Value Line index and the S&P 500 Equal Weighted Stock Index) or share price weighting (i.e. in proportion to the price of individual shares – for example the Dow Jones Industrial Average).

Of course, professional investors (and their clients)  expect to be able to outperform the established market indexes.  RA’s claimed contribution is to use measures of company size other than market capitalisation, or the other common weighting methods, to determine the proportions of shares to hold in a portfolio made up of a selected set of stocks.  In other words, RA uses its own index with weightings derived from its own research into improved ways of measuring economic performance.

Assuming this works, it is clearly a very powerful and valuable idea.  To a fund manager, a method of constructing an improved index – and thus a corresponding portfolio – is of comparable significance to giving an auto mechanic an improved supercharger!

This is essentially what the RA applications claim, and what IP Australia has determined in not patentable.  However, as far as we can see this is just as wrong as if the Patent Office had rejected a claim to a new and inventive supercharger – which, of course, it would never do, because mechanical things are always manners of manufacture!

WHY IS IP AUSTRALIA WRONG?

It is very easy to demonstrate that the Delegate has got this one wrong.  It is not even necessary to analyse his convoluted reasoning (although we will come to that in due course).

Here is a representative rejected claim of the RA divisional application:

28. A computer system including:

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

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

                (b) is not based on market capitalization weighting, equal weighting,
                share price weighting and any combination thereof.

And here is a representative claim as considered by the US Court of Appeal for the Federal Circuit in State Street Bank & Trust Co. v Signature Financial Group 149 F. 3d 1368:

1. A data processing system for managing a financial services configuration of a portfolio established as a partnership, each partner being one of a plurality of funds, comprising:

1. computer processor means for processing data;
1. storage means for storing data on a storage medium;
1. first means for initializing the storage medium;
1. second means for processing data regarding assets in the portfolio and each of the funds from a previous day and data regarding increases or decreases in each of the funds’ assets and for allocating the percentage share that each fund holds in the portfolio;
1. third means for processing data regarding daily incremental income, expenses, and net realized gain or loss for the portfolio and for allocating such data among each fund;
1. fourth means for processing data regarding daily net unrealized gain or loss for the portfolio and for allocating such data among each fund; and
1. fifth means for processing data regarding aggregate year-end income, expenses, and capital gain or loss for the portfolio and each of the funds.

Now, as far as we can see these two claims are directly analogous.  Both define an invention in terms of a computer-implemented system using ‘means plus function’ type language (the means generally being some combination of hardware and software).  Both involve manipulating data representing assets within a portfolio in order to produce information for use in managing an investment.  Neither one results in any ‘physical effect’ in the outside world, although both clearly cause changes in physical state of information storage elements within the respective computer systems.

Here is what the US court had to say about the claim in State Street (at p 1373):

Today, we hold that the transformation of data, representing discrete dollar amounts, by a machine through a series of mathematical calculations into a final share price, constitutes a practical application of a mathematical algorithm, formula, or calculation, because it produces "a useful, concrete and tangible result" — a final share price momentarily fixed for recording and reporting purposes and even accepted and relied upon by regulatory authorities and in subsequent trades.

That seems pretty clear – in the US, at least, the transformation of data representing financial and investment information, using a computer, to produce a ‘useful, concrete and tangible result’, in the sense of something that people actually use as a basis for further action, is patent-eligible.

And here is what the Full Federal Court in Australia had to say about the State Street decision in Grant v Commissioner of Patents [2006] FCAFC 120, at [21-24]:

In State Street … the United States Court of Appeals for the Federal Circuit held that a business system, a data processing system utilising a mathematical algorithm for implementing an investment scheme, constituted a practical and concrete application. This involved the transformation of data by a machine through a series of mathematical calculations to provide a share price fixed for recording and reporting purposes. The Court was careful to emphasise that a mathematical algorithm, formula or calculation was not patentable but the transformation of data by a machine that produced a useful, concrete and tangible result was patentable.

The United States Court of Appeals revisited the subject of patentable subject matter in the context of the manipulation of numbers in the field of computer technology in AT&T. The basic proposition was repeated, that a mathematical formula alone, viewed in the abstract, is unpatentable subject matter. … The Court cited its earlier decision In re Alappat … in which it explained the ‘straightforward’ concept on which previous decisions, including State Street were based, ‘namely, that certain types of mathematical subject matter, standing alone, represent nothing more than abstract ideas until reduced to some type of practical application’ (emphasis in original) …. An application is not limited to a physical transformation, which was said to be an example of a useful application but not an invariable requirement. Where the claimed process did nothing more than take several abstract ideas and manipulate them together, nothing was added that was patentable.

Despite the broad terms of the United States Code (35 U.S.C. SS 101), the courts of that country have held that abstract ideas, laws of nature and natural phenomena are outside the categories of patentable invention until reduced to some type of practical application ….  The same tests apply for patents for business methods as for any other claimed invention. The distinction was between the employment of an abstract idea or law of nature and the idea or law itself.

… While the development of US patent law is derived from the Constitution of the United States rather than the Statute of Monopolies, … the policy behind the two provisions is the same. In both jurisdictions, the courts have confirmed that a broad approach to subject matter should be taken in order to adapt to new technologies and inventions but that does not mean that there are no restrictions on what is properly patentable.

And further, at [32]:

A physical effect in the sense of a concrete effect or phenomenon or manifestation or transformation is required. In NRDC, an artificial effect was physically created on the land. In Catuity and CCOM as in State Street and AT&T, there was a component that was physically affected or a change in state or information in a part of a machine. These can all be regarded as physical effects.

This can only be read as approval of the US approach.  In short, the State Street claims (in fact, the patent was owned by Signature Financial Group) would have been accepted as patentable in Australia by the Full Bench of the Federal Court.

And that being the case, it seems pretty clear that the Research Affiliates claims must also be patentable.  Certainly the Australian Delegate’s finding that (at [16] of the latest decision) ‘there is nothing of substance in the specification as a whole from which claims to patentable subject matter could be drafted’ appears manifestly untenable.

HOW DID IP AUSTRALIA REACH A WRONG CONCLUSION?

We cannot see a flaw in the above reasoning.  The inevitable conclusion is that computer implemented methods and systems, which take input information and process it to produce new information as an output, at least when that information is ‘useful, concrete and tangible’ in the sense that it is not abstract, but rather can provide the basis for specific further action, is patent-eligible.

So how did an experienced Hearing Office at IP Australia go so far astray?

We would say that this was firstly because he failed to look at what the Grant decision clearly tells us is patentable.  As explained above, the decision is pretty clear on the point that the claims in the State Street case would have been patentable had they been up for consideration in Australia.  In our view, this should have been the starting point for the analysis in Research Affiliates.

However, Mr Grant’s claims were not patentable.  They related to abstract relationships involving legal arrangements between people and entities (such as trusts).  They did not involve, or require, the use of any computer processing.  No input information was received, processed, or transformed into new and useful output information.  Perhaps because of the negative outcome in Grant, the Delegate has been unduly concerned to look in the decision for reasons to reject the Research Affiliates’ claims?

Indeed, at paragraph [22] of the first Research Affiliates decision, the Delegate appears to have been distracted by references to cases relating to printed information, which were cited in the Patent Office’s own earlier decision in Iowa Lottery [2010] APO 25, from which he cited the following passage:

Information even if represented in a physical way has never been considered sufficient for patentability save for some material advantage or mechanical effect in the arrangement of the information. See for example Re Cooper’s Application (1901) 19 RPC 53 and Re Virginia-Carolina Chemical Corporation’s Application (1958) RPC 35. (Emphasis added in Research Affiliates.)

Both of the old cited decisions indeed relate to whether particular information (‘printed matter’) was patentable.  In Virginia-Carolina Chemical Corporation the relevant principle was explained as being that the intellectual or visual content of a paper, film or other medium related to the fine and not the useful arts.  In Cooper the claims related to an invention whereby one or more spaces were left across the printed page of a newspaper so that it could be folded and remain readable.  This particular arrangement of information was held to be patentable.

Clearly, neither of these cases have anything to do with the situation in Research Affiliates.  The claimed method and system for index generation does not relate to ‘arranging’ information.  Rather, it relates to conducting specific data processing steps to produce new information.  It is the production of this new information, and the fact that it is ‘useful, concrete and tangible’ in nature, which confers patentability.  ‘Arrangements’ of information are utterly irrelevant to the invention.

The Delegate concluded (at [25] of the first decision), that:

The requisite physical phenomenon or effect is not present in the method of claim 1 nor is there one resulting from the working of the method. That is, there is no material advantage or mechanical effect evident in any arrangement of information in the present case…

It is apparent, however, that this conclusion is based on a manifestly false premise, namely that an ‘arrangement of information’ has anything relevantly to do with assessing the patentablility of the Research Affiliates’ claims.  It plainly does not.

CONCLUSION

In our view, IP Australia’s assault on ‘business methods’ is not only producing outcomes that are plainly incorrect in law, it is unduly complicating and obfuscating an area of law which was actually quite clear and well-settled.

In this case, a simple comparison with the Australian Full Federal Court discussion of State Street leads quickly and cleanly to the conclusion that Research Affiliates’ claims are patentable.

In contrast, trying to establish that they are not patentable sends the Delegate off on a round-about path that not only leads to the wrong conclusion, it incorporates irrelevant considerations and citations of case law that have nothing to do with modern data processing applications.  This contributes to confusion and uncertainty for applicants and their advisors.  When we look to the relevant authorities and conclude, on a sound basis, that particular subject matter is patentable, we are not accounting for the possibility that a hearing officer will conclude otherwise on the basis of a 1902 decision relating to the layout of newspaper pages!

If any further evidence is required that all this serves only to complicate and obscure the law, we note that the Delegate’s decision in the first Research Affiliates case is about 200 words longer than the US Federal Circuit court decision in State Street, and about 1000 words longer than this blog post.

It is now up to the Federal Court to set matters right.

Tags: Australia, Business processes, Computer programs, Patent Office, Patentable subject matter