As I mentioned in my previous post, quant and TOBAM funder Yves Choueifaty’s journey towards securing a Canadian patent for his process of constructing “anti-benchmark” securities portfolios has not been easy.
But first, a quick detour. Many companies have secured Canadian financial system patents. For example:
- TD-Bank: 36 patents in block chain, machine learning, recommendations, transaction processing;
- TSX (Toronto Stock Exchange) Inc.: 8 patents in latency tolerance and failover management;
- Accenture: hundreds of patents in systems that provide futures order throttling and order notification;
- JPMorgan Chase: 6 patents in settlement systems and stored value cards;
- NASDAQ: 14 patents in block chain systems and social media analysis;
- American Express: 14 patents in dynamic loan bill payment, e-commerce systems, and transaction processing;
- Thomson Reuters: 55 patents in electronic trading test beds, latency floor determination, price-maker speed differential management, and quantification of trading temporal fairness;
- NYSE: 8 patents in liquidity seeking, order slicing, and securities creation; and
- Chicago Mercantile Exchange: 10 patents in credit default swap clearance systems, market controls, and systems that automatically generate orders.
The Choueifaty case has piqued my interest because, unlike the financial system patenting mentioned above, the Choueifaty invention is about a financial product – a thing that can have millions of copies and be bounced around within a financial system, subjected to buy orders and sell orders and all of the liquidity mechanisms the financial system facilitates. This has not been the sort of thing we have gotten used to seeing patented in Canada. However, Choueifaty has now demonstrated that it is indeed possible, and commercially desirable, to do so.
Choueifaty first applied for his Canadian patent back in 2008. He, like many applicants, framed the steps and calculations of his portfolio construction process as a set of steps and calculations to be implemented by a computer, and to a computer configured to execute the particular steps and calculations.
Despite the mention of a computer, the Canadian patent examiner rejected his process under Section 2 of the Patent Act for being outside of the definition of invention. Machines like computers and methods that are computer-implemented are indeed fully contemplated as inventions in Canada. However, the Examiner seemed to take the position that Choueifaty had not himself actually invented anything to do with computers. So, the Examiner embarked on a problem-solution analysis to distill out the computing aspects in the analysis, leaving behind only the residue of disembodied steps and calculations as targets for the rejection.
Choueifaty appealed to the Patent Appeal Board. He filed a declaration by TOBAM’s Head of Research stating that the claimed steps and calculations, when executed on a computer, would significantly increase the computer’s functionality, in terms of computational speed gains.
Initially, the Board was not swayed, so Choueifaty appealed to the Federal Court of Canada. Fortunately, for Choueifaty, the Court admonished the Board for having used the corrosive problem/solution approach, and remanded the case back for a “fresh assessment”. Fortunately, for the rest of us, this led to the abandonment of the problem/solution approach to examination. More details on this may be found here: https://www.dickinson-wright.com/news-alerts/powell-patent-claim-construction-decision
On remand, the Patent Appeal Board accepted that the steps and calculations “permit a computer to carry out the portfolio optimization procedures with significantly less processing and greater speed”, such that the Section 2 rejection could be withdrawn.
So to recap: in addition to dog beds, engine cogs, coffee cup lids and paper, it is indeed possible to secure Canadian patents for computers and computer processes that produce financial products. To be successful, the Canadian IP Office will need some indication that the processes have something like a differential technical impact on the computer running them, as compared with some alternative. And, despite Chouiefaty’s success in filing a declaration after the patent filing, I recommend this be explained within the patent application itself.
Financial products can be interesting for all sorts of reasons. However, to be patentable in Canada, the patent office has to be convinced that the processes giving rise to the financial products are computationally interesting. This can take some extra prep work, but having exclusivity to financial products can be very lucrative.