Prior Art

An eagle-eyed colleague recently pointed me to the case of CLS Bank v Alice, where Alice Corporation is trying to enforce a set of patents against CLS Bank, the little-known (but systematically vital) central settlement point for interbank FX trades. 

The judgment itself is interesting because it tries to rule on the patentability of computer-implemented inventions (that is, business methods implemented “on a computer” instead of “in real life”). Alice’s patents specifically refer to computer-based implementations of certain processes, rather than the processes themselves, and the Federal Court was asked to rule whether the computer-based implementation might be patentable even when the business process itself isn’t. 

The end result was a hung bench - five people saying that computer-based implementations are patentable, and therefore Alice has a case; five people saying they’re not. The end result is that they’re apparently not, though there’s a lot of talk that Alice will appeal.

I’m going to leave the “they were wrong! they were right!“ arguments over jurisprudence to the people who know better. But it seems to me that Alice’s patents - even if you look at them purely as computer-based implementations, not as claims on the underlying processes - might have problems with prior art. That is, the patents aren’t valid because someone else already did it. 

Have a look at the patents linked off the PatentDocs article (the very first link in this post). Here’s the abstract of 5,970,479:

Methods and apparatus which deal with the management of risk relating to specified, yet unknown, future events are disclosed. Sponsor stakeholders specify a particular product relating to an event or phenomenon for which there is a range of possible future outcomes. Ordering stakeholders then offer contracts relating to the predetermined phenomenon and corresponding range of outcomes. The offered contracts specify an entitlement or (pay-off) at the future time of maturity for each outcome, and a consideration (or premium) payable, in exchange, to a counter-party stakeholder.

Independently of the offered contracts, the counter-party stakeholders input data as to their view of the likelihood of occurrence of each outcome in the predetermined range into the future, or specifically at the predetermined date of maturity. Each offered contract is priced by calculating counter-party premiums from the registered data, and a match attempted by a comparison of the offered premium with the calculated premiums. Matched contracts can be further traded until maturity, and at-maturity processing handles the exchange of entitlement as between the matched parties to the contract.

Here’s a decoder ring: “ordering stakeholder” = “market maker”; “probability” = “delta”; “premium” = “premium”. The claimant is describing an options exchange. Or, specifically, he’s describing an electronic implementation of an options exchange. (Have a read further down in the Claims section if you’re suffering from insomnia, it’s gold.)

All well and good. But the patent’s dated 1992… and CBOE implemented an electronic options limit order book in 1978.

Another patent - the one at issue in _CLS Bank_ - is 6,912,510:

A method of exchanging an obligation between parties where the exchange is administered by a supervisory institution that ensures real-time settling of obligations between parties by updating shadow records in real-time and instructing one or more exchange institutions to effect, from time to time, the exchange of obligations in accounts maintained external to the supervisory institution. Updates to the exchange institution accounts may reflect the net obligations of parties over a nominated period of time. The role of the supervisory institution is to ensure that obligations are only settled where parties have sufficient balance in their shadow records to complete the transaction.

Obligations that can be exchanged include, but are not limited to: shares in financial or physical assets, participation rights in wagers, national or synthetic currencies, exchange settlement account deposits, taxation account deposits, and deposits of financial instruments or precious metals.

The decoder ring again: “shadow balances” = “bank accounts”; “exchange institutions” = “banks” (and credit card companies, but you get the idea). The claim describes real-time gross settlement systems. (Claim 58 reads to me like they’re trying to claim nostro accounts as part of the patent, which seems remarkably ambitious.)

And again, the patent’s dated 1992. But CHIPS - a privately-run American settlement system that competes with Fedwire - has been doing electronic RTGS settlement since the 1970s, and Fedwire itself seems to have been doing the same thing long before CHIPS appeared

I’m not sure where I stand on the issue of software patents. I mean, I have no problem with patenting an innovative business process. But I think saying “we’re patenting this already-existing process… on a computer!” is a bit of a stretch. And attempting to patent something that’s already been in existence for twenty years is a _real_ stretch - and I think that’s what Alice has done, at least in these two cases. 

Have I misread these patents? Or is there a possible serious issue of prior art here?