PTAB reverses 103: statement that combining pieces is predictable doesn't explain why a POSITA would combine

Takeaway: The Applicant appealed an obviousness rejection of claims to a currency trading user interface, including the feature "determine a currency in which to quote skew for each of the risk reversals." The Examiner relied on a combination of two references. The Board found the Examiner's rationale for combining was deficient. "Apparently, the Examiner finds that using market trend indicators, such as skew, is predictable with currency swap market transactions, including risk reversals. This is hard to quarrel with as such, but does not find why one of ordinary skill would determine a currency in which to quote skew for each of the risk reversals." (Ex parte Breitenbach, PTAB 2015.)

Details:

Ex parte Breitenbach
Appeal No. 2013-005344; Appl. No. 12/197734; Tech. Center 3600
Decided:  Jan. 15, 2016
 
The application on appeal was directed to a user interface that displays currency trading information. The Examiner rejected as obvious using a combination of Rodgers ("Trading Options") and Thompson ("Displaying Trading Trends"). An independent claim on appeal read:

1.  A system comprising [a computing device operable to:]
[allegedly in Rodgers] retrieve market data for a plurality of risk reversals for a currency pair at a first time ...
[allegedly in Thompson] determine a currency in which to quote skew for each of the risk reversals;
[cause] an interface screen to be displayed [on another] computing devices
     a listing comprising the first and second risk reversals, the market data, and an indication of the determined skew for each of the risk reversals;
...

Thus, according to the rejection, the Examiner relied on Rodgers for the entirety of one limitation (retrieval of risk reversal data), and Thompson for the entirety of another (currency determination). The Examiner gave this explanation of the reason to combine: "obvious to use skew as a market indicator in the Rodgers system, in view of Thompson's teaching that skew is a market trend indicator that the investor can consult to make better trade decisions."

On appeal, the Applicant argued that Thompson merely mentioned skew as a market trend indicator, did not include the phrase "risk reversal" anywhere, and did not mention currency in the paragraphs relied on by the Examiner. Thus, Thompson did not disclose or suggest "determine a currency in which to quote skew for each of the risk reversals," as specifically asserted by the Examiner.

In the Answer, the Examiner first clarified his position on Thompson with respect to currency. The Examiner noted that Thompson contains many instances of both "currency" and "currency pairs," then explained that "a prior art reference must be considered in its entirety, i.e., as a whole," citing W.L. Gore & Assocs v. Garlock, 721 F.2d 1540, 220 (Fed. Cir. 1983). Moreover, Thompson's Summary explicitly describes the invention as related to currency pairs. Therefore "the details of the invention recited in Thompson relate to currency pairs." (Emphasis added.) With respect to risk reversals, the Examiner cited to several mentions in Rodgers – despite the reliance in the Final Office Action on Thompson for the entire "determine ... for each of the risk reversals" limitation.

The Board first made some Findings of Fact:

Rodgers describes displaying prices for currency pair risk reversals. Thompson is directed to providing a visual display of a two-line cross-over method signaling buying and selling opportunities of foreign currency pairs  in the foreign exchange.   

The Board then reversed the rejection under § 103, with this explanation:

Each of these paragraphs [in Thompson relied on by the Examiner] lists a variety of market trend indicators, among which is included skew [but do not] describe how a skew is used nor describes determining a currency in which to quote skew for risk reversals. Rodgers describes risk reversals, but does not discuss using skew at all. Apparently, the Examiner finds that using market trend indicators, such as skew, is predictable with currency swap market transactions, including risk reversals. This is hard to quarrel with as such, but does not find why one of ordinary skill would determine a currency in which to quote skew for each of the risk reversals.
(Emphasis added.) 

My two cents: Great save by the Board. I characterize this as a "save" because the real issue wasn't deficiency in the individual teachings of the reference, but the deficiency in the combination. The parties focused on the former, while the Board zeroed in on the latter. 

The Applicant focused exclusively on Thompson's failure to teach risk reversal in the context of a currency determination, as claimed. Understandable, because the rejection appeared to specifically rely on that feature being taught in Thompson.  There was one sentence in the Answer that hinted that the Examiner had instead combined the risk reversal in Rodgers with Thompson's teachings about currency and skew. But that hint was buried in the point-counterpoint structure of the Answer, and the Applicant didn't pick up on this in the Reply Brief. 

Reading in between the lines, I think the Examiner really did rely on the combination to arrive at "determine a currency in which to quote skew for each of the risk reversals." Seems like the display of risk reversal for currency pairs from Rodgers' FIG. 1 was added to Thompson's mentions of and skew as a market indicator – to arrive at "determine a currency in which to quote skew for each of the risk reversals."

But suppose we generously give the Examiner credit for what he might have been thinking (as opposed to what he actually put in the record). As the Board found, even that wasn't enough to make the combination obvious. 

The Board's analysis focused on the lack of connection between the teachings in the two references. Each reference had various pieces of the claim, but these teachings weren't related enough for a POSITA to combine them as claimed. 

Many Applicants dealing with disparate teachings from multiple references either argue non-analogous art or just simply quote the teachings and say "Hey, look, these are different things." I found the Board's explanation for non-obviousness to be persuasive, because instead of simply listing differences, they put these differences in context: Yes, reference X taught A and B, but not in the context of C. Yes, reference Y taught C, but not in the context of A and B. Simply knowing about the pieces isn't enough of a reason to combine because the contexts are different and/or have little overlap.

The Board used the magic word "predictable" in its analysis, which to me is a clear reference to KSR's "predictable results" rationale. What are we to make of the Board's reversal of obviousness despite the apparent concession that "using market trend indicators, such as skew, is predictable with currency swap market transactions, including risk reversals."

Is the Board saying that generally combining the pieces at a high level is predictable, but the claimed combination is more specific than that and so a high level predictable result isn't enough?

Thanks to Robert K S for calling my attention to the Breitenbach decision.