[00:00:00] Speaker 03:
Cases 2,022-1286.

[00:00:05] Speaker 03:
Ms.

[00:00:05] Speaker 03:
Maynard.

[00:00:06] Speaker 00:
Thank you, Your Honor, and may it please the Court.

[00:00:08] Speaker 00:
Dan Maynard for Teradata.

[00:00:10] Speaker 00:
I'd like to reserve three minutes for rebuttal.

[00:00:13] Speaker 00:
In granting summary judgment on Teradata's antitrust and trade secrets claims, the District Court erred as a matter of law and decided disputed factual questions.

[00:00:24] Speaker 00:
I'd like to start with the antitrust claim.

[00:00:26] Speaker 02:
carried it out later today that that that's right okay can we start with jurisdiction which uh... right here uh... trivial issue here so uh... let me just tell you that the things that uh... are uh... that i'm thinking about and concerned about uh... which have

[00:00:45] Speaker 02:
maybe can be summarized with reference to the discussion in the reply filing at the motion stage of Chamberlain.

[00:01:00] Speaker 02:
The argument I think by SAP is that what it said about its patent counterclaims

[00:01:13] Speaker 02:
was said in response to TerraData having asserted up to that stage, I think the number was 482 trade secrets.

[00:01:23] Speaker 02:
Anyway, a very, very large number.

[00:01:27] Speaker 02:
What it did not say was that its patent counterclaims were directly responsive to the single technical trade secret claim left

[00:01:42] Speaker 02:
in your case, by the time you were responding to the summary judgment motion.

[00:01:48] Speaker 02:
And that without that, the compulsory counterclaims standard is not met, and that under Chamberlain and the case relied on in Chamberlain, the dropping of all the technical trade secret assertions, but for the one remaining one, leaves the parties

[00:02:12] Speaker 02:
with respect to all the dropped claims in the same position as if they had never been made and therefore they don't count for purposes of the jurisdictional determination.

[00:02:25] Speaker 02:
Now there's a lot packed into that, but that's what I'm worried about.

[00:02:29] Speaker 00:
Yes, Your Honor.

[00:02:29] Speaker 00:
So I think just out of the gate, Chamberlain is distinguishable because they're

[00:02:33] Speaker 00:
The dropped claims were the very patent claims that were the hook for jurisdiction.

[00:02:38] Speaker 00:
And once they were dismissed... Right, right.

[00:02:40] Speaker 02:
You know, absolutely, one has to translate Chamberlain into the counterclaim context.

[00:02:44] Speaker 02:
Absolutely.

[00:02:44] Speaker 02:
But obviously, one can do that.

[00:02:49] Speaker 00:
On the facts here, Your Honor, the timing that they assert in their reply brief is mistaken.

[00:02:58] Speaker 00:
Teradata served several amended narrow trade secrets claims.

[00:03:03] Speaker 00:
By December 2020, we had narrowed the trade secret case to 19 trade secrets, more than half of which related to batch merge, which is, as you know, the trade secret here today.

[00:03:13] Speaker 00:
And in SAP severance opposition, which they filed in March of 2021, they stated that what SAP said in 2019 holds true today.

[00:03:26] Speaker 00:
quote, Teradata's trade secret claims are intertwined with SAP's counterclaims.

[00:03:31] Speaker 00:
That's an APPX 8881.

[00:03:35] Speaker 00:
And then I would point, Your Honor, to a chart that SAP presented to the district court.

[00:03:40] Speaker 00:
It's not in the joint appendix here, but you can find it at ECF 175-4.

[00:03:46] Speaker 00:
And we cite it in our opposition to the motion to transfer at page 12.

[00:03:52] Speaker 02:
I'm excited, but I'll show it, right.

[00:03:55] Speaker 02:
Dealing with this fundamental jurisdictional issue has been a challenge because almost nothing was actually supplied to the court.

[00:04:03] Speaker 00:
So if you do have a chance to review the SAP chart that they filed with the district court to convince the district court that their counterclaims were arising out of the same transaction or occurrence and should not be severed, that's ECF 175-4.

[00:04:19] Speaker 00:
On page 6 of that chart, they deal with alleged trade secrets 24 to 30.

[00:04:28] Speaker 00:
And on page 12 of that chart, they deal with trade secrets, I mean, sorry, page 15 of that chart, they deal with trade secrets 31.

[00:04:37] Speaker 00:
Those are 8 of the 10 trade secrets now asserted that relate to batch merge.

[00:04:42] Speaker 00:
the two that they don't discuss had not been asserted at that point.

[00:04:46] Speaker 00:
So I do think that all that SAP said about how their claims interrelate, the same technologies are at issue, the same witnesses are at issue, the same themes are at issue, apply equally to the case as narrowed.

[00:05:00] Speaker 02:
So and just to be clear, so my understanding is the district court in its severance order didn't actually rule on

[00:05:09] Speaker 02:
on the same transaction, it said something like, it appears today, yes, it appears from the argument presented to this juncture that SAP's counterclaims arise out of the same transaction or occurrence.

[00:05:28] Speaker 02:
That didn't quite settle the issue, but this is also something, because it's jurisdictional,

[00:05:35] Speaker 02:
would have to decide for ourselves.

[00:05:37] Speaker 02:
What do you think is, in fact, the relation between any of the particular patent assertions and the trade secrets at issue here?

[00:05:49] Speaker 02:
I understand that they would clearly be enough if the patents either claimed or described from a filing time before

[00:06:04] Speaker 02:
the asserted trade secrets were turned over or were shared, I guess, that actually taught those things and then they wouldn't be trade secrets because they'd be in the public domain.

[00:06:17] Speaker 02:
I don't understand that to be the assertion, so I'm left unsure what it means as a concrete matter to say they were related.

[00:06:27] Speaker 00:
Well, I think the standard under this court's decision in Reardon is quite broad, as I'm sure your honor knows.

[00:06:33] Speaker 00:
And the question is whether there's a logical relationship between the counterclaims and the claims at issue.

[00:06:40] Speaker 00:
And here, that's clearly the case.

[00:06:42] Speaker 00:
SAP alleges it invented what we claim as our trade secrets.

[00:06:47] Speaker 00:
And so they claim these patents show, and I point you to this chart that they created, which goes into much more detail with a chart about the patent claims and a chart about our trade secrets that makes that tie.

[00:06:59] Speaker 00:
But at the level that matters for the broad standard that is the compulsory counterclaim, the federal rules encourage related claims to be brought together.

[00:07:10] Speaker 00:
It is a broad standard for compulsory counterclaims.

[00:07:13] Speaker 00:
If we were to win the trade secrets claim against them,

[00:07:16] Speaker 00:
They could not bring a separate suit and assert that these patent claims show that, in fact, they invented that trade secret first.

[00:07:23] Speaker 00:
That's their claim here, and they need to bring that in this action as they have done.

[00:07:28] Speaker 02:
And you say that's their claim here, I think, on the basis of some ECF 175 document that hasn't been supplied to us, right?

[00:07:38] Speaker 02:
Because I don't think anything else you cited says that.

[00:07:42] Speaker 00:
We cited on page 12 of our motion to transfer opposition, Your Honor.

[00:07:47] Speaker 00:
We briefed the jurisdictional question in our opening brief, and they did not respond.

[00:07:51] Speaker 00:
I took them to have somewhat abandoned this argument.

[00:07:54] Speaker 02:
Well, they did say, see our motion papers in their jurisdictional section.

[00:07:59] Speaker 02:
And you didn't actually respond to their reply about Chamberlain and so on in the motion paper.

[00:08:07] Speaker 02:
Can you supply this crucial ECF 175 document to the court, please?

[00:08:12] Speaker 00:
Yes, of course, Your Honor, we will do so.

[00:08:17] Speaker 00:
Thank you.

[00:08:17] Speaker 00:
If I may turn to the merits of our inter-trust claim, I would like to start there.

[00:08:21] Speaker 00:
The district court aired, so SAP is engaging in a classic illegal tie.

[00:08:34] Speaker 00:
Our evidence shows that they are forcing their locked-in ERP customer base to buy a product that they don't want

[00:08:41] Speaker 00:
and charging them premium prices for a substandard product, and that that tie is causing the very harm of such arrangements.

[00:08:51] Speaker 00:
It's foreclosing competition from others, like Teradata, that offer competing products.

[00:08:56] Speaker 00:
It's foreclosing competition on the merits.

[00:08:58] Speaker 00:
And under any standard, per se, or rule of reason, Teradata has sufficient evidence to go to a trial.

[00:09:05] Speaker 00:
But the Supreme Court has made clear that it is the per se rule that applies here, and under the per se test,

[00:09:11] Speaker 00:
There's only one element in dispute, which is, at summary judgment, which is SAP's market power.

[00:09:16] Speaker 00:
And SAP does not dispute that if Professor Asker's definition of the tying market comes in, then SAP, then Teradata has sufficient evidence to create a dispute, a fact for trial on its per se claim.

[00:09:29] Speaker 00:
And here, that testimony should have come in.

[00:09:32] Speaker 00:
Professor Asker's a well, well,

[00:09:36] Speaker 00:
recognized economists who applied well-established methods, as the amicus briefs show, to determine that SAP has market power over a certain subset of customers, namely large enterprises who are purchasing core ERP services.

[00:09:54] Speaker 00:
And the District Court cared... Ms.

[00:09:57] Speaker 02:
Naderj, so there are a number of different elements of

[00:10:04] Speaker 02:
a tying claim, whether it's rule of reason or per se, that are in play here.

[00:10:11] Speaker 02:
Some of them, I think, discussed more fully than others in the district court.

[00:10:17] Speaker 02:
I want to put aside for the moment the question of whether the definition of the tying product market was proper, even the question of whether the definition of the tied product market was proper.

[00:10:32] Speaker 02:
and focus on the question of whether there was sufficient evidence even to go to a jury of effects in the Tide product market of the only specific thing Tide, which was the runtime license for HANA itself, not an EDW product.

[00:11:01] Speaker 02:
And this is, I guess, summarized or illustrated, it's probably the right term, in the figure shown on page 48 of the red brief that picks apart the several different elements of the analysis, the different kinds of licenses to HANA, the different kinds of things that

[00:11:29] Speaker 02:
Hannah does and that compliments of Hannah like the BW product would do and illustrates the contention that there's no reasonable connection between the limited Tide product and effects on the Tide product market.

[00:11:51] Speaker 00:
So there's a lot to unpack there, Your Honor, so if you could just bear with me while I try to march through it.

[00:11:56] Speaker 00:
So first,

[00:11:58] Speaker 00:
The harm to competition, if the per se rule applies, the harm to competition that's required to be shown is only de minimis.

[00:12:04] Speaker 00:
The Ninth Circuit has said $100,000.

[00:12:06] Speaker 00:
They've never disputed that in the per se context.

[00:12:09] Speaker 00:
So the only element at issue for summary judgment was SAP's market in the time market.

[00:12:15] Speaker 00:
And we do believe the per se rule applies.

[00:12:17] Speaker 00:
But I'll take your question as assuming that perhaps the rule of reason applies, and therefore we have to show substantial harm to competition.

[00:12:24] Speaker 00:
And I believe we did so.

[00:12:26] Speaker 00:
The chart on page 48.

[00:12:28] Speaker 00:
There's multiple problems with that chart.

[00:12:30] Speaker 00:
But at most, it shows a triable dispute of fact on substantial harm to competition.

[00:12:35] Speaker 00:
So first, just to note, that chart's not drawn to any sort of scale.

[00:12:40] Speaker 00:
It also talks about the percentage of customers.

[00:12:43] Speaker 00:
And it claims that only 12% of customers use the full use license.

[00:12:52] Speaker 00:
Our expert put that at 20%.

[00:12:54] Speaker 00:
So there's a dispute of fact on that point.

[00:12:56] Speaker 00:
Um, but there's a dispute of fact about whether or not, um, the, the, so, and also, I would take issue, Your Honor.

[00:13:05] Speaker 00:
The HANA is tied.

[00:13:06] Speaker 00:
All of HANA's analytical capabilities are the tie.

[00:13:10] Speaker 00:
The HANA is tied regardless of whether it's bought under a runtime or a full use.

[00:13:15] Speaker 00:
Yes, they give an option, but you have to buy HANA.

[00:13:18] Speaker 00:
You have to buy it in one of its forms.

[00:13:20] Speaker 00:
And the first time they've even suggested that the full use is not where the tie was in this court, that argument is forfeited.

[00:13:30] Speaker 00:
But in any event, I think you believe the runtime license is tied, correct?

[00:13:38] Speaker 02:
Well, as I understood it, the only thing that anybody is actually, and I didn't understand this to be in dispute, the only thing that SAP enterprise resource planning customers need to buy contractually is a runtime license.

[00:14:00] Speaker 02:
That is, it's completely optional whether any such customers pay the extra for the full

[00:14:07] Speaker 02:
the full-use license that would then enable the use of the HANA data with other products, basically.

[00:14:18] Speaker 00:
Well, the first time they ever suggested that, Your Honor, was in this court, and they're not even clearly making it here.

[00:14:23] Speaker 00:
It's just in the parentheticals that go with the chart.

[00:14:26] Speaker 00:
I'm sorry.

[00:14:27] Speaker 00:
Was it ever in dispute that ERP customers did not have to buy a full-use license?

[00:14:34] Speaker 00:
They have to buy one or the other, and they're given an option.

[00:14:37] Speaker 00:
But as our expert's testimony shows, the harm to competition flows from the tie of HANA, which requires them to buy one or the other.

[00:14:47] Speaker 00:
Because it's the purchase of HANA's analytics that distorts the competition and causes foreclosure in the marketplace.

[00:14:56] Speaker 00:
And he estimated that of the sales of HANA, and he did a differences and differences regression,

[00:15:05] Speaker 00:
that proved that the sales of HANA that were tied to the S4 HANA were causing harm to Teradata.

[00:15:21] Speaker 00:
And he also estimated that there was

[00:15:27] Speaker 00:
close to half, half to three quarters of foreclosure in the EDW market as a result of the forced purchase of HANA, a product that had very, very poor uptake in SAP's own experts' words before the tie.

[00:15:45] Speaker 00:
And now the vast majority of SAP's revenues from HANA come from the tied sales.

[00:15:52] Speaker 03:
Ms.

[00:15:53] Speaker 03:
Mander, your time is up.

[00:15:56] Speaker 03:
We'll give you your rebuttal time back.

[00:16:00] Speaker 03:
Judge Tarano, if you wish to ask any further questions at this time.

[00:16:06] Speaker 02:
OK, can I just ask?

[00:16:07] Speaker 02:
I guess I'd like Ms.

[00:16:09] Speaker 02:
Mander to have a short opportunity to highlight her best point on her trade secret appeal.

[00:16:18] Speaker 00:
Thank you, Your Honor.

[00:16:19] Speaker 00:
I appreciate that.

[00:16:20] Speaker 00:
So the district court aired.

[00:16:21] Speaker 00:
in concluding that we had not marked and also that somehow these agreements had given SAP the keys to one of our most important trade secrets, which is batch merge, which is something that we developed and have been using for years in conjunction with TerraDasis flagship database product, which is essential to its massively parallel processing.

[00:16:46] Speaker 00:
There's nothing in the

[00:16:50] Speaker 00:
MNDA, which requires all the details of the trade secret to be marked in order to protect it.

[00:16:55] Speaker 00:
The point is to put it on notice.

[00:16:57] Speaker 00:
And we had at least a dispute of fact as to whether we had provided sufficient notice that it was marked, because the design document where we first disclosed key aspects of batch merge was marked.

[00:17:09] Speaker 00:
And then secondly, batch merge is partner materials, which the SDCA clearly reserves

[00:17:19] Speaker 00:
to Teradata, and 9.2 of the SDCA gave them only a limited license to use it and only during the bridge project.

[00:17:34] Speaker 00:
At a minimum, there is a discrete effect on that.

[00:17:36] Speaker 00:
They recognize that we have an email from an engineer who recognizes it was confidential SAP engineer and not allowed to share it.

[00:17:44] Speaker 00:
If you have any questions, otherwise, I would like to have the time

[00:17:47] Speaker 03:
Thank you, Respected.

[00:17:48] Speaker 03:
We will restore your time.

[00:17:50] Speaker 03:
Let's hear from the government.

[00:17:52] Speaker 03:
Mr. Kuhlman.

[00:17:53] Speaker 04:
May it please the Court.

[00:17:55] Speaker 04:
Patrick Kuhlman on behalf of the United States.

[00:17:57] Speaker 04:
We're here today to address the correct legal principles for defining relevant markets in antitrust cases.

[00:18:03] Speaker 04:
The District Court departed from the flexible inquiry into market realities that should guide market definitions.

[00:18:10] Speaker 04:
because the types and the quality of the available evidence vary from case to case.

[00:18:15] Speaker 04:
Courts define more.

[00:18:16] Speaker 02:
Can I just double check?

[00:18:18] Speaker 02:
You're not asserting a position about per se versus rule of reason as the applicable standard for assessing the tie that's at issue here.

[00:18:31] Speaker 02:
Is that right?

[00:18:32] Speaker 04:
that is correct judge to the judge rondo we're just addressing uh... a number of errors in the uh... district courts uh... market definition analysis okay thank you and i want to focus on uh...

[00:18:46] Speaker 04:
two errors this morning where the district court improperly limited the use of important market definition tools.

[00:18:54] Speaker 04:
The district court's treatment of brown shoes practical indicia and the district court's treatment of aggregate diversion ratio analysis or ADR analysis.

[00:19:03] Speaker 04:
So first, the district court wrongly confined the practical indicia to submarket cases.

[00:19:09] Speaker 04:
There's no rigid dichotomy between submarkets and primary markets.

[00:19:14] Speaker 04:
A submarket is simply a small irrelevant market that resides within a broader market.

[00:19:19] Speaker 04:
Same tools apply, same market definition tools apply, regardless of how the candidate market is denominated.

[00:19:28] Speaker 04:
The district court was also wrong to require three or more practical indicia to define a relevant market.

[00:19:36] Speaker 04:
Ninth Circuit explained that the indicia are practical aids and shouldn't be used in a talismanic fashion.

[00:19:43] Speaker 04:
They're to be used pragmatically.

[00:19:46] Speaker 04:
And indeed, in ITT, the Ninth Circuit affirmed the definition of one of the relevant markets, market for telephone equipment, using less than three indicia.

[00:19:58] Speaker 04:
So second, the district court wrongly questioned the validity of aggregate diversion ratio analysis, or ADR analysis.

[00:20:06] Speaker 04:
This methodology is used by courts, endorsed in antitrust fetuses, and in the academic... Can I just double-check?

[00:20:14] Speaker 02:
We're seeing any citation to a court of appeals decision using aggregate diversion ratio.

[00:20:21] Speaker 02:
I'm vaguely recalling two district court opinions.

[00:20:24] Speaker 02:
Is that what we're talking about here?

[00:20:26] Speaker 04:
That is correct.

[00:20:27] Speaker 04:
That is correct.

[00:20:29] Speaker 04:
OK.

[00:20:29] Speaker 04:
Yes.

[00:20:29] Speaker 04:
There's two.

[00:20:30] Speaker 04:
There's actually, I can believe, there's Wilhelmsen, Cisco, Aetna, H&R Block, Bizarre Voice, and Dial Corporation cases.

[00:20:44] Speaker 04:
you know, the district court statement that ADR analysis has been rarely accepted by courts was wrong.

[00:20:51] Speaker 04:
And it is accurate, as Your Honor points out, that there aren't that many cases addressing ADR analysis.

[00:20:57] Speaker 04:
But in those cases, it uniformly accepts the general validity of the methodology.

[00:21:02] Speaker 04:
It's a standard methodology in the words of the H&R block court.

[00:21:08] Speaker 04:
I'd also like to turn to, the district court was also wrong in

[00:21:15] Speaker 04:
thinking that actual pricing data was required to implement an ADR analysis.

[00:21:22] Speaker 04:
That type of data often is not available, and it's not imperative.

[00:21:26] Speaker 04:
So aggregate diversion ratio requires an estimation of where customers are.

[00:21:31] Speaker 02:
Can I just ask, would it be a reasonable decision on a district court's part, either exercising downward authority or JMO or summary judgment?

[00:21:43] Speaker 02:
to say that pricing data was either available or could have been but wasn't used and when it's available and not used, that's a reason to reject a conclusion based on a subset of the relevant but available data.

[00:22:06] Speaker 02:
I don't take your position to reject that.

[00:22:10] Speaker 04:
No, no, no.

[00:22:12] Speaker 04:
We're not saying that the district court in exercising its discretion couldn't consider all of the available evidence and determine the weight of that evidence.

[00:22:26] Speaker 04:
Our position is rather that the district court endorsed a categorical, hard and fast rule that actual pricing data needs to be used.

[00:22:36] Speaker 04:
that actual pricing data is one way to estimate diversion, but it's not the only way.

[00:22:41] Speaker 04:
Other types of data and evidence can do so.

[00:22:44] Speaker 04:
For example, let's say general motor salespeople are hearing all their customers saying, we're considering GM trucks and we're considering Ford trucks.

[00:22:54] Speaker 04:
So that suggests that for GM customers, Ford trucks are their next choice.

[00:23:00] Speaker 04:
And that's strong evidence of where GM customers would go if Ford increased the price of its trucks.

[00:23:09] Speaker 04:
And indeed, in Cisco, Wilhelms, and an H&R block, courts define markets using ADR analyses that did not include actual pricing data.

[00:23:22] Speaker 04:
The court doesn't have any further questions.

[00:23:24] Speaker 04:
We ask that the court correct the errors in the district court's market analysis.

[00:23:28] Speaker 04:
And we thank the court for the opportunity to present our views.

[00:23:32] Speaker 03:
Thank you, Mr. Coleman.

[00:23:35] Speaker 03:
Mr. Shanmugam.

[00:23:37] Speaker 05:
Thank you, Judge Lurie, Candidate Shanmugam of Paul Weiss for the SAP Appellate.

[00:23:42] Speaker 05:
I'd like briefly to start with the issue of jurisdiction, which Judge Taranto raised, and then I will go to the merits of the district court summary judgment decision.

[00:23:50] Speaker 05:
On the issue of jurisdiction, while we recognize that it is a close question, we believe that the better view is that the Ninth Circuit has jurisdiction here.

[00:23:59] Speaker 05:
Now, I understood my friend Ms.

[00:24:01] Speaker 05:
Maynard to be quibbling with the timing of when Teradata narrowed its trade secret claims.

[00:24:08] Speaker 05:
But I didn't understand her to dispute the proposition that by the time of the district court's ruling on the issue of summary judgment, those claims really had been narrowed to what is before the court now, at least with regard to the technical trade secret claim, and that is the claimed trade secrets concerning the batch merge method.

[00:24:27] Speaker 05:
That narrowing in our view is critical for the simple reason that the patents that were the subject of our counterclaims were various patents that concerned the HANA database itself and not the subject of the trade secrets claim which is the method for communication with that database.

[00:24:46] Speaker 05:
And so, whatever the timing, our fundamental submission is that this is effectively a constructive amendment of the complaint.

[00:24:53] Speaker 05:
Essentially, what has taken place here is that Teradata has abandoned the literally hundreds of trade secrets that it initially asserted, which is what precipitated the filing of the patent counterclaims.

[00:25:07] Speaker 02:
And I would... Just to be clear, Mr. Chen, if I remember right, the Chamberlain analysis

[00:25:13] Speaker 02:
deems a constructive amendment to be one in which the parties are left in the position that they would have been in had the matter never been litigated.

[00:25:22] Speaker 02:
Okay, so I just want to make sure, SAP agrees that all of the trade secret matters asserted but later abandoned remain available to TerraData to assert.

[00:25:40] Speaker 02:
Because if not, then it falls into the actual chamber of holding

[00:25:43] Speaker 02:
that there was essentially an amendment like a dismissal with prejudice, in which case the rationale

[00:25:53] Speaker 02:
for disregarding the initially filed but dropped claims would no longer apply.

[00:26:00] Speaker 05:
I think what I would say about that, Judge Toronto, is that it does appear that Teradata has abandoned those remaining trade secret claims.

[00:26:10] Speaker 05:
And again, in some sense, we're sort of piecing this together from, for instance, their opposition to the summary judgment motion.

[00:26:18] Speaker 05:
But it does appear at this point as if the only trade secrets they are asserting with regard to the technical claim are the batch merge method claims.

[00:26:29] Speaker 05:
And I would note that to the extent that the parties took.

[00:26:33] Speaker 02:
And what does your ECF 175 document, which I'm afraid, since I haven't seen, I'm just remembering what was asserted about it, to me, the heart of Ms.

[00:26:46] Speaker 02:
Maynard

[00:26:47] Speaker 02:
point on this is that you actually did make assertions about the close relationship between these asserted patents and that the still live technical trade secret assertions regarding batch merge methods.

[00:27:05] Speaker 05:
So I don't I don't have a document in front of me in part because it was obviously not a focus of the opposition to the motion to transfer.

[00:27:14] Speaker 05:
But I think that what I would say is that to the extent that we made assertions below more generally first that there is no estoppel obviously in the context of jurisdiction and I would rest on our submissions for that proposition.

[00:27:30] Speaker 05:
But beyond that, even if there were any sort of estoppel, the district court did not issue a definitive ruling on the issue of severance.

[00:27:37] Speaker 05:
And I would point you to the document that I think you were referring to earlier, docket 418, where the district court essentially said, I'm not going to sever at this time.

[00:27:48] Speaker 05:
That issue remained a live issue below.

[00:27:52] Speaker 05:
And on the substance of what this court has to decide, the application of the Reardon standard, again, our fundamental submission here is, again, whatever was said below, that if all that is left here is the claim for the method of communicating with the database, that was quite simply not the subject of the asserted patent counterclaims.

[00:28:14] Speaker 05:
Those were, again, claims variously to the database itself.

[00:28:19] Speaker 05:
Now, I do want to, and of course, we look forward to whatever Teradata may wish to submit with regard to DOC at 175, and we're happy to respond to that.

[00:28:29] Speaker 05:
I do want to turn to the merits, given the limited amount of time.

[00:28:32] Speaker 05:
And if this court concludes that it has jurisdiction, our fundamental submission is obviously that summary judgment was correct, both with regard to the antitrust and the trade secret claims.

[00:28:42] Speaker 05:
Now, at the outset, just to provide a roadmap to the court, I think there are two ways in which the court could affirm the ruling on the antitrust claim.

[00:28:51] Speaker 05:
hold that the district court correctly exercised its discretion in excluding Dr. John Asker's qualitative and quantitative analysis supporting his market definition for the tying product.

[00:29:04] Speaker 05:
Or it could conclude in the alternative, and the district court made clear that this was an alternative basis for its summary judgment ruling, that Teradata simply failed to bear its burden of coming forward with evidence of anti-competitive effects in the EDW market.

[00:29:19] Speaker 05:
And under the rule of reason, which in our view applies, Teradata would obviously have to meet that element in order to move forward on its antitrust claim.

[00:29:28] Speaker 03:
Why does the rule of reason apply?

[00:29:31] Speaker 05:
So in our view, the reason that the rule of reason applies is that it is clear now, and the Supreme Court has indicated as much in Illinois Tool Works, a case with which I know that this court is familiar, that not all time claims are subject to the per se rule.

[00:29:46] Speaker 05:
Instead, the question is whether

[00:29:48] Speaker 05:
you are in a context where there are plausible, pro-competitive justifications, such that judicial experience provides a basis for believing that there might be a pro-competitive justification, and therefore that the rule of reason applies.

[00:30:03] Speaker 05:
And we rely, as the court is aware, on the DC Circuit's decision in Microsoft, but perhaps more relevantly, the Ninth Circuit's more recent decision in Epic Games.

[00:30:11] Speaker 05:
Now to be sure, in both of those cases, the courts made clear that their holding was limited to the specific context in front of it, I think with appropriate judicial modesty.

[00:30:22] Speaker 05:
When you look at the reasoning of those decisions, they made clear that in a technology market like this one where the time product and the tide product are technologically integrated,

[00:30:32] Speaker 05:
And whereas it's true here that the integration provides benefits in the form of improved efficiency, that application of the rule of reason is appropriate.

[00:30:41] Speaker 05:
And I would note that for purposes of this question, the question of which standard applies, I think the question is really just whether there are plausible pro-competitive benefits.

[00:30:50] Speaker 05:
Obviously, the question of whether those pro-competitive benefits justify the allegedly anti-competitive practice is a question for the merits.

[00:30:59] Speaker 05:
And really, at bottom, we think that there is simply no valid reason for limiting the application of the rule of reason.

[00:31:06] Speaker 05:
to cases where, as was true in Microsoft and Epic Games, the tying product is the platform or where the benefits in your deferred parties.

[00:31:15] Speaker 05:
The court will be aware of the reasoning of the Ninth Circuit in the recent Epic Games decision, where the court warned against per se treatment in software markets like this precisely because those products are highly innovative.

[00:31:27] Speaker 05:
They involve short product lifetimes with a constant process, as the Ninth Circuit put it, of bundling, unbundling, and rebundling of various functions.

[00:31:35] Speaker 02:
Now, if the course... Just to be clear, when you say integrated, you don't either mean or need to say that the integration is what I would call absolute.

[00:31:47] Speaker 02:
That is, the idea that HANA cannot be made to work with any other ERP, and yes.

[00:31:56] Speaker 05:
That is... Judge Tronto, that is correct, and obviously the best evidence of that in some sense is that SAP did sell HANA

[00:32:04] Speaker 05:
separately with modest success in the marketplace.

[00:32:08] Speaker 02:
My point is to say that... At somebody, at least, I think, in the record, did get it to work with something else, right?

[00:32:15] Speaker 05:
Yeah.

[00:32:16] Speaker 05:
Well, I'll come to that when we get to the issue of anti-competitive effects.

[00:32:20] Speaker 05:
I think my point for purposes of the application of the rule of reason is simply that we did present unrebutted evidence that the integration of S4HANA with HANA came with benefits in the form of improved performance.

[00:32:33] Speaker 05:
And that that, in our view, is what triggers the application of the rule of reason, the fact that there was that plausible pro-competitive justification.

[00:32:42] Speaker 05:
And there may be contexts in which integration would not have that benefit, but here, again, there was evidence to that effect.

[00:32:48] Speaker 02:
And that brings us back to- So on that assumption, can you now talk about why some total of Teradata's evidence from Dr. Asker and others

[00:33:04] Speaker 02:
isn't enough to create a triable issue on anti-competitive effects.

[00:33:09] Speaker 05:
Sure.

[00:33:10] Speaker 02:
I'd be glad to.

[00:33:10] Speaker 05:
And then with the court's leave, I'll spend literally a minute on the trade secret claims and on the government's submission.

[00:33:17] Speaker 05:
On the issue of anti-competitive effects, as a threshold matter, Teradata failed to identify a single customer that used HANA when purchased together with S4HANA

[00:33:31] Speaker 05:
the allegedly product in the tying market as an EDW product, much less to replace Teradata's EDW product or the EDW product of any competitor.

[00:33:45] Speaker 05:
And to the extent that Teradata relies on non-tied sales of HANA, we would submit that that is irrelevant to the analysis, that the fundamental question here is with regard to the alleged tie

[00:33:58] Speaker 05:
whether that tie foreclosed market competition.

[00:34:02] Speaker 05:
Now, as you pointed out, Judge Toronto, the vast majority of S4 users were subject to these runtime licenses, which restricted the use of HANA as an EDW product.

[00:34:14] Speaker 05:
My friend, Ms.

[00:34:15] Speaker 05:
Maynard, says, well, maybe there were 20% of customers who had full use licenses.

[00:34:20] Speaker 05:
We believe the correct number is 12%.

[00:34:22] Speaker 05:
But either way, the vast majority, again, were subject to these restrictions.

[00:34:27] Speaker 05:
Now, Teradata argues that those customers might nevertheless have been able to use HANA as an EDW product by virtue of another product, BDW for HANA, which SAP also made available.

[00:34:40] Speaker 05:
But again, there was no evidence that that actually took place.

[00:34:43] Speaker 05:
And with regard to Dr. Asker, whose testimony on this issue was rightly excluded, Dr. Asker

[00:34:51] Speaker 05:
simply assumed that every customer who uses S4 HANA with HANA would use HANA as an EDW product and would cease using Teradata's and other customers' EDW products as a result.

[00:35:07] Speaker 05:
And that is how Teradata argues that there would be market foreclosure of 30% or more across competitors in the tied market.

[00:35:15] Speaker 05:
What Asker basically did was just bluntly say, look, it appears that

[00:35:20] Speaker 05:
Somewhere in the neighborhood of 48 to 73 percent of Teradata's customers use these SAP products and we assume that if they purchase the Tide product that they would all switch.

[00:35:33] Speaker 05:
That simply is an unsupported assumption here and to the extent that Asper conducted a regression analysis that showed some decline in customers spending on Teradata products.

[00:35:45] Speaker 05:
That obviously says nothing about

[00:35:47] Speaker 05:
the relevant causal question, which is whether customers were substituting HANA for Teradata's EDW products.

[00:35:54] Speaker 05:
So again, our fundamental submission on the antitrust claim is that that's one route for this court to say

[00:36:00] Speaker 05:
that the summary judgment on the antitrust claim was appropriate.

[00:36:04] Speaker 05:
If you look at footnote 16 of the district court's opinion, it made clear that it would have ruled on that ground.

[00:36:10] Speaker 05:
Of course, the alternative ground is to uphold, under the abuse of discretion standard, the exclusion of ASCR's market definition.

[00:36:18] Speaker 05:
And maybe I'll turn to the government's submission on that, since it was really the government that made some suggestion that the district court committed legal error in that regard.

[00:36:27] Speaker 02:
Can I just ask, is there any, I took the government to be making a series of there shall be no absolute rule and then, you know, one, two, three, four, maybe five.

[00:36:39] Speaker 02:
Do you disagree with the submission that there shall be no absolute rule?

[00:36:43] Speaker 05:
You know, we don't.

[00:36:44] Speaker 05:
We don't actually.

[00:36:45] Speaker 05:
I think I was in vigorous agreement with about 99% of what Mr. Kuhlman said.

[00:36:51] Speaker 05:
And so

[00:36:52] Speaker 05:
You know, with regard to, for instance, the brown shoe factors, Teradata really didn't try to make a brown shoe argument below until the oral argument on summary judgment.

[00:37:03] Speaker 05:
But really, even there, as the district court noted in footnote four of its opinion, it only really relied on one of

[00:37:10] Speaker 05:
the brown shoe factors, and that is industry or public recognition of the market.

[00:37:16] Speaker 05:
And we don't think that merely relying on one condition would be sufficient, but we really read the district court simply to be saying

[00:37:25] Speaker 05:
that the very case you cite, Teradata, requires three or more such indicia.

[00:37:29] Speaker 05:
And that arose in the context of a sub-market.

[00:37:31] Speaker 05:
But we obviously would acknowledge that you can use the brown shoe factors when defining the market as well.

[00:37:37] Speaker 05:
The other thing that Mr. Culeman focused on was the issue of ADR analysis.

[00:37:42] Speaker 05:
And Judge Tronto, I think you rightly exposed the fact that ADR analysis has not been invoked very often by courts.

[00:37:48] Speaker 05:
We're not aware of a case that has actually invoked it in the context of

[00:37:52] Speaker 05:
tying claims as opposed to in the context of a merger.

[00:37:56] Speaker 05:
And certainly Teradata cites no case in which a court has accepted ADR analysis under even remotely similar circumstances.

[00:38:04] Speaker 05:
But we don't read the district court to say you have to have actual pricing data.

[00:38:07] Speaker 05:
We simply read the district court as saying

[00:38:09] Speaker 05:
that the consumer resource management data, the CRM data on which you relied, data that Asper himself recognized may not be a reliable indicator is insufficient because that was simply a database indicating who the competitors were for

[00:38:27] Speaker 05:
SAP and others.

[00:38:29] Speaker 05:
It indicated nothing about whether customers would switch in response to a price increase.

[00:38:35] Speaker 05:
And the district court decisions cited by Teradata and the government were cases where there was more robust data, even if that doesn't rise to the level of actual pricing data.

[00:38:51] Speaker 05:
Great, thank you, Your Honor.

[00:38:53] Speaker 05:
And so I think, again, that we agree largely with the district courts, with the government's submission on those legal principles.

[00:39:02] Speaker 05:
The government takes no position on the bottom line here concerning the exclusion of Asper's testimony, I think because some of the relevant underlying material was under seal.

[00:39:14] Speaker 05:
But I think even with regard to the unsealed material here, the district court's exclusion was pretty clearly proper.

[00:39:21] Speaker 05:
And I'll say just a word about that.

[00:39:23] Speaker 05:
And of course, Ms.

[00:39:25] Speaker 05:
Maynard can respond to this on her rebuttal, because she didn't get as much of an opportunity to get into it in her opening.

[00:39:32] Speaker 05:
I think with regard to the exclusion of Asker's testimony, the exclusion of his qualitative analysis was appropriate, because there he defined the market in terms of large enterprises, which in turn he defined in terms of the number of

[00:39:48] Speaker 05:
employees and users at.

[00:39:52] Speaker 05:
And after simply failed to offer a methodology justifying that aspect of his market definition he relied on.

[00:40:02] Speaker 05:
two pieces of evidence that didn't even really support that definition, and he failed to come to terms with the conflicting evidence and conflicting industry practice.

[00:40:12] Speaker 05:
That is an appropriate application of Daubert, and notwithstanding Teradata's efforts to suggest that this very experienced district court judge, Judge Oreck, was somehow weighing the evidence.

[00:40:22] Speaker 05:
I think fairly understood the discussion, particularly at page appendix 28, was a garden variety application of Daubert.

[00:40:29] Speaker 05:
Now, Teradata spends a lot of time in its brief on the quantitative aspect of the analysis.

[00:40:36] Speaker 05:
And frankly, it's not even clear to what extent Asker was really relying on that.

[00:40:42] Speaker 05:
If you take a look at page 14, 121 of the appendix, he said that none of his opinions relied on the precise aggregate diversion percentages implied by the CRM data.

[00:40:53] Speaker 05:
But again, to the extent that he did, our fundamental submission is that the district court's decision can be sustained.

[00:40:59] Speaker 05:
on the ground that whatever the propriety of ADR analysis more generally and I think the district court rightly recognize that this is rarely used and again it's used predominantly in the merger context that ASCR did not base that analysis on reliable

[00:41:14] Speaker 05:
transactional data, the sort of data that you would need in order to make confident predictions about whether or not customers would, in fact, switch such that you could assess whether or not diversion is occurring.

[00:41:30] Speaker 05:
And if so, whether the amount of diversion would make allegedly anti-competitive practices rational.

[00:41:38] Speaker 05:
And so again, this seems like a fairly conventional application of Daubert to the facts of this case and not some sort of categorical ruling that ADR analysis is improper.

[00:41:51] Speaker 05:
Finally, with regard to the trade secrets claims, I think our fundamental submission with regard to the trade secrets claims is simply that the district court's decision was appropriate and appropriate really on both of its grounds.

[00:42:04] Speaker 05:
First, on the factual ground.

[00:42:06] Speaker 05:
that Teradata did not sufficiently mark its claim trade secret as confidential.

[00:42:11] Speaker 05:
The design document on which Teradata relies merely used the phrase batch merge.

[00:42:16] Speaker 05:
It did not disclose the actual trade secret.

[00:42:19] Speaker 05:
And the author of that document himself, John Grass, testified that he did not believe, quote, that he ever said that he conveyed the batch merge method in this document.

[00:42:31] Speaker 05:
Instead, he testified that he verbally conveyed the method

[00:42:34] Speaker 05:
during design calls.

[00:42:36] Speaker 05:
That is insufficient.

[00:42:37] Speaker 05:
That is pages 21970 and 21978 of the appendix.

[00:42:42] Speaker 05:
But even if you didn't accept that and somehow thought that there was a factual issue

[00:42:46] Speaker 05:
in dispute here.

[00:42:47] Speaker 05:
The relevant agreements here gave SAP the power to use this method because they drew a demarcation between, on the one hand, Teradata's own programs, software, or other material, which was provided under a project plan, that's section 9.2 of the agreement, and inputs under section 9.4 on the other.

[00:43:10] Speaker 05:
and proposed modifications to SAP's own software clearly fall under the latter category, as Judge Oreck correctly determined in his opinion, and that was a construction of unambiguous terms of the contract as a matter of law, and therefore... Can I just double-check something just to make sure I understand?

[00:43:32] Speaker 02:
The claim of trade secret misappropriation here is a claim of improper use, not of any

[00:43:40] Speaker 02:
disclosure by SAP, or is that not an applicable distinction here?

[00:43:46] Speaker 05:
I believe that that is correct.

[00:43:50] Speaker 05:
In other words, the claim here is that by incorporating this trade secret into SAP's software, that SAP was

[00:44:01] Speaker 05:
infringing on Teradata's trade secrets.

[00:44:05] Speaker 05:
That having been said, Judge Toronto, and this will be my final point because I see that my red light is truly on,

[00:44:11] Speaker 05:
Everybody knew that SAP was going to make these changes to its software available to others.

[00:44:19] Speaker 05:
I would point you to the evidence we cited on pages 75 and 76 of our brief.

[00:44:25] Speaker 05:
And indeed, Teradata went so far as to suggest that Teradata's own employees emphasized that fact in order to induce SAP to make the change to its own software.

[00:44:38] Speaker 05:
And in light of that evidence in particular,

[00:44:41] Speaker 05:
It seems a bit odd that Teradata would dispute the proposition that this falls within the suggestions of changes to SAP's own software side of the line.

[00:44:54] Speaker 05:
And so we would respectfully submit that if this court does exercise jurisdiction, it should affirm the district court's judgment.

[00:44:59] Speaker 05:
And otherwise, it should transfer the case to the Ninth Circuit.

[00:45:03] Speaker 03:
Thank you.

[00:45:04] Speaker 03:
Thank you, counsel.

[00:45:04] Speaker 03:
Ms.

[00:45:04] Speaker 03:
Maynard, you have four minutes for a bottle.

[00:45:09] Speaker 00:
Are you ready, Your Honor?

[00:45:11] Speaker 00:
We're hearing you.

[00:45:13] Speaker 03:
We hear you.

[00:45:15] Speaker 00:
Thank you, your honor.

[00:45:17] Speaker 00:
So I appreciate the indulgence and the extra time.

[00:45:20] Speaker 00:
So on the per se versus rule of reason, it's certainly not the holding of Illinois tool that the per se rule does not apply.

[00:45:27] Speaker 00:
If anything, Illinois

[00:45:29] Speaker 00:
Illinois tool reaffirmed that Jefferson Parish is the rule for coercive ties where, like here, a party like SAP has market power in the tying market.

[00:45:39] Speaker 00:
Epic and Microsoft are fundamentally different cases that drew a very narrow exception for when the tying product is a software platform.

[00:45:49] Speaker 00:
And the reason they did so, the reason they drew an exception from Supreme Court precedent, which is of course unusual,

[00:45:55] Speaker 00:
is because they saw benefits to third party app developers who were using those platforms.

[00:46:01] Speaker 00:
So it was a new form of competition.

[00:46:04] Speaker 00:
There's nothing like that here.

[00:46:06] Speaker 00:
All SAP does here is say the same things that every tying defendant says.

[00:46:11] Speaker 00:
We sell these because the products work better together.

[00:46:14] Speaker 00:
If that's true, then people will buy the separate product, regardless of whether it's tied.

[00:46:21] Speaker 00:
Even if you were to look at the pro-competitive benefits here, the hoist declaration doesn't have any cognizable pro-competitive benefits.

[00:46:29] Speaker 00:
They're just throwing up their hands and saying, we can't license.

[00:46:32] Speaker 00:
There's a dispute of fact about that.

[00:46:35] Speaker 00:
Couldn't grant summary judgment under the rule of reason without resolving that dispute of fact.

[00:46:42] Speaker 00:
The evidence shows they can indeed follow who's doing what on their products.

[00:46:49] Speaker 00:
On foreclosure,

[00:46:51] Speaker 00:
Judge Toronto, even if you think that the HANA with the runtime license is the only tied product, we have substantial evidence to create a dispute of fact on that.

[00:47:02] Speaker 00:
First, there is a dispute of fact about whether or not that can be used to compete with other EDWs.

[00:47:10] Speaker 00:
We lay out the evidence in our reply brief on 16 to 17, but there are SAP documents that explain exactly how to do that.

[00:47:20] Speaker 00:
The only thing that the runtime license prevents is exporting or importing data into HANA.

[00:47:26] Speaker 00:
It doesn't prevent using HANA's analytics to analyze enterprise-wide data.

[00:47:33] Speaker 00:
And I point you to all the sites that we have on pages 16 to 17 of our reply, which shows

[00:47:39] Speaker 00:
HANA can be allowed data loading from non-SAP solutions via the application layer.

[00:47:48] Speaker 00:
That's the BW product to which Mr. Shamigan referred.

[00:47:51] Speaker 00:
You can use, it is an inferior product, but you can use it.

[00:47:56] Speaker 00:
And the question is not whether any single customer has switched over or is even using HANA as an EDW.

[00:48:03] Speaker 00:
The question is whether it's foreclosing competition on the merits in the EDW market.

[00:48:09] Speaker 00:
So is it distorting consumers' choices?

[00:48:12] Speaker 00:
And Professor Asker's evidence shows that it is.

[00:48:16] Speaker 00:
Even if you take only their chart on page 48 by their own

[00:48:20] Speaker 00:
assertion, the vast majority of people are taking the EDWs with a runtime license, which, Judge Toronto, if you assume, and I'm just assuming with you, that that is the only tied product.

[00:48:33] Speaker 00:
Professor Asker did a differences in differences regression analysis that, despite what Mr. Shanigan says, did show the causal effect for Terradicca's

[00:48:46] Speaker 00:
lower sales as a result of the tie.

[00:48:51] Speaker 00:
So sales on HANA.

[00:48:52] Speaker 00:
That's why you do a differences and differences regression to separate out sales that weren't affected.

[00:49:00] Speaker 00:
And we have a dispute of fact on that.

[00:49:03] Speaker 00:
Even if you only look at the runtime license, it's still, as they say, the vast majority of customers.

[00:49:10] Speaker 00:
There's no reason to exclude the tying market here either.

[00:49:14] Speaker 00:
Professor Asker used a well-established qualitative methodology and carefully explained all of his opinions.

[00:49:20] Speaker 01:
Can I ask one question?

[00:49:21] Speaker 01:
Yes.

[00:49:23] Speaker 01:
Ms.

[00:49:23] Speaker 01:
Vayner, I just have one question on the trade secrets.

[00:49:26] Speaker 01:
If Teradata, and correct me if I'm wrong on the facts, but if Teradata offered batch merge as some kind of solution to HANA or SAP's own products, why isn't that as a matter of law

[00:49:41] Speaker 01:
an input rather than partner information that's controlled and remains with Teradata and can only be used for the bridge project.

[00:49:52] Speaker 01:
Or why is there a trialable issue on that?

[00:49:54] Speaker 00:
Because the way that the agreement is structured, Judge Hughes, is that Teradata came into the project under Section 10.

[00:50:01] Speaker 00:
It retains all of the IP it brought to the table.

[00:50:04] Speaker 00:
Batch merge is part of that IP.

[00:50:06] Speaker 00:
It's something they've been using for years with their Teradata-based product.

[00:50:10] Speaker 00:
And then the question is, did SAP get any kind of license to it and what?

[00:50:15] Speaker 00:
And under 9.2, partner materials, which includes tools, Teradata tools, are licensed only during the bridge project for a limited time.

[00:50:28] Speaker 00:
And that's the license that governs here.

[00:50:30] Speaker 00:
Their reading of input, which would essentially be in this project where everybody's working back and forth together all of the time during the course of the project, would subsume that reservation of rights.

[00:50:41] Speaker 00:
Because anything Terra Data has said to them to try to help the project work would then for be input.

[00:50:48] Speaker 00:
And it would, as a matter of law, the agreement shouldn't be read that way.

[00:50:53] Speaker 00:
At a minimum, there's a dispute of fact whether this is a tool that would fall within partner materials.

[00:50:58] Speaker 03:
Thank you, counsel.

[00:50:59] Speaker 03:
The case is submitted.

[00:51:01] Speaker 03:
We appreciate both arguments.