0:09 Uh as Brendan said, today's session is
0:12 on the connection between uh digital
0:14 assets and OTC markets. So we're very
0:17 fortunate to have Amy here with us today
0:18 from uh Goldman Sachs, but here in our
0:21 capacity as the chair of ISDA. So
0:22 welcome Amy.
0:24 >> Thank you. Thanks Peter. It's a pleasure
0:26 to be here. And I wanted to start by
0:28 asking you to maybe give a little bit of
0:30 a background on IST itself. Uh it's
0:31 obviously a wellestablished
0:33 organization, 40-year anniversary here
0:37 last year. Congratulations. Um but maybe
0:39 less familiar to onchain uh
0:42 participants. So uh can you give us a
0:43 sense of the types of things you've done
0:46 and are doing and the breadth of your activities?
0:47 activities?
0:50 >> Yeah, absolutely. Well, ISTA is the
0:53 global derivatives trade association and
0:55 our mission is to foster safe and
0:58 efficient derivatives markets for our
1:00 thousand plus members around the globe
1:03 and the purpose there is to make sure
1:06 that our members can manage risk with
1:08 confidence. Um, you know, Peter, as you
1:10 mentioned, ISDA was founded over 40
1:13 years ago, and it was founded because
1:15 the derivatives markets were growing
1:17 rapidly, and they were highly decentralized.
1:18 decentralized.
1:21 We're talking about bilateral contracts,
1:24 bespoke terms, and real question with
1:27 regards to legal enforcability.
1:30 Since then, ISTA has delivered on legal
1:33 documentation as well as market
1:35 protocols that the industry has adopted
1:39 at scale. And more recently, we've
1:41 actually extended that standardization
1:44 into data and processes through what we
1:47 call the common domain model. And that's
1:50 a digital representation of products,
1:54 trades, and life cycle events. And as
1:55 this group knows, that's really
1:57 important as markets continue to digitize.
1:59 digitize.
2:02 Today, ISD's work spans legal
2:05 documentation, market structure, public
2:09 policy, digital transformation, and we
2:12 regularly engage with regulators around
2:15 the world as markets and technologies
2:17 continue to evolve.
2:20 In our experience, markets scale when
2:23 there's legal certainty and operational
2:26 consistency and interoperability.
2:28 And that's exactly what ISD does for our members.
2:30 members.
2:32 >> And so tying that more to the digital
2:35 asset space specifically, um do you have
2:37 a perspective on things that are
2:40 happening digital asset markets now and
2:44 um the way forward in 2026?
2:46 >> Yeah, thanks for that question. ISDA is
2:49 focused on where digital assets
2:51 intersect with regulated derivative
2:53 markets to make sure that there are
2:57 clear usable standards. Uh first we
2:59 published standardized digital asset
3:02 derivatives definitions which is a
3:04 little bit of a tongue twister initially
3:06 covering instruments like
3:09 non-deliverable forwards and options on
3:11 Bitcoin and ETH. And this was really
3:13 important because what we did was
3:15 establish a common set of standards that
3:17 the industry could rely upon as opposed
3:20 to using bilateral bespoke one-off documentation.
3:22 documentation.
3:24 Second, we've done a lot of work on
3:27 tokenized collateral and that spans both
3:30 documentation as well as guidance so
3:33 that firms and institutions can use
3:35 tokenized securities as collateral with
3:39 confidence. And third, we engage
3:42 regulators globally on innovation with
3:44 an eye towards ensuring that we don't
3:47 undermine legal certainty and existing
3:50 riskmanagement practices as we continue
3:54 to fuel innovation. Overall, ISDA's role
3:56 has been to translate digital assets
3:58 innovation into frameworks that our
4:01 market participants and members can rely upon.
4:03 upon.
4:06 You mentioned a few of the drivers of um
4:10 markets that uh you were important from
4:12 your perspective at ISDA. Are there ones
4:13 that you would single out though that
4:15 are particularly different for onchain
4:18 markets versus trady markets that um we
4:19 should be more aware of?
4:21 >> Yeah, that's a good question. So, at the
4:23 end of the day, markets are markets and
4:25 so there are a lot of shared
4:28 commonalities and similarities between
4:30 regulated derivatives markets and
4:32 onchain markets. And so the same
4:35 fundamentals apply in either instance.
4:38 But there are also key differences. Um
4:40 you know one difference which this group
4:43 will be very well aware of is in the
4:45 infrastructure itself.
4:48 Onchain markets are built around programmability,
4:49 programmability,
4:52 instantaneous settlement and 247
4:56 continuous operations from day one. On
4:59 the other hand, regulated traditional
5:01 derivatives markets um we're we're
5:04 looking to find ways to retrofit these
5:06 capabilities onto existing
5:08 infrastructure and that's not an easy
5:11 nor simple matter. Another important
5:14 difference is in fragmentation.
5:16 In traditional markets, fragmentation
5:21 tends to be jurisdictional or venuriven.
5:23 Whereas in onchain markets,
5:25 fragmentation tends to be architectural.
5:28 And so ultimately, while the problems
5:30 rhyme, I would say that the tools that
5:32 we need to use to solve these problems
5:34 are actually very different across the board.
5:35 board.
5:37 >> Interesting. And I know you've taken a
5:39 particular interest in modernizing
5:41 derivatives markets during your time as
5:43 chair. Um so can you tell us a little
5:45 bit about your perspective on the role
5:47 of technology generally uh in financial
5:49 markets and tokenization in particular?
5:52 >> Yeah absolutely. Um so when I think
5:54 about tokenization I like to
5:57 differentiate between digitization and
6:00 core infrastructure change. A lot of
6:02 tokenization today is effectively
6:04 digital rappers and that can be really
6:07 beneficial for distribution,
6:10 programmability, composability.
6:14 But tokenization becomes truly powerful
6:17 when we start to compress the trade life
6:19 cycle. And that includes issuance,
6:23 pledge, margin, substitution, all the
6:24 way through to settlement. And we're
6:28 able to do this truly end to end. Um so
6:31 you know really that integration um is
6:34 is what's what's key. Uh it matters most
6:38 of all products I would say for cashlike
6:40 and highly liquid instruments because
6:42 that's where speed and certainty is
6:46 everything and matters the most. Um
6:48 along those lines though um I think it's
6:50 also important to note that tokenization
6:53 doesn't replace legal foundations. Um
6:57 the goal is actually law plus code,
6:59 right? So enforceable rights in
7:01 regulated markets paired with verifiable
7:05 controls in code. Um so it's really when
7:07 those things work together that we can,
7:09 you know, look at things that can be
7:11 truly transformational.
7:13 >> Terrific. And so one of the particular
7:15 projects that I know you is has been
7:17 involved in and continues to be involved
7:21 in is uh collateral mobility um
7:23 including the use of stable coins and
7:24 tokenized money market funds as
7:27 collateral. Uh can you give us a sense
7:29 of why that use case is so important and
7:31 how tokenization is helping?
7:33 >> Yeah. Well, collateral is absolutely
7:36 central to derivatives markets. It's how
7:38 we manage counterparty risk and it's
7:41 also how we ensure that markets remain
7:45 um stable. So in recent years we've seen
7:48 why collateral mobility is not just an
7:51 operational matter but can be a systemic
7:54 issue during times of stress like the
7:58 2020 dash for cash. External shocks
8:01 caused market volatility which resulted
8:04 in spikes in margin calls. And when that
8:06 happened, we had market participants who
8:09 had to transform large amounts of um who
8:12 had to source large amounts of cash in
8:15 tight time frames in order to to post
8:17 post that cash as collateral. And that
8:19 created a liquidity strain and in some
8:23 instances even amplified market stress.
8:25 Today there are well over a trillion
8:28 dollars in variation margin in the
8:31 system. And as that continues to grow,
8:34 even small frictions, small operational
8:37 frictions, um, you know, manual
8:40 processes or settlement delays can
8:42 become destabilizing when markets are
8:45 under stress. And so that's why ISDA is
8:48 focused on collateral and tokenization
8:50 of collateral. Collateral movement can
8:53 be operationally intensive with
8:55 workflows that weren't designed for
8:59 speed nor continuous operation.
9:02 um in ISTA has responded in a couple of
9:05 ways. First and foremost, we updated
9:08 documentation to ensure that tokenized
9:10 securities as well as stable coins can
9:13 be used as collateral and we followed
9:15 that with legal guidance on the
9:18 enforcability of tokenized collateral.
9:20 Um I I think it's also important to kind
9:23 of call out an example. Money market
9:25 funds are a good illustration of why
9:28 this matters. When you think of money
9:30 market funds, you think of trillions of
9:34 dollars globally, low volatility, and
9:37 small regulatory haircuts. But when you
9:40 actually try to use money markets, money
9:42 market funds for collateral purposes,
9:44 you'll actually find that the workflows
9:46 and processes are quite cumbersome,
9:49 resource intensive, and can be costly.
9:52 because in most instances you actually
9:54 have to transform the money market fund
9:56 into cash before you can post it as
9:59 collateral. And so, you know, this is an
10:01 area where tokenization has the
10:04 potential to really improve some of
10:06 these operational efficiencies so that
10:10 you can actually post direct shares or
10:11 interest in money market funds as
10:13 collateral instead of having to go
10:15 through the cumbersome conversion process.
10:16 process.
10:19 tokenization can help address another
10:21 challenge which is making sure that
10:23 collateral gets where it needs to be
10:26 quickly and you know this of course
10:29 leans on um the the accelerated
10:31 settlement mechanisms that tokenization
10:35 can offer um which removes or eliminates
10:39 industry reliance on T1 T2 types of
10:42 processes and so for ISDA it's really
10:43 about strengthening riskmanagement
10:46 plumbing through sound legal framework
10:48 frameworks, interoperability, and close
10:50 engagement with regulators so that we
10:52 can function with uh with modern
10:55 technologies going forward.
10:58 >> Well, that's great. I mean, uh it sounds
10:59 like that's just a very challenging use
11:01 case with a lot of moving parts. I know
11:03 another technology you're particularly
11:05 interested in is artificial
11:07 intelligence, so AI. Is this the type of
11:10 use case where AI can play a role? um
11:13 and if so um how do you think about both
11:15 their capabilities and then the the
11:16 risks that might come along with the use
11:17 of an AI tool?
11:21 >> Yes. Yes. And absolutely. Um collateral
11:24 management today is hugely data and
11:27 process intensive and you know this is
11:31 where um standardized data models like
11:33 ISD's common domain model can be so
11:36 helpful is because AI can only be as
11:38 good as the quality of data that it
11:42 operates on. And so when it comes to um
11:44 the collateral management space, you're
11:47 dealing with large portfolios, frequent
11:51 changes in valuation, eligibility rules,
11:54 and regulatory constraints, all of which
11:57 are moving rapidly, especially during
12:00 times of market stress. And so, you
12:02 know, AI can alleviate some of these
12:05 complexities. For example, AI can be
12:07 used for things like documentation analysis,
12:09 analysis,
12:11 um you know, exceptions management as
12:14 well as um collateral optimization, just
12:17 as some examples. Um but it can also
12:21 help firms anticipate market stress by
12:23 monitoring patterns in things like
12:26 margin calls and capacity.
12:28 Now, one thing that I think is really
12:31 important to recognize is that when it
12:34 comes to things like risk management and
12:37 margin, AI is meant to be a co-pilot. It
12:40 can augment human judgment but shouldn't
12:42 replace human judgment. And so, you
12:45 know, I I jokingly call it the bionic
12:48 arm because it's meant to aid the people
12:50 who are making these types of decisions
12:53 at various firms. Um, across the board,
12:55 I think that strong governance,
12:58 highquality data, and clear
13:00 accountability are absolutely paramount
13:02 as we continue to forge forward with
13:04 these novel capabilities.
13:06 >> Fantastic. Well, we're just about at
13:09 time, but I wanted to ask if if there's
13:11 um a closing thought or two you might
13:13 have about lessons learned from your
13:15 experience in markets over time and the
13:17 like the times we're living in now with
13:19 the increasing role of digital assets.
13:21 >> Yeah, absolutely. In our experience,
13:24 scale tends to follow legal certainty
13:27 and um you know the same is true for
13:30 interoperability. And so as technology
13:33 continues to move quickly, I think it's
13:35 absolutely critical that legal
13:38 frameworks, operational processes, and
13:42 technology evolve together. And ideally
13:45 that'll be through collaboration between
13:47 traditional institutions and digital
13:50 native builders. Well, fantastic. Well,
13:51 thank you so much for joining us, Amy.
13:53 It's been a nice discussion, and we'll
13:54 let everybody go to lunch. Uh, with
13:55 that, thanks so much.
13:56 >> Thanks, Peter.
