 We're live. Hi. I'm Dazza Greenwood, a scientist at MIT Media Lab, and I run law.mit.edu there and this is a lecture for our upcoming computational law workshop course at MIT something that we've done every every January for the last four years and With us today is a lecture so good That he lectured last year and we want to back this year for Bit of a deeper dive in his topic. I want to introduce Chris Berent, a partner at Drinker Biddle and a colleague and a friend to talk about the concept and the practice of computational Contracts such a critical part of the law and perhaps some of your experience Applying and practicing law such that these computational contracts can support and reflect Your deals and transactions in the energy market as an example With that Chris if you if you'd like to maybe introduce yourself a little more of your background and please jump right in and thank you for joining Thank You Dazza. That's a kind introduction My name is Chris Berent as Dazza mentioned I'm a partner at Drinker Biddle and Reeth I run the advanced energy practice here My my practice my team focuses mainly on transactional matters and the advanced energy space microgrids advanced clean energy systems DER aggregations basically think of smart grid applications more intelligent assets that are more flexible more dispatchable and When we're not doing deals, we are working regulatory dockets in front of the Federal Energy Regulatory Commission State Public Utility Commission and so on focused on grid modernization and advancing the tariff environment and the rural environment so that we can Have better deployment of more advanced more cost efficient higher environmental Performing energy systems. I got to know Dazza in my work with clients in both the energy and environmental markets Regarding applications for enterprise blockchain automated contract systems smart contract systems and so on so Computational contracts automated contract systems smart contracts What are we talking about when we're talking about all this? Well, it's certainly more than just the contract account on Ethereum The ability to establish contractual relationships that have automated properties that can look at market conditions asset statuses and Make decisions to optimize asset portfolios is something that extends Well beyond blockchain although blockchain technology offers a recordation and a Ledger that can support these automated contract systems Which are increasingly taking in more data analyzing more data and taking Folks who are using them to ever more complex positions and opportunities and so on so a Computational contract. What what does that look like? Well at the most simple level. It's a writing Well, you might say but what happens if that contract exists electronically on servers whether it's distributed or semi-distributed Whether it is, you know permissioned or you know wide open for anyone to download the clients You know to start up a node and so on Regardless of how it exists If that contract exists in writing even if it is electronic it is still widely viewed as binding you can have An email a bind two parties together You can have contracts electronically signed. I'm sure many of you have used a lot of the popular features out there That have evolved. They're the last several years like doc you signed and others to design electronic agreements and Electronic writing can bind just as a a written copy could could bind so much of the question as to the application you know the legal enforceability of smart contracts Computational contracts automated contract systems whether or not they reside And execute on blockchains or not Is fairly solid in that they're electronic writings and if that the intent is there and if you have all of the elements That allow for a contract to be formed within your jurisdiction You will have a contract form And in many cases it's a lot easier to do than folks expect in terms of creating a binding contract And if you miss a few of the elements And you have enough of the core elements in that contract You'll have the jurisdictions common law or civil law system fill in the gaps And so here in the United States that usually means that the uniform commercial code would kick in and address it and you know You eat a you know the uniform electronic transfer acts Uniform electronic acts that are out there that allow for those electronic writings to be binding You know really don't have too much of Question marks around them as to the enforceability of that signature of that you know that you know binding nature of those agreements so Much of what we're left with is really are the Recordation systems around smart contracts adequate is the vehicle itself and the mechanism around a smart contract or automated contract system You know adequate to perform in the market in operational environment What I've done traditionally is help clients you have complex asset portfolios Optimize those portfolios and what I mean by a complex asset portfolio is you have assets that in any given You know hour of the day, you know day week or so on can Take different operational positions Some of the best examples of this are Integrated energy and manufacturing facilities think industrial parks where you could have someone own Generation on site combination of solar storage high efficiency co-generation you could also have Manufacturing production that's happening on site at any given time that industrial park Might find it more lucrative to make the widgets that the manufacturer is producing Because they can sell them under contract at that time of production for a higher value Then they might have not making the widgets and say doing something like Providing the power that the manufacturing facility would have used out to the power market If you have a gas natural gas fire co-generation plant on site You know that plant might decide that instead of producing power and steam It's more lucrative at that time to sell out its natural gas to another party or sell out some natural gas Transportation position to another party The natural gas commodity and the the transportation rights are two different things That that same facility could also decide to work and make power But not just sell power into the power market. It could sell ancillary services frequency regulation. It could sell Spinning reserve services. It could sell any number of different goods and products and services So much of what I've been working on Over the years is helping clients where you have these highly flexible and dispatchable groups of assets I've I've worked with folks to essentially Look at how can we create? complex multi-point arbitrages to maximize the the value of how those assets can position in any given a market condition and You know market day so, you know, how do we do this? Well from a legal standpoint, it's more simple than you might expect We use what are called master contracts Master contract vehicles are very common. You have a lot of model agreements in the energy sector In the manufacturing sector other sectors Where you set up a master agreement that basically says that the parties agree to these terms and conditions when we undertake a Transaction and that the transactions themselves are governed by confirmations which are basically, you know short descriptions of here's how much I'm selling you at what price and when Where you take where are you taking delivery and what's the nature of that product? Or that service that I'm selling you That type of master agreement To create an automatic contract system You often will stack several different types of master agreements that focus on the different markets Products and services that can be accessed by that complex asset portfolio I was talking about either a group of assets physically located or a portfolio of assets spread out, you know around the country around the world you can stack those master agreements together and the confirmation process can get handled Computationally on an automated basis What you can do is you can have the master agreements say that under certain conditions I have the right to sell you this product or service or buy from you this product or service and what We often have looked at doing is creating what's called a married order algorithm And what the married order algorithm basically does is just create scenarios Based on the different types of positioning that that asset portfolio can a step can take and what it does is it pulls from What some people have referred to oracles what other people would call simply market marks That are you know signals in the market this market is pricing here. This service is pricing there and so on Calculates those in Runs all of the various scenarios that that asset portfolio could take in any given time period Based on all the various market conditions across multiple different markets some of them unrelated And then we'll basically rank on a dollar basis. What is the most profitable? Overall positioning of this group of assets given all the various markets that they can access and sometimes it takes You know more than one asset to truly access that market Sometimes a market won't be lucrative unless I get a certain amount of threshold of volume into it Sometimes a service discount doesn't kick in unless I'm buying a certain amount so it's the combination of different resources that help to make all of this possible and what a married order algorithm basically helps folks do is Rank given all the various market conditions. What is the most lucrative scenario? That is physically functional operationally That accesses various combinations of markets It might be in anyone given day that the best thing is to just make as many widgets as you can It might be in another given day that it might be the best thing you can do is produce as much power and sell as much power The wholesale power market as you can it can vary it, but it's also often fine combinations of different market offerings service purchasing and so on and so what the married order algorithm does is assesses all those Those those potential permutations based on the various market conditions. Does it automatically? builds the scenarios ranks the scenarios based on revenue possible and then reports back out saying This is the most lucrative positioning of our asset portfolio today or in many cases in the energy sector We're talking about the next day. It's a day ahead market and then basically that Algorithm kicks out to the various systems Electronic confirmations to the various parties the different master agreements Kicks out electronic confirmation saying I want to buy this much natural gas I want to sell this much power to the power market I want to produce this much widgets and sell them to this customer this customer for this this, you know shipment in order as opposed to this one and Essentially those confirmations can flow out to multiple counterparties get processed and then come back Now this is similar to what a lot of people are used to in the security space And so on when you think of things like limit orders where you tell your broker Hey, when the market reaches this point sell or when the market reaches this point buy But what this what's different here is that? the amount of data information coming in across multiple markets to run All of the various scenarios rank them in the merit order algorithm and then kick out a set of confirmations that reflects a Complex multi-point arbitrage of your assets into the most valuable markets that's something that's getting increasingly more complex and When we're what we're seeing in the power sector is a push towards more distributed energy More energy at the distribution level more energy resources Generations storage, you know demand response in particular at the building campus, you know You know a level at the district level more micro grids What we're what we're seeing in the energy sector in particular is the opportunity for there to be more complex arbitrages happening amongst more asset portfolios at different times as well as consumers Pulling together and creating new market layers trading amongst themselves Prosumer is one of the terms that we've been using in the power sector But all it really means is that instead of just buying power you are actively buying and selling power Utilizing your own assets that you have on your building often solar panels coupled with battery storage And I have a number of clients that have done some very interesting things in that space and who are actually opening up with DLT enabled apps New market layers and the power sector that haven't really existed before what's new products and services That go with those new market layers that haven't existed before And all of that drives greater opportunity for efficiency in the system Because unfortunately our energy system traditionally is widely inefficient From you know at for whether it's at end use in terms of how we use energy whether it's at production much of what's characterizing the future in the advanced energy world is not just Renewable energy clean energy battery storage, but it's also high efficiency combined heat and power systems that operate in the upper 80 percent on their energy efficiency as opposed to the lower 40 percent on traditional big power plants and so on and so We really have an evolution and why am I talking to you about these computational? contracting principles and multi-point arbitrage systems You know in in this class with Daza, it's not just because they're interesting computational You know ways to undertake contractual relationships legally It's also because the the breadth of the datasets the aggregation in the nesting that can occur at the you know at the Asset grouping portfolio level and then amongst different arms length third parties who are aggregating the you know similar assets into a portfolio voluntarily and so on or through an aggregating entity and Such we're just talking about large large datasets Multiple different parties and again some of the value propositions We've seen with the LT technology in terms of the ability to have certain trustless transactions the ability to have immutable ledgers The ability to run Faster settlements because settlement speed is still not really a characteristic of blockchains right now But the ability to run faster settlements that record those complex arbitrage positions And not only the strikes and when the confirmations execute But also what is the statuses of the various markets and other conditions that are playing? How does all that data you know fit together and and what do those arbitrage points look like? As they cycle through the day and the markets change and volatility moves through the various systems different needs arise In short, it's a lot of data and that's why having Distributed ledger technology can be a big help and in particular what I've been working on with folks is not the full bore Fully distributed, you know, anyone can get down to load a node view But rather because we're dealing with things that are utilities that are providing You know, you know important goods and services We're not, you know, we're not looking at it as a you know Just kind of what can happen on a wild west basis, but rather we're looking at enterprise blockchain technology permissioned blockchains semi-distributed nodes You're looking at hashing algorithms that are going to be running on things that are relevant to the underlying markets Products goods and services that they're serving as opposed to just who can run basic math the fastest And therefore you're going to see these enterprise blockchains really being run and set up among groups of market participants You know who have all of the other You know, let's call it, you know gears in the machine that allow for What's happening from the enterprise blockchain dynamic to articulate through to actual market operations market trades? And you know all the way down to the SCADA systems That are actually telling specific assets ramp up ramp down fire up fire down and so on so It's it's a brave new world. We're going into and there certainly is a lot of interest and dynamics Around how the energy sector can benefit From this technology But I tell a lot of people that especially when you look at large international manufacturing Corporations and others who have a history of running Operations facilities where there's multiple different assets power production manufacturing natural resource extraction you know any number of whether it's a You know telecom, you know you name it There's any number of various applications that can go on a lot of those industries have already been running Early versions of automated contract systems and that as we get better recordation better computational dynamics running In our contractual space, you know, you're just going to see those becoming more complex with more and more data applying But there's still a wonderful amount of white space. I have run into clients That are large international entities that are running, you know Substantial operations holding companies with with many different operational subsidiaries and sometimes those entities have yet to fully explore What types of multi-point arbitrages and portfolio optimization? Strategies are available to them within their own portfolio sometimes until we start that analytic view of Where are you doing transactions and what data are you sourcing to do transactions? Sometimes we we find out that some of the best data available is with another affiliate and you had no idea And that that other affiliate might have a arbitrage capability that you weren't aware of that when put together at the portfolio level can make a big difference so Overall, it's just better ways to use more data But from a computational legal standpoint I would encourage people to think of computational contracts as things that combine and part of the part of the challenge here Is just making sure that when those venture into Let's call it traditional Legal areas where we have mechanisms beyond just the signature Notarization as an example on let's say we were doing a lease long-term lease a deal for For a new facility, you know, you're gonna want to get those needs at least is notarized How that notarization can get done from electronic basis is still something for the future And to get determined And who knows maybe the Secretary of State's office is in the future will figure out that collectively they have a decent interest in creating their own enterprise blockchain but As things stand now you still need to have lawyers involved to know what boxes need to get checked To get deals done binding and enforceable But there are more opportunities out there, especially in the operational in the market side Where often all you need is a binding signature from an authorized officer to get a binding deal And you don't need a lot of the traditional Notarization title filings and so on that need to happen physically In hard copy, you know a locally or at the state level or whatnot a lot of what happens operationally in the markets can get done simply with electronic signature And if you're willing to to get creative and take a holistic view of your asset portfolio There's a lot of people who are realizing that a DLT technology And looking to begin to automate their contracting contracting systems can open up these multi-point arbitrages and bring You know substantial Incremental revenue to existing asset portfolios. And so with that Daza, I'll I'll kick it back to you to see if any questions Okay. Thanks, Chris. That was Spectacular, I appreciate you taking the time to lay out some of the basic building blocks of computational contracts and how it kind of fits into this concept of computational law and and the sort of extension from You know people are familiar with high-speed trading and the heavy automation that's occurred in the security Markets and being able to show how that is playing out in the energy markets Is is really helpful. So before I ask a question just a couple of them a couple more housekeeping matters When you say DLT for those of you that may not be familiar that stands for distributed ledger technology Which is another take on blockchain but using but really when you're applying the blockchain not so much as a log of cryptocurrency Kind of transactions, but you know, perhaps as a ledger for other types of transactions Like including fiat currency or perhaps units of energy or you know other types of commercial Exchange of rights to some asset. So that's one thing. The other thing was you use the phrase complex Multi-point arbitrage a few times and I think from the context people can probably get it But it would be useful if you could just take, you know, like five or ten seconds just to Share what you mean by that phrase sure well at the simplest level of an arbitrage is a way to Reposition what you're doing to access one market or one client versus another market or another client It's the act think of it as as taking a good or a service that's going to be provided Or consumed and pivoting to another one and having the legal rights and operational capability to be able to do so A classic one in the power sector is that if one fuel for a power plant is getting very Expensive and that power plant is able to fire on two different fuels. They will arbitrage and switch to firing with another fuel Another example is I might sit on the edge of two different regional power markets And I might decide at any given time that I'm going to put my power into one versus another The ability to make that decision and not be bound to one offtake Is a example of an arbitrage and the simplest ones You know are are basically You know do I go here or do I go there the a complex multi-point arbitrage looks at the overall value of assets in a group Or in a portfolio in a manner to say What are all the variety of positions that these assets can take to physically provide different services to different markets? To different customers and then basically says Looking at all those various scenarios pulling from maybe five six seven eight, you know or more different market conditions What is the overall optimal position for that asset grouping at any one time and the complexity is Running all through all those different markets through all those various different positions the different ways you can position that asset And then saying at any given time Here's the most lucrative positioning, you know for those assets in the different markets I'll note. I have clients who run the same merit order algorithm, but they don't do it for What is the highest revenue basis for the asset? They do it on what is the way that we can make a decent revenue And what is the way that we can position the asset portfolio for the lowest level of carbon intensity? In other words they operate their micro grids and other assets in a manner to reduce the overall amount of carbon Admissions that they're putting out so there's different ways you can put rule sets in these merit order algorithms Maybe you have one supplier who's favored more than another for whatever reason for other affiliate reasons So that you program it to give them more points to the scenarios that favor selling to them or buying from them But the the complex arbitrage simply refers to the ability to put all of your assets in a Large variety of different positions as opposed to saying Two or three different ways I can position all of this when you zoom out to the portfolio There can be hundreds of different positions that all of your assets can take at any given time and the most lucrative ones Will be based on whatever the markets are At at any given time and those markets those market marks will get automatically pulled into the algorithm Through the you know through the the merit order system that's developed And those can get developed on on you know basic Excel spreadsheets and people can do that and and run a simple ranking Algorithm to rank the output of different tabs and then there's also other programs In such that that folks have developed that can just do a simple kind of scenario ranking where you have manual scenario Formulation and you're just basically kicking in numbers, but The future is going to show these arbitrages does a happening in at more Portfolio more aggregated levels with lots more volumes of data with lots bigger data sets and with you know a Kind of Kevin Bacon like you know new relationship amongst affiliates and others Especially at holding companies as they begin to figure out and better analyze What not only are the goods and services that are flowing through my different subsidiaries and the different affiliates But what are also the data that they're taking in at different stages in different levels? I think much of the future of enterprise blockchain and asset portfolio optimization is going to be started with Exercises at the holding company level Looking at what are the data flows product and service flows that are coming in throughout all of our Portfolio companies throughout all of our subsidiaries and so on Analyzing that on a roll-up basis and then looking at where are the common data sets? Where are the where are the common flows? I think a lot of people are going to realize that there's you know a lot of just efficiency and savings You know from becoming more aware and having better visibility into their overall asset data Operations and how that relates to goods and services in market position. So a long answer But multi-point arbitrages are just going to get more complex the more assets that are involved and the more markets that are involved Awesome. Thanks. I appreciate that because some people may not be familiar with with you know What that phrase means so that's very useful. So So it's a sort of synthesize and and start to wrap when when you talk about merit order algorithms and data sets and Scenarios some other words that people might be familiar with there in data science would be like Prioritization for an optimization algorithm of some kind and and in terms of the data You know the first two lectures that that students will have seen by now I think are the high-level overview of computational law from a talk I gave in China and then a more recent sort of Survey of computational law in a deep dive into a open source platform called doc assemble that Co-instructor Brian Wilson did a few days ago and what we really tried to focus there on data and Computation to create more automated functions and flows That are legal For different legal instruments, I think to me in my world the classic legal instrument is a contract Obviously you have wills and deeds and you know on and on down the line, but contracts uber Alice You know rid of Hammurabi forward. Yeah, and there's so common all across the economy and society And so I'd like to encourage students to be thinking Uh In the context of what you've just said Chris about what other types of contracts can you imagine where there could be data inputs that would Where you can where we could construct a smart legal contract of some type that can have some listeners, you know over interfaces For events, you know such as you know a market mark as as Chris said or You know an inventory or whatever just be creative in the context here so that it could the event It basically could construe like going through a threshold or some type of data as an event that would trigger a process That would you know could be something like sell or price it this way or send a proposal or whatever Or do this and then that and then And and and share your ideas in the pigeonhole a little Engagement widget that you're probably seeing right now. I'm Chris's session page And then upvote on the ideas that you think are most interesting and that you'd like to do a deeper dive in when we get to class so So think about the data think about the the underlying legal context and How you could use listeners and events and triggers for more automated processes that Don't simply that aren't just the technical layer that's disconnected from the contract But that that are the contracts. So how do we structure these contracts so that they're very explicitly Supporting and reflecting these automated processes and then as Chris said Quite rightly and we're going to get more into this in a future lecture, you know, the traditional components of contracts like a scent through a signature and The clauses in terms of a contract absolutely can be expressed electronically thanks to electronic transactions act and e signatures and National commerce in global and national commerce act in the US and all the equivalent legislation from unto Charles model on Electronic commerce in other jurisdictions And what we'll do is we'll get a little bit creative about how to structure the transactions and structure the deals So that we're meeting all the legal requirements, but then seeing how far we can Go with with more creative Adoption and adaptation of this technology. So it so it is an expression of the legal contract itself So everybody be thinking about that no take and what Chris just told you is street legal I'm like he's done this many times in the energy industry This is not futuristic and then the last thing I wanted to highlight is it does not require Cosmic technology when you said spreadsheet, I almost went off mute and I didn't want to interrupt your flow but amen like a spreadsheet especially these online spreadsheets Google sheets and and Office 360 they have interfaces Application programming interfaces that can automatically update data and you know sell by sell basis and sheet by sheet basis and that can send Triggers or it can send information and can be interrogated by other services to see what is the current value of a given You know kind of a number or or other information in the sheet And so we can begin to prototype some of these computational contracts using spreadsheets And it's not different in kind from what one would do by building a huge Python or or you know or or blockchain based automated system so With that I want to thank you Chris very much for for joining us from your from your Washington, DC office there And and sharing your knowledge and and I can promise that as we start to get the feedback from students and from others In our community to your lecture and they start to brainstorm ideas or have other questions or comments We'll wrap up. We'll wrap those up using our own merit order algorithm, including the upvoting On on pigeonhole and ensure those with you and we'd be very very grateful For any feedback that you have on on the input from our students. Yeah, and I look forward to it. Thanks, Taza Okay, thanks, so we'll see you online