SGX's Rehan Ahmed and Temasek's Pradyumna Agrawal discuss the joint venture between the two organisations to build a full fledged infrastructure for digital assets in Asia, its growth potential and drivers as well as issues and challenges surrounding building an end-to-end infrastructure and adoption by issuers and market participants
Singapore Exchange (SGX) and Temasek, the Singapore government owned investment company, announced in January 2021 a joint venture (JV) that claims to be Asia Pacific’s first exchange-led digital asset programme focused on capital markets workflows through smart contracts, as well as ledger and tokenisation technologies. The JV called Marketnode builds on a previous initiative between the two organisations and HSBC, as arranger bank, that issued a public syndicated digital bond for Olam International in August 2020, said to be the first of its kind in the region.
SGX’s digital asset issuance, depository and servicing platform was used to issue four digital bonds by several issuers, with a total size of over SGD1 billion ($754.26 million).
The Marketnode partnership combines SGX’s multi-asset experience operating in regulated market infrastructure, together with Temasek’s expertise in blockchain technology and ecosystem connectivity. The venture had also acquired a minority stake in Covalent, B2B financial technology company specialising in capital market solutions.
It will tap on the accelerated need to digitalise and optimise workflows to meet demand for capital in Asia, starting with fixed income (FI). The partnership is poised to consolidate primary FI workflows as bond issuances continue to grow in Asia ex-Japan, where G3 (US dollar, euro and Japanese yen) primary issues have grown 46% to $387 billion over the last two years, totalling more than 800 new issues in 2020 with more than 450 issuers and 200 arranger banks.
Marketnode and Covalent will work together to streamline the listing, straight-through processing and settlement of bonds and activities in bond lifecycle management. It will connect Covalent’s OMAS platform, which is a front-end issuing, data, book building and allocations tool, with SGX’s listing, post-trade and asset servicing capabilities, thereby providing the Asian bond market with a one-stop listing, issuance and lifecycle management platform.
In this TABLive interview, Rehan Ahmed, head of fixed income products and digital assets at SGX and Pradyumna Agrawal, managing director at Blockchain@Temasek discussed the Marketnode joint-venture and partnership with Covalent to create a full-fledged infrastructure to service digital assets for financial market participants in Asia Pacific.
Here are the key points discussed:
Here is the full transcript:
Foo Boon Ping (FBP):
We are excited to have with us for this TABLive interview. Mr. Rehan Ahmed, head of fixed income, product and digital assets of Singapore Exchange (SGX), and Mr. Pradyumna Agrawal (Prady), Managing Director at Blockchain@Temasek. The two have recently come together to form a joint venture to create digital asset infrastructure for the financial markets. As we know, the evolution of digital assets or the use of blockchain, distributed ledger technology for the financial markets is gaining momentum. SGX and Temasek is one of the more recent initiatives at creating or furthering infrastructure development in this area. And they have also issued the first public syndicated digital corporate bonds for Olam International in August and since then they've issued a total of $1 billion worth of digital bonds. We are very interested to find out from both of them, how this joint venture came together. But before that, we want to explore the potential for digital assets in Asia Pacific. And I’d like to ask both of them the potential and growth trajectory for digital bonds as one of the many digital assets as a new asset class in Asia Pacific, and how is it being currently viewed by issuers and investors?
Digital bonds are not a new asset class
Rehan Ahmed (RA):
Thanks Boon Ping. First off, whether or not digital bonds are really a new asset class and what our approach here is. If you look at the Asian bond market, it's close to, when you look at hard currency, Asian bond market - US dollar, euro, Singapore dollar, it's a $2 trillion outstanding asset class annual issuance of $400 billion. What we're not out there to do is create a whole new asset class called digital bonds and fragment that market even more Our approach is looking at how bonds are issued and serviced given their current workflow, how bond issuance works with arrangers, lawyers, paying agents and investors, and really try to digitise that process. So as opposed to creating a whole new asset class, this is about taking an existing asset class and servicing it in a digital fashion, which is something that doesn't exist within the primary markets today.
FBP: For Blockchain@Temasek, it's very interesting how you come into this joint venture. Could you tell us in terms of looking at blockchain technology and applying that to financial markets, why is Temasek involved?
Pradyumna Agrawal (PA):
This is part of a broader effort at Temasek to firstly look at emerging technologies, blockchain being one of those technologies. If you look at blockchain itself, the view has been that there are a few use cases which can have quite a game changing effect. And therefore, can Temasek build expertise, which we have over the last few years, and also bring up productivity, connectivity to accelerate the development of those use cases. Our efforts are divided today into three areas. We hope to expand that in other areas. One, obviously, which we're talking about today, which is digital asset infrastructure, and we'll go into more depth. Second is actually somewhat related, and eventually hope to bring together which is money itself - How do you make money more programmable? Our involvement with Project Ubin, and subsequent efforts are very public. And the third is around identity and data. We see these as foundational pieces where we are bringing in a combination of our expertise, capital, attracting talent, and then working with partners to create foundational technologies, which can then help accelerate this space. That's really how, when you're thinking about capital markets and building digital asset infrastructure. That's where we landed up developing first a POC (proof of concept), and then now commercialising the technology with Singapore Exchange, and very excited with what this holds for the future.
FBP: Looking at the technology today. Taking the Olam digital bond for example, the technology today is used for workflow, the application of smart contract and so on and so forth. You also mentioned that you'll be exploring tokenisation as well, creating existing assets as well as new emerging assets. Could you tell us more in the area of tokenisation? Are you also potentially looking at other areas like cryptocurrencies as asset classes? How would you define blockchain assets?
Lack of common standards and potential bifurcation of liquidity hamper tokenisation of traditional capital market assets
RA: Sure, so I'll cover what we've done in bonds and we can build on from there. Within bonds, we have actually not gone down the tokenisation route. The first step we actually did in the creation of our platform was to look at what is being done today. Because our guiding principle always has been to make life easier for everyone that interacts with us as both a listing venue and a depository. And mapping out that entire process was the most interesting exercise. A typical bond issuance, believe it or not, has more than 650 individual steps involved during the issuance process. And all of this can happen in a matter of, anywhere from one to one and a half weeks. Quite a compressed timeline. Lots of fragmented processes. Our approach was really looking at how can you automate some of these processes through usage of smart contracts. Generally, when you look at a tokenisation play for assets, it has a distribution play in there, so maybe you want to open up new pockets of liquidity. For the pilots that we ran. That certainly wasn't our approach. We did a very deep dive study on how primary processes work for syndicated bonds in the Asian markets. And we found our specific areas that we could automate. I would say rather, we created a digital security that rests on a ledger that we operate. when you start looking at tokenisation of traditional capital market assets like bonds, I don't think we are landing yet. Certainly from my standpoint, it would be difficult for me to describe what a token standard in today's market would be like. We're aware of at least three different token standards that are being used. So really, the focus was on making the issuance process easier, investor liquidity. We'll discuss this a little bit later. But that can be accessed through the investor custodians, we try to make what they do with us a lot simpler, versus trying to create a token and distribute within another ecosystem. Because one of the challenges you have when you start to create markets for tokenised bonds versus analogue paper bonds as we have them today is you bifurcate the liquidity, which certainly wasn't our intent. Now that the use case is there, and we've been through a pretty complex mapping out of the entire primary process, what can be automated, I'd actually say it's quite easy to add on other asset classes. Bonds as I mentioned, 650 plus steps involved in the creation of a new issue. You can imagine that maybe adding on simpler fixed income securities is not that much of a lift. Going forward, it's really about having a ledger of securities, bonds are a fairly complex instrument, they've been the first to be added, and to expand horizontally across asset classes.
PA: I’d just add one point there. Once we've got some of the core infrastructure being built out, the ability to extend that into other asset classes is much easier. Even on bonds, for example, we can now bring on much shorter-term paper, as an example. Some early conversations that we've had, it was not easy to service, and actually create a market around some of these shorter tenure bond instruments. Funds is another area where we're already beginning to do some early work, where you will deliver many of the benefits that have been created in the market infrastructure for corporate bonds that we see. So the intent here is to quickly fast follow into other areas and tool it out for the various market participants.
FBP: Now, in terms of the potential and the growth trajectory for digital bonds, we understand so far, $1 billion of such bonds have been issued. How are you looking at the market for the coming years? Obviously, this being one of the assets, and there will be more as we discussed?
Growth potential and trajectory of digital assets will be Asia centric
RA: 2020 has been a very positive year for bond issuance. Earlier I'd mentioned, the total bond issuance in 2020, was around $390 billion USD equivalent, that was a 10% increase over 2019. Obviously, year to date, bond issuance has been very strong given a conducive rates environment as well. Our intent is to, to look at the growing Asian bond market and see how do we create an infrastructure through which the issuance can be done quicker, and the asset can be serviced quicker. But if you're asking me specifically about digital bonds, it's interesting to talk about their history over the last two and a half years. The first bond issued on a blockchain was actually in 2018. And the issuer, that was World Bank. 2019, you really started looking at some of the European banks are doing the POC. Santander and Societe Generale are two names that come to mind. And then in 2020, you start seeing Asia, a lot of activity in Asia on pure digital bonds. And some examples of this are Philippines and Thailand. But really the use cases there were around retail distribution. And then our platform went live towards the end of last year in September. This was the first example of a proper syndicated deal being done. If you think about the numbers, this is a SGD500 million deal. 2020, certainly is a very Asia centric year for digital bonds. But I would say that each market has had a very different twist. Our twist is that we are focused on the wholesale market, we are focused on institutional investors, but other markets look at this as a retail play. So how do I see the growth going forward? I think the growth will still be driven by Asia. There's a lot of activity going on both in the smart contracts and tokenisation and ledger deployment. Also, I see Asia playing a greater role in cross-border connectivity. I can imagine links between Asia and the international depositories, both based in Europe, in the US. That's something I see a definitive move towards. I won't comment on the growth in digital bonds. But can I see a world where more bonds are issued and serviced digitally? With cross-border connectivity? That's certainly a theme I expect to see in 2021. One of the other pieces to think about is how does the regulation support digital securities? The way we traditionally issue securities is a bond certificate, which is a very heavy document is usually given to us at the end of an issuance process, and you're supposed to store it and lock it away. It's really the new paradigm that all of the information contained within that documentation stored on an electronic ledger that we operate. Some of the regulatory changes are likely to be catalysts for a move towards more bonds in the $400 billion of issuance eing digital. I think that'll play a big role.
FBP: To that point, you need a new kind of regulatory framework to look at the documentation, also legal framework for it. How is that progressing? Is there discussion with the MAS (Monetary Authority of Singapore) on the Securities Act?
Introduction of new regulatory and legal frameworks will be catalyst for growth
RH: We've always taken a very collaborative and pragmatic approach to regulation. The pilots that we ran in 2020, we're really there to prove that the technology worked, and for the regulation had to come in to support it. One thing we've done in the beginning of the year, in January, there was a thought piece in the Business Times about what would need to change from a Singapore law standpoint, to support digital bonds, or following that thought piece, we have partnered up in a working group, with three law firms and Singapore Law Academy, to put there a thought piece that we intend to present to the regulators and other industry bodies such as ACRA (Accounting and Corporate Regulatory Authority), about what would be required to have a full end-to-end digital issuance. So that's something we're working hard on during Q1 (first quarter). But certainly, we are guided by the fact that we should be the first market, Singapore really should be the first market in Asia, to have a full-fledged digital bond framework that allows high frequency high grade issuers to come in and use our infrastructure to fundraise.
FBP: Your expectation of the timeframe? Within 2021, third quarter?
RH: You're putting me on the spot, Boon Ping. Certainly, I would hope that this is something we manage to achieve this year. It's not something unknown at a global precedent level. Luxembourg, for example, has a electronic securities regime, Germany has an electronic securities regime, Austria has a government bond regime that is electronic. These things are not unknown. I think they have to be customised to the Asia bond market for sure. But there is existing precedent globally.
FBP: Looking at the digital workflow, using digital ledgers. In terms of some of the upstream activities, there are increased efficiencies. But in terms of how does that accrue to issuers and investors in terms of lowering the cost of issuance in terms of the financial incentive to go the digital route, apart from just pure process efficiency?
Efficiency gains, including faster access to proceeds, better price discovery and ease of use of infrastructure are key incentives for issuers
RA: If I go back to my famous 650 plus step stats, you may be interested to know why 650 individual steps, you actually realise it is because everyone doesn't have the same information source. Think about who is involved in the bond issuance? Sure, you can have anywhere up to 10 to 15, syndicate banks involved in a deal. You can have multiple legal counsels, you can have a paying agent and a trustee, you have the depository. And really what we found out is that the reason you have 650 steps is because it's a duplication of sending the same information to the lawyer or to the paying agent, or to the depository, or to all of the investors. Fundamentally, the approach here is that if you think about ledgers, the thinking over there is that it's a common data source. And how do you get everyone involved in that process access to the same data source at that point in time. Zooming out from an issuer standpoint, out of the 650 steps, it's not just for the issuers, it is for the arrangers, for the investors. We've really had to look at how does a common platform bind all of them? I would actually argue that maybe the answer is not a financial incentive. But it's an efficiency incentive. Also, I think, whatever infrastructure is put together has to work as a utility for the market. Now, when I look at the fixed income vertical, it really looks like a lot of microservices added on that work in conjunction with each other. But the thought here is to have a common infrastructure for the market, that the market has a say in the development of. And from there on, we actually think they themselves would be incentivised to use it, because they've been involved in the design. They've been involved in the product development cycle. Typically, when you look at such infrastructure in the market today, the development cycles may be driven outside of Asia, in EU, UK or the US. You really have to wonder do Asian issuers, do Asian arrangers, Asian investors, really have a say in building that.
PA: The last point I would make here is that keep in mind that through Marketnode, we've built essentially post trade infrastructure. But the intent here, which we've already shown through a partnership with Covalent, is to create truly comprehensive end-to-end infrastructure, starting with the issuance process. The improvement in the issuance process that we are likely to see is also likely to result in better price discovery. And therefore, clear benefits coming through to the issuers. Both in terms of access and price discovery, there will be benefits. But as Rehan has said, obviously, the primary benefits are very much bigger efficiency gains for everyone in the chain, which will also make the process of coming to market much easier, access to capital much easier.
FBP: With the Olam digital bond, Olam being the first issuer and HSBC the arranger bank, getting more issuers more arranger banks, the entire ecosystem on the new platform. How are you doing that? And what are the issues and challenges in getting them on the platform? From a technology standpoint, from an investment, infrastructure standpoint, from the bank side?
PA: This venture has come out of a lot of conversations with clients, and issuers. And I think, firstly, therefore, the focus on fixed income has been really driven by the fact that we've heard from the banks that this is a segment, which, if we're able to deliver the solution, they will see a lot of benefit. I think the key challenge we anticipate here is, really, how do we get everyone to participate on a common platform end-to-end. If we simply just create the depository, and the asset servicing capabilities, then it would not be so attractive, even though. While the infrastructure clearly has demand from our conversations on the testing. And we've been able to actually quantify and demonstrate the value that it can generate. The key thing really is the front end, and making sure those common workflows are truly digitised end-to-end, I think that's really the primary challenge. I think that's more a business challenge, as opposed to a technology challenge. And Rehan, you can share a little bit more detail on just the quantification of the benefits that came through as a result of the work that we've done over the last one year, particularly with issuers resulting, in fact, in some of the deals actually being upsized through the process.
RA:If you think about what is the major draw for the issuer? Obviously, they want ease of usage of infrastructure, but also, after an issuance has gone through, how long does it take for them to receive their money, their proceeds? Today, the typical Asia new issue market is T + 5, so from the date that your deal prices to the date that you receive your money and the bond is issued is five days. So obviously, as the maturity gets shorter, so five days is very different looking at a 10-year bond versus a 30 day certificate of deposit. One draw is most certainly the reduction in time to receive proceeds. it allows for more flexible asset liability management for the issuer. Absolutely. For the issuer as well throughout the lifecycle. We did an interesting study on how long it takes to process a single coupon payment for an issuer. And if you think about the complexity involved there, the depository has to speak to the paying agent. The paying agent has to speak to the issuer. There's a back and forth, usually a lot of reconciliation involved. For everyone in the ecosystem, it's around 220 minutes to process a single coupon payment. Most bonds have semiannual payments in a year. That's 450 plus minutes that are spent servicing one bond. Imagine multiplying that by 800 issuances in Asia per annum. The numbers really add up. In terms of the savings, I really see them as more efficiency gains. A lot of these activities are done by current market intermediaries like paying agents, for example, when we want to get a more scalable solution, they should be able to service more bonds on a daily basis. So that, again, is the intent and that from a quantification standpoint, smart contract enabled platform can bring these 220 minutes down to a few minutes. That's the scale of improvement that you're likely to see.
FBP: Getting more participants into the ecosystem, issuers. What can we expect in the coming months?
Marketnode partnership and its strategic stake in Covalent are key to building an end-to-end digital asset infrastructure and ecosystem
RH: It's quite important to have the right upstream partner as it was mentioned by Prady, because in a way, the more invisible the backend technology is, the better the adoption is. What people want to see is that they're getting the data on their front ends at the right time, the documents are being generated, deal identifiers are there at the time, whether a ledger is doing it in the background, or it continues to be a manual process is something that can't really be seen by the users. In terms of adoption, the way to build that is to really embed yourself in the workflow. Our focus of the pilot has been on post trade and asset servicing. We don't build primary issuance workflows. So from the start, our approach really has been to work with a platform partner, a neutral third party fintech, whose sole focus is primary issuances to where you can give the end users two options, you can say, do you want to continue doing a paper bond like you do today? This is the workflow. Would you like to do a digital security that is, for example, as a paperless security, that doesn't require that it's fully automated from a servicing standpoint? That is also an option on the platform. For adoption, you really have to try to embed yourself in the workflow. Thereafter, the adoption comes naturally. Some of the other digital asset infrastructures I've seen, actually try to build the entire vertical by themselves. They would say I do the front-office, I do the middle office and back office, I do the custodian. We recognise what our competency is, and we recognise the need for partnership where it's not our specialty. That's led to this approach where we partner with platforms like Covalent. So for example, they have 200 plus clients and arranger banks on their platform. They have 1,000 plus existing users, and that's already part of their workflow. To be able to add on the choice of additional security on top of that is a very different proposition than telling someone, here's a new piece of software, please use this for additional securities going forward.
FBP: Tell us more of the joint venture between Blockchain@Temasek and SGX as an infrastructure provider, in terms of financial terms. Do you generate revenue and how is that all taken care of between SGX and Temasek?
PA: What we publicly disclosed is, it is majority owned by Singapore Exchange. And Temasek has a significant minority. We've really seen this as a commercial business joint venture, Boon Ping, in that we didn't want to first build the infrastructure and then hope people will come and use it. We spent actually more than a year creating a prototype, testing it in the market, testing the commercial feasibility, and then decided to convert it into a permanent joint venture. We used to use the word virtual joint venture earlier, and now it is an official and commercial setup. The intent is that there are certain things which Singapore Exchange only brings by way of core functionalities and making that infrastructure and access available. A good example of that is Euroclear connectivity, which Rehan and the team are able to bring as part of the product group - we leverage off that. But this has pretty much been set up as a joint venture, in that the management team members are either being dedicated from our institutions, or will be hired from outside fresh to run the venture. It pretty much will stand on its own feet. It has its own board. It has its governance. And as any other joint ventures would operate. So that's really how we've approached it. We've also made an estimate of what the funding requirements are. And we've had contributions coming in from both our institutions in order to support the growth of the business. From a business development perspective, the combined reach of Singapore Exchange and Temasek has been already very helpful in generating interest from issuers and market participants. And our focus here is the build a solution for the broader markets and to drive adoption collectively. That is generally how we are approaching it. Very engaged operational effort from both sides. Rehan, do you want to add anything from your side?
RA: You've articulated it well, I would just say that given that we've worked together over the last 18 months or so and the virtual JV as Prady calls it. It really gave us a chance to evolve the working relationship quite well as to how the biz dev would work, how we would be able to leverage on Temasek’s blockchain ecosystem community. The model had already been running for a while, the formalisation of the joint venture, obviously formalises and announces it to the public. But I'd say the operating rhythm has been there for the last 18 months or so.
FBP: Tell us more of Marketnode’s stake in Covalent, the technology partner that puts together a lot of the technology behind the platform.
RA: I should clarify that Covalent is the upstream bond issuance platform. If you think about bonds from issuance to settlement, the core hash infrastructure that we have that the joint venture in this virtual form had built together, continues to be operated by us. The upstream workflow technology I was speaking earlier about, which is how an issuer comes to market, how do they do the workflow with the arranger and the investor, that piece sits within Covalent. Obviously, for us, the idea is to marry the two. We continue to focus on our technology, which is post trade, listing, security creation, linkages into ICSDs (international central securities depositaries). Covalent focuses on its strengths, which is upstream workflow.
PA: The clarification is important. Marketnode has been built as open infrastructure. And what I mean by that is, we're really looking to enable a broad set of market participants. Covalent has already invested a few years in building that front end issuance capability. And therefore, we look to partner with them, especially given the capabilities in the region and the Asian time zone. Which makes it also easier to provide that capability to customers, and to have those development and adjustments that we need to make along the way.
FBP: Could you comment on the other asset classes that you're working on in terms of other specific funds, especially the sustainable finance assets.
PA: The applicability is very broad based. And the approach we've always taken is, to say, let's first prove out the value in one area, and therefore the single-minded focus on corporate bonds, in particular, in the fixed income space. Now, having gotten the first asset class off the ground, and having built on the core infrastructure, the partnership with Covalent to begin that engagement on end-to-end basis, there are two, three areas that we are now beginning to develop, like I said, shorter term, bond issuances, or shorter term paper. We've already seen interest from the various market participants in the fund space. So that is another area that we're looking to explore. Sustainable finance is something which is fast growing, and a real need to grow as an entirely new asset class in of itself. And therefore, we've had some early conversations to see how some of the market core infrastructure can be made available. But it's at a nascency as an industry in and of itself. So really, what we're doing is having the conversation to see how we can make the technology available, but will obviously be dependent on the development of the front end in of itself as an industry.
RA: On the fund side, we are in deep discussions with the industry. Funds are an existing asset class, we are not trying to build some new like sustainable finance. We're trying to see what are the pain points in there. Very much like what Prady mentioned, have a singular focus, get a certain workflow right, solve a certain need, and then scale that. There are some are very particular pain points in the fund industry, especially on the underlying infrastructure side. So that's one of our focuses for 2021.
FBP: We discussed the current infrastructure for existing workflow, smart contracts, the underlying infrastructure. We talked about new asset classes in terms of tokenisation. Is there a timeframe for that? And how are you looking at that aspect of using that technology?
Adoption of digital asset workflow requires existing distribution mechanics to change
PA: The technology will be available at scale in a relatively short period. We cannot predict how fast some of these would come, Boon Ping. Take alternative assets, let's take real estate, let's take infrastructure. A lot of the issues that need to be solved are at the distribution level. So even as we make the technology available, we do need some of them to follow fast and for the distribution mechanics to change. Our hope is that the people who are trying to build the full stack themselves find it prohibitively expensive, and it’s not that feasible to connect to multiple custodians and create that into an end-to-end infrastructure. This takes off some of the development timelines and makes it more feasible. We will follow where the customers are, and we are doing our best to engage with a broad set of parties. And this is similar to the approach we take with Covalent where we showcase certain use cases, and then wait for applications to be pushing in that direction. Some interesting early conversations. But again, it's difficult to put a timeline. Rehan, you may also have a perspective given SGX is involved in a few of the platforms more directly.
Marketnode aims to serve as a master ledger of record
RA: I think the concept of a master ledger is very relevant over here. Certainly, we see a proliferation of digital exchanges, each with a different flavour. Some of them are focusing on early stage investments, some of them are focusing on late stage, some even focused on bonds. That is a space that Marketnode is probably not going to compete in, which is trying to build a marketplace for retail and AI (artificial intelligence). But where we do see ourselves playing a role is “Can you connect all of these different venues through having a master ledger of record?” What's being traded in these venues are digital securities. Could it be that the end investors’ trading goals actually need traceability to regulated depository in the future, I can see that. Does that mean that the depository has to have a say in the tokenisation connectivity standards, I can see that. That's certainly a role that we play today. So, again, back to Prady's point, we are likely to follow our clients. If you think of us as a master ledger, we can be seen as the servicing agent for any digital infrastructure marketplace that comes up in the future. In bonds, the use case is very specific, where we're actually partnering with one to get the connectivity. And start looking at others, for example, that focus on short dated securities in the future. I expect the proliferation to continue. But I also expect that at some point, these will grow beyond a certain point when there is regulatory backing, some kind of traceability and element of trust for the underlying asset. I think that'll help catalyse.
FBP: Thank you. Rehan and Prady for your insights and comments and to bring us a closer understanding of the JV that you have and how you're developing the infrastructure for financial markets. Thank you.
PA, RA: Thank you.