Transcription
[music] >> Hello. Welcome to Coffee Time, a podcast series on markets and economies from DBS Group Research. I'm Taimur Baig, Chief Economist, welcoming you to our 175th episode. This is a special one. I'm delighted to welcome back Piyush Gupta, CEO and Director of DBS Group to this podcast. It has been about a year since Piyush Gupta took over the top job at the bank, and what a year it has been.
But it is not just about the last 1 year or the year that's about to befall us. It's about the profound changes taking place in markets and economies, in banking and tech. So, we will dive into the both short-term and long-term in this podcast. Piyush Gupta, welcome back to Coffee Time.
Thank you, Taimur. So great to be back. >> It was 100 episodes ago. No, 100 exactly 4 years ago. >> Time has flown. Time has flown, but the world of wars is a constant state. >> 4 years ago, you and I talked about the war in Ukraine. Oh my god. Its impact on inflation, and today we're talking about the war in Iran, its impact on inflation. Deja vu.
Unfortunately, yes. How are you guiding the bank through this sudden episode of burst in inflation fear and associated issues? So, we've really had a few stress tests in the last 5-6 years, starting with COVID on supply chains and and liquidity. Uh and then the Ukraine war for LNG and energy and power and now Iran. So, we have a playbook. And we take the playbook out and we execute and we stress test according to the playbook. First order impact, second order impact, third order impact. What, who, how, why. And that's what we're doing.
There's some shades of similarity with Ukraine. Uh and that is inflation on energy prices, inflation on the whole energy construct, including sulfur, helium, all the gases. Uh and then the follow-down the the follow-through impact on food and supply chain around food, right? So, food, power, energy, shipping, logistics, warehousing. So, that's, you know, the the the industry by industry, sector by sector, and then segment by segment kind of first-level stress test. Second level would be the fallout about about around slower consumption growth, consumer confidence, business confidence. Um that's our second-order impact, which I think will be harder and more severe for, you know, more people. Um and we just have to hunker down.
In the first-order side of the effects, you're not that particularly worried about the bank's exposure because we're not exactly that exposed to the Middle Eastern economies. >> Yeah, I think our exposures right now to the Middle East fairly limited. It's limited to sovereigns and and really high-grade, high-investment-grade credits and fairly short-term. Um so, we're not too So, the first-order impact isn't too big. It's quite limited. The second and and and and of course, as I said, credits are very good. The second-order impact I do worry about. I worry about supply chains and inflation and the cost of supply chains. The good news is because of COVID and because of the last and because of of of Ukraine, everyone's diversified their supply chain. So, everyone has been intentional about rearranging their supply chains, their suppliers, backup plans, inventory. So, everyone's gone from just-in-time to just-in-case. Cost has already gone up the last few years for for people managing for diversification of risk. But this was a real wake-up call around energy. I think this is it, right, Taimur? Don't you see that?
>> Absolutely, Piyush Gupta. Two things. One is that I think this underscores the importance of green energy. Green energy, renewable. And we've seen that in the Chinese example. >> China, right? Yeah. >> People are not panicking about China's fragility, although they import a lot from the Middle East, is because they have built such an amazing green ecosystem.
>> But at the same time, they didn't give up on their coal infrastructure. So, in terms of emergency use, they can tap into both fossil fuel and green. Yes. And therefore, when we see around the world fears about uh spending around the pandemic about the crisis pushing up bond yields, in the case of China, the bond yields are not going up. >> They're quite muted, aren't they? >> Exactly. Yeah, it's all about the availability, diversity, affordability, and diversity well of of of energy sources, right?
>> That's right. Yeah. Uh and Piyush Gupta, the second issue in terms of the lessons from this thing is that, of course, as you said, the cost of just-in-case has gone up. No more just-in-time. But perhaps another layer of cost will be associated now. Uh you and I were talking about this before the recording that even for a country like Australia, they might have to rethink about the refining capacity because they do rely a lot on countries like Singapore for the refined products, despite being a big energy producer. Yeah, it's not just about the source of energy. It's about refining it, the quality. And then also the infrastructure and the transport, right? And and and countries are now thinking, "I must get my infrastructure sorted out. I must get my diversity of sources." And also the refining and the more downstream impact. It's so important.
>> Absolutely. I'll just share with the audience that one of this piece of analysis that we have been doing is that at $110, it is not the end of the world for Asia. Although the newspaper headlines are all about, you know, it's going to be hitting Asia very hard, we have taken this shot before. Yes. >> Back in 2009, 2022. I think Asia's resilience is understated. Uh we probably are better off sort of handling this now than we could have 5 years ago. Yeah. I would worry at 150. So, hopefully, fingers crossed, we don't go in that direction. >> Hopefully, we don't go there. Right? Yeah.
Piyush Gupta, there was an acronym that you used 100 episodes ago. Uh OODA. Observe, orient, decide, act. Is it like your mantra even now, 5 years later? Absolutely. So, just and and also don't be scared to pivot, right? The thing about today and even more so than 5 years ago is that the change is so rapid. You've got the geopolitics. You've got the market volatility. But you also have an underlying fundamental shift in technology with the world of AI, generative any generative AI now upon us. I think 2026 is the year of reckoning. It's the year that enterprise, businesses, SMEs, governments, households, families, students realize, "Oh my god, things have changed." And if this is a year of reckoning, and with each drop, things just get so much better, faster. It's recursive. It's self-improving. Then we have to really be nimble, agile, and be able be able to switch at a dime. That's not easy for enterprises trying to plan for the long term. So, you have to plan for the long term. You got to plan for contingencies, exigencies exigencies, and, you know, diversity of supply chains. And then you got to plan for this big kahuna of, you know, structural change in the way business is done. The good news is I think this tech tonic shift in technology is actually helpful. It should ultimately bring costs down. It should give you better informed information and help you make better judgment calls. So, that's actually not a bad thing. It's good.
>> Right. But it will shift the way work is done. It will shift jobs. It will shift job scopes. And we're going to have to learn stuff real quick. I will would like to go a little deeper on both the technology and the job thing a little later, but since you mentioned the moment of reckoning, a year ago, when you took over the big job at DBS, we had almost immediately a big moment of reckoning with the reciprocal tariffs. >> I know, right? April 2nd, Liberation Day. Absolutely. So, it's been a year. It's been a year. Uh DBS clients and DBS itself, how has the experience been the last 1 year dealing with these tweets and tariffs? I think um people are getting used to this, you know, big shifts. That's it's good and bad. I mean, it's, you know, markets now absorb these these shifts quite easily. Um our clients have grown to like the fact that we're dependable. I like to talk about the four Ds. DBS is the Development Bank of Singapore. That's the old name. But really, we want DBS to be known as a dependable bank, the digital bank, the diversifier bank, and a disruptor bank. And that's come into play in the last year in a big way.
Um you know, what struck me was the fact that all the more when things are tough, people want a safe, dependable lighthouse in the sea of volatility. We want to be that lighthouse. We want to be the safe. We are the safest bank in Asia. We continue to be rated one of the safest banks in the world. And that does count for something, of course, right? Um we want to be that bank that helps you diversify in the world that's getting more and more bifurcated. You're going to have to not just diversify your supply chains. You're going to have to diversify your banks, your payment rails, your counterparties. And as the world gets more and more bifurcated, then in the past, you would have one big global bank and one big network bank. But now you'll be thinking, "You know what? Asia for Asia, Europe for Europe, US for US. And, you know, the Middle East and and and and rest of the world is will also all be connected in some way, but you do want to spread your risk a little bit. Does that make sense? So, diversify your kind of, you know, counterpart. So, you remember when we were gaming the scenarios in the really uncertain times a year ago, we came up with an acronym TOTUS. Trade outside the United States. That was great. I used it a lot. You were spot-on on TOTUS and I've used it. And look at the last one, you Sushan. We cannot divorce ourselves from the US. It's too large and consequential. But for the non-US world, there's a lot of prosperity to be had with >> Trade outside the US has indeed grown, especially intra-regional trade here in Asia, north-to-south Asia, uh the global south, right? Middle East to Asia. That's been really great. Yeah, that's been really great.
Now, we're having this conversation in the first day of April. Most of March you were in various planes around the world. You went to the US West Coast, you went Beijing, you went Hong Kong. Tell us a little bit of a takeaway from visiting the key consequential areas of the world. So, you you know, a year ago, it was data centers and and and brag awards, right? That's what um my friend who runs Equinix calls it, where there really was a lot of big ambitions, you know, all the LLMs were fighting each other to see who could build more data center, more compute power. Um but the world is energy constrained. This year it's all about how many tokens are you burning. So, it's now about, you know, bragging about token usage. And as the cost of tokens go down, it will the use of tokens will go up exponentially. It already has in every market. So, what struck me in the West Coast was number one, the speed of adoption. When Open Claw dropped, suddenly the way all these, you know, AI companies had to pivot, they pivoted on a dime. Right. The LLM companies as well, all sort of pivoting to do to do more enterprise work, to get more domain knowledge. At the end of the day, I think everyone realizes it's all very well to have an agentic AI world, but what matters is the domain knowledge and the data. Because if you have the right domain knowledge, if you have the right data, and you couple that with great talent, and knowing what questions to ask, and also, of course, guardrails and safety and security, bang, you've got a moat. And everyone's rushing to create that moat quick and fast. So, I was just stunned by the speed. Even when we were there, things were dropping.
>> [clears throat] >> I was also interested to see the different strategies adopted Mhm. by different big tech. Some have chosen to go full stack, Google, for example, with their TPUs, Google Cloud, their multimodal, they have LLMs, it's, you know, And both consumer and enterprise. >> Consumer, enterprise, a full stack kind of, you know, steady, large, but pretty darn powerful. Then you've got Anthropic and and OpenAI, both incredibly innovative, fast-moving companies, great talent, um both looking to do a lot more in enterprise. Of course, OpenAI is also very focused on the consumer. And then lots of startups just, you know, coming in and doing different bits of the ecosystem from connecting with browsers to keeping you safe, creating control planes for agents, creating observability, auditability of agents, etc. So, so many use cases. And I was struck at how young the talent was.
>> Mhm. You know, everyone in their 20s, maybe maximum early 30s. Uh felt like a bit of a relic, to be honest, Taylor, but it gave me energy. I have to say I was struck by the talent, the speed, the momentum, and the creativity and innovation of it all. Well, and after the West Coast trip, you didn't even come back to Singapore. You went straight to China. >> Straight to China for the Beijing China Development Forum, where I was just struck by the clarity, predictability, stability of the narrative. It was consistent, it was clear, it was predictable. Li Qiang started, the premier started with this is the GDP plan. We're at 140 trillion, we want to go to 170 trillion. So, four four and a half, 5% compound average growth rate for GDP. And here are the different focuses. Still very focused on renewable, green, still very focused on AI, but more enterprise use AI. They've chosen to go open source and open models, and they are very much about usage, about efficiency, and of course, hardware and software. So, not just software, but hardware. Dark factories, robotics, but also health care, biotech, use of AI to be more instructive in the in in in in in health care. And they don't seem to be as obsessed with this race toward AGI like some of the tech titans in the US seem to be putting so much resources into. Yeah. I you know, but if you look at, you know, Deep Seek moments and now ByteDance and and and Ali, they've they've also come up with great offerings.
>> Absolutely. And um I think they are more focused on use cases for everyday life, for for business. I kind of prefer that in some ways, uh because it seems to me that it is a very powerful technology, and I think we should deploy it for the most naughty problems that we have, as opposed to the trying to force ourselves to come up with solutions that probably are not that urgent. I think social media, the last thing social media needs is more AI. [laughter] Whereas, I think on health care, on agriculture, or on meteorological prediction, I think AI can be far more value-added. I agree. >> Admire what the the Chinese are sort of pushing. >> Yeah. Um and I went to visit the Xiaomi factory, and I was just struck at how quick the car could be assembled by robots, like literally in minutes, you know? I was just struck at how clean, how fast, how progressive, how beautiful the end product looked. >> These are really nice cars. The SU7 that they're I mean, they're breaking all the records in the world, yes, indeed.
Um we have, of course, a much bigger presence in Hong Kong than we have in mainland China. Every time in the last 5 years I've been to Hong Kong, I've seen a progressive improvement in sentiments. 5 years ago, they really seemed to be on their knees. The pandemic recovery hadn't gone well. Property market was under a lot of stress. What's your sense of Hong Kong? Hong Kong is back. Hong Kong is booming, I think, especially in the last few weeks, months, days, and Hong Kong is um is doing great. I and I'm happy to see that. There's been a lot more flow of people, talent, into Hong Kong. They're attracted by the low taxes, they're attracted by the capital markets, especially equity capital markets. And talent begets talent, right? And the flows have been really good. Wealth flows, corporate flows, capital flows been really good. So, Hong Kong is back. And Hong Kong MA's been very vocal in its support for growth capital, gold, equity, of course, and they've also issued the ordinance around stablecoins, etc. So, I think there is a momentum in Hong Kong. After a long time. After a while, but, you know, never never never underestimate the power and the hunger and the energy of of of the Hong Kong people and Hong Kong market. We always I used to live there, and I I can see that it's it's doing great, and and I'm happy to see that.
What about this notion that the problems in the Middle East is a boon for Singapore, is a boon for Hong Kong. Singapore versus Hong Kong, what do you make of all that? Singapore and Hong Kong have a healthy competition, but we're good at different things. I think Singapore capital markets has also been good. The the equity development fund's been very helpful to kickstart a lot of activity in the SGX, but Singapore's always been known as an FICC market. We're good at it foreign exchange, good on rates, good on commodities, and also very keen on increasing our our role in the world as a hub for gold. It could be physical gold, it could be paper gold, digital gold, tokenized gold, whatever. I think um both markets want to be hubs for commodities, both markets want to be hubs for equity and and debt. And there will be some competitive tension, but I do think Hong Kong's equity capital market's always going to be bigger. It's got the north in Asia flows, got China Stock Connect, Bond Connect. It will always be a big hub for North Asia. And Singapore's complementary. Singapore has found a niche in REITs. We've got, you know, good banking stocks, we've got some tech stocks, we've got some manufacturing stocks. So, it's fairly complementary. And um Hong Kong has all the new economy stocks as well. So, I would say it's a complementary, hopefully symbiotic relationship, but there will both will benefit from wealth flows. Both will benefit from diversifi- diversification and potentially de-dollarization, if that were to happen. I think both markets will benefit from that.
Absolutely. Susan, it seems to me that on the equity market side, of course, Hong Kong has a much deeper, longer pedigree. But, because of SGX's concerted A+ effort over the last few years, we're beginning to see uh the needle moving to some extent. The IPO pipeline is interesting. I'm part of a WhatsApp group which has about 150 entrepreneurs. And and I'm beginning to see a higher frequency of people closing deals, series A, series B, which was really far and few between even 2 years ago. >> Yes. So, on the margin, there's some glimmer of hope. Even the area where Singapore has been lagging, there might be some finally some momentum. Oh, yes, for sure. I mean, you see also a lot of talent coming in, whether in the private capital world, private debt, private equity, uh venture, growth capital. That's also growing. And and and and I am quite hopeful that that will continue to be a structural growth arena for us. The financial industry creates what, 11% of GDP in Singapore? It's an important part of um of our ecosystem and our our our economy. And we have been successful in attracting talent to this. I think Singapore's move to create a wealth management hub has been successful. I remember being part of the setup of the private banking industry group years ago, where we came up with the blueprint for talent and standards, right? So, money laundering, talent standards, education. And you know, it's just snowballed in the last 10 years. And I'm really happy to see that, because once you get the talent, you get the AUMs, you get the AUMs, you get more talent, you get more of the whole service providers, whether it's wealth tech, fund managers, distribution. It it all comes together. And I'm glad to see that we've created in Singapore that whole hub of wealth. And and then the numbers speak for themselves.
So, on one hand, we have a good sort of foundation for talent coming into Singapore. Even the locally grown talent is sort of, you know, upping their game. But, at the same time, we have this big concern that AI will eat jobs. Yes. >> And for the 25-year-old, unless they're an entrepreneur, there's nothing for them to do going forward. How do you reconcile this? And as a bank, what is the signal you want to give to the community in Singapore? So, we've come up with a message message house in DBS, which is we're here to save people, not jobs. And AI's going to change jobs. But, it's also going to create capacity. And that capacity should be utilized in a constructive manner. And what matters is attitude. What matters is if the way you used to do your job has changed, it's now easier, faster, better to use AI, then heck, use it. Learn to use it and be on top of it. Make sure it's safe. Make sure it's constructive. And make sure it's not hallucinating. Then, free up that time to have more of these person-to-person interaction. Free up that time to get to know your customers better. Free up that time to know your industries better, to create more domain knowledge. Right? If you think about it, does anyone code anymore? Teemu, what do you think?
>> It's changing very dramatically, yes. Right? So, if you ask any tech company, does anyone code anymore? Some do it just, you know, to work with the agent. But, by and large, coding has changed. Coding is not needed. So, if you don't code, but you're an engineer, and you're in a domain, then this is your time to get to know your domain better, not just be a coder, but or or be a production-level engineer, but be a domain-level expertise. For example, if you're in trading tech or wealth tech, get to know the domain. Go spend time with the with the with the clients, Absolutely. >> with with with with the fund managers, with the traders, etc. So, I think this is a great opportunity for us to not do those mundane jobs and, you know, data pull and research and you still have to read and understand it in your brain. Your brain has to synthesize information. I think that's important. We don't want to lose that human understanding. 100%. I tell my son who just turned 13 that if you don't understand the world, you will not be able to deploy AI in a productive manner. You have to understand. Therefore, it's not a question that AI can do math for me. If you don't understand math, how do you then get the AI to do it rightly? Yes. >> Right? What do you think, Teemu? I think the world will change, but new jobs will be created, right? Let me know. Was it the World Economic Forum that said 97 million jobs will go, but 140 million will be created, something like that?
>> Right. So, Susan, when you were traveling, Leslie Teo, who's a deputy director of Singapore AI, and I, we published an article [clears throat] in Business Times about the future of jobs. And Business Times also asked us to write an article about how we wrote the article, because the various agents that we used to synthesize the research. So, we had five LLMs querying 200 research >> the case, then being thoughtful about young people and jobs, I I worry about the millions of STEM graduates around the world. I mean, you got millions in in some of the highest growth population countries. They're all talented, but they need, you know, they need to be that energy needs to be channeled into constructive, meaningful work. Right.
I want to extend the discussion a bit toward the future of banking, because when we talk about AI-related disruption, there's also been a lot of fintech-related disruption taking place over the past decade. >> [snorts] >> So, cryptos to stablecoins, wealth management apps, tokenization, how will a bank like DBS look like through all these disruptions 5 years from now? Would they become smaller, or would they be able to harness all these things and make themselves keep themselves relevant? So, Teemu, you know, DBS, we had a head start. We were the first bank, I think in the world, to to to get into the whole digital asset ecosystem. We started looking at it thinking, well, if banks want to stay relevant as financial intermediaries, first you need to keep the trust of the customer. And if in you know, people trust us to to safeguard their money. It's their hard-earned money. If that money becomes a digital asset, a token, an STO, a Bitcoin, will they custodize that digital asset with us? And we thought, well, if we created a safe, solid, secure custody [clears throat] service, they'll come to us. So, we created cold wallet, warm wallet, hot hot wallet. You know, we created the whole sort of gambit of of of custody for clients on digital assets. Having done that, we realized as people want to go on and off ramp or they want to trade between fiat and digital assets, they wanted a safe place to go. And so, we created exchange. We work with the regulator. We create an SGX. We created the exchange so that our customers could go on and off from off ramp safely with us. Then we realized, you know what? Going further up the the the the chain, if people want to tokenize their assets, what it's whether it's a building, a money market fund, um or even tokenize their deposits onto the blockchain, onto a private permissioned blockchain, we need to help them. So, we've now done the whole thing. And we are really in the picks and shovels business of this whole digital asset ecosystem. We provide not just custody. We provide reserve management. We provide banking services for for for the digital players, for for fintechs, etc. So, we've gone end to end.
Now, is the world going to move from fiat to stable coins in a hurry? Well, when I looked at the data, there's some 300 and something billion in stable coins, of which 90% is used to facilitate crypto trading. >> Right. Right? There's some on the fringes of you know, not so savory flows, but the tiny percentage is used to settle real genuine commerce trade. I do think there is a use case for stable coins for for countries where your own currency is volatile, maybe depreciating in a hurry, and you need to use a stable coin to buy and sell goods or services. >> Acts as an anchor. Yeah. I think there is a use case. [clears throat] But the problem is regulations have not landed. Countries need to figure out interoperability, security, and money laundering rules. So, that has not landed. Right. So, if everybody starts to issue their own stable coin, and it's not interoperable, then what is the use case for trade? Correct. So far, the use case has been USDC, USDT on and off ramp, or or or trading for Bitcoin or Ethereum. That's been the dominant. So, we've been active, as you know. We've tokenized deposits for clients. We own with some other banks and Temasek, Partior, which is a cross-border blockchain platform. And we've worked with governments, Singapore particularly, also Hong Kong on their various projects around the DLT. We'll continue to do it. We'll continue to learn. But we have to see how regulations evolve in order to make big bigger plans around this.
My entrepreneur friends who are in the non-bank financial sector, they always complain that the large banks will never allow them to be disintermediated, and they will lobby, whether it's the government of Singapore or the government of the US, to make sure that the DBS and JP Morgans of the world remain at the center of financial intermediation. Uh and and my response to that sort of fear or frustration is that there's a reason why banks exist, whether it's in the US or in Singapore, because it is a regulated entity, which provides to your, you know, core message that a trusted hub for the most important thing of modern capitalist society, which is transactions and custody. You can't have it in a wild west situation where there's limited visibility of the regulator, or limited uh enforceability by the regulator. And this is why I don't worry too much about the future of banking. I think it plays a pretty essential role. Yes, because you will have if you're not regulated, there will be dodgy flows. By definition. By definition, because you can be anonymous in in in in in a public blockchain. You have a code. Um although there is transparency, it's it's, you know, it can move very quickly. It's 24/7. Right. And that's why for banks, when we get into this, we have to be very studied, and we have to realize, can you really KYC every wallet, hosted or unhosted? Can you really be, you know, can you do this without, you know, be be in line with the high standards that we hold ourselves to? And that's why we've been cautious. Um we had a head start. We consolidated our position. We want to We think there is scope to to do stuff in the middle, being this picks and shovels ecosystem player. And we'll continue to do that. But we do think that tokenized deposits could be the way forward versus maybe stable coins. Because there is a use case for tokenized deposits. There is a use case for 24 by 5, soon 24 by 7 uh settlements, atomic settlements, real-time settlements, smart contracts. Yes, there is a use case. And we want to be there. But we're taking the route of tokenized deposits for now.
Great. I've saved the hardest question for the last. >> Ooh, what is that? >> Um because going back to your trip through the US and China, both countries are leading the world in both hardware and software solutions around cybersecurity to 5G, 6G, to uh you know, you know, customer risk management, to everything. If the Chinese offer you a much cheaper and as good a technology as a Western firm does, do you still have to say no to them because you have run the danger of being bank based in Singapore, which is also under the radar of the US, and they don't want us to use Chinese technology? How do you balance that? So, I I do think the world is becoming more bifurcated. And you'll have different payment rails for different currencies, potentially different rules for different spheres. So, I like the idea of layering. I like the idea of staying neutral and open for business, being rules based, but keeping your domain and your data safe, staying neutral, but doing business on all sides as long as everyone is rules based. Keep to your regulations. Keep to your standards, but learning from both sides. If you can get the best of both worlds, that is a great outcome. And I want to take that optimistic outcome. I love that. That's a great note to end on. Su-Shan, thanks so much for coming to our 175th episode. Thank you, Taymour. Thanks to our listeners, too. Coffee Time was produced by Ken Delbridge at Spice Studios. Daisy Sharma and Violet Lee provided additional assistance. This podcast is for information only and does not constitute any trade advice. All 175 episodes of Coffee Time are available on YouTube, as well as on all major platforms of podcasts, including Apple, Google, and Spotify. As for our research publications and webinars, you can find them all by Googling DBS Research Library. >> [music] >> Have a great day.