WEBVTT - Anthropic Draws Investor Offers at Over $800 Billion Value 0:00:02.520 --> 0:00:13.480 Bloomberg Audio Studios, podcasts, radio news. Bloomberg Tech is live 0:00:13.520 --> 0:00:17.319 from coast to coast with Carolline Hide in New York 0:00:17.600 --> 0:00:19.600 and Eva Low in San Francisco. 0:00:22.680 --> 0:00:26.440 This is Bloomberg Tech coming up and Thropic rebuffs investors 0:00:26.480 --> 0:00:29.600 pushing offers on funding that would value it at more 0:00:29.640 --> 0:00:31.760 than eight hundred billion dollars plus. 0:00:31.840 --> 0:00:35.080 Metro expands a multi billion dollar partnership with Broadcom to 0:00:35.159 --> 0:00:36.480 design and build custom. 0:00:36.280 --> 0:00:40.159 Chips and ASML the supply of cutting edge chip making 0:00:40.240 --> 0:00:44.440 machines raises, it's for your sales forecast. Is AI demand 0:00:44.640 --> 0:00:45.720 fuels growth. 0:00:45.760 --> 0:00:48.640 And that demand is fueling growth. When it comes to 0:00:48.680 --> 0:00:52.680 the benchmarks today, ed despite the ongoing anxiety about conflict 0:00:52.720 --> 0:00:55.480 in the Middle East, but hope optimism that some sort 0:00:55.520 --> 0:00:58.040 of extension to the ceasefire might be found, we are 0:00:58.320 --> 0:01:00.680 at or near record highs. The S and P five 0:01:00.760 --> 0:01:02.840 hundred is at a new record high. We hold onto 0:01:02.880 --> 0:01:04.839 these games towards the closed we're up three tens of percent. 0:01:05.000 --> 0:01:07.000 Then aw's that one hundred ever so closed to a 0:01:07.080 --> 0:01:09.480 record high that it had last year, up six tens 0:01:09.520 --> 0:01:12.880 of percent. But look at the length of games that 0:01:12.920 --> 0:01:14.600 we have had then, as that one hundred is now 0:01:14.680 --> 0:01:17.880 up for eleven straight days. It is increased by thirteen 0:01:17.880 --> 0:01:19.919 percent over that stretch. So long as win in streets 0:01:19.920 --> 0:01:20.920 is twenty nineteen. 0:01:20.720 --> 0:01:25.039 N single movers, this is what's driving the market right now. 0:01:25.319 --> 0:01:28.280 Look at Broadcom up almost four percent, Metas actually also 0:01:28.319 --> 0:01:30.040 pushing high. I think that's probably more in line with 0:01:30.080 --> 0:01:33.120 where the market's at right now. This is an expansion 0:01:33.160 --> 0:01:37.240 of an existing agreement. Broadcom works with Meta for their 0:01:37.280 --> 0:01:39.959 custom silicon, and later in the block we're going to 0:01:39.959 --> 0:01:43.280 talk to Mandate saying who leads our research and Bloomberg Intelligence. 0:01:43.400 --> 0:01:44.440 It was going to take a look at some of 0:01:44.440 --> 0:01:46.720 the pictures from when I visited Meta and get into 0:01:46.760 --> 0:01:50.680 detail about what those chips actually are. In private markets 0:01:50.720 --> 0:01:53.440 is where the big story is right now, and Tropic 0:01:53.960 --> 0:01:58.280 is drawing investor interests at evaluation of eight hundred billion 0:01:58.360 --> 0:02:01.960 dollars or higher. That would be a massive jump from 0:02:01.960 --> 0:02:04.480 its three hundred and fifty billion dollar price tag. Just 0:02:04.520 --> 0:02:07.200 in February, bluemokes that Tasha Mascarene is part of the 0:02:07.240 --> 0:02:08.960 team that broke this story. 0:02:09.200 --> 0:02:10.240 Let's be super clear. 0:02:10.760 --> 0:02:15.600 This is investors going to Anthropic and saying, you guys 0:02:15.600 --> 0:02:18.280 should raise money at this valuation, we want to do this, 0:02:18.720 --> 0:02:21.360 and Entropic saying we do not want to do this 0:02:21.560 --> 0:02:22.280 at this time. 0:02:22.560 --> 0:02:25.600 Yes, you're absolutely right. I mean, we've both been covering 0:02:25.639 --> 0:02:28.280 the AI boom long enough to know that the investor 0:02:28.320 --> 0:02:31.960 interest can soon turn into an actual deal. But in 0:02:32.000 --> 0:02:34.280 the case of Anthropic, I mean, they close their employee 0:02:34.320 --> 0:02:38.400 tender at around half of this new valuation just earlier 0:02:38.440 --> 0:02:41.280 this month, and so right now we're seeing Nthropic push 0:02:41.400 --> 0:02:44.000 back on those prices. But these are serious term sheets 0:02:44.080 --> 0:02:46.920 that we're hearing or even encroaching higher than eight hundred. 0:02:47.040 --> 0:02:48.840 And investors think the evaluation. 0:02:48.440 --> 0:02:51.560 On the valuation, Yes, and investors think that that's a 0:02:51.560 --> 0:02:52.000 good deal. 0:02:52.160 --> 0:02:55.560 Right now, Natasha, what sort of amount would they want 0:02:55.600 --> 0:02:58.120 to raise? Could they need to raise? Because as much 0:02:58.160 --> 0:03:01.280 as they steal all the headlines, if ideas that Opus 0:03:01.280 --> 0:03:03.960 four point seven is coming soon according to reporting, or 0:03:03.960 --> 0:03:07.440 indeed that we see Mythos being very staged rolled out, 0:03:07.560 --> 0:03:09.480 there is this concern that maybe it's stage because there 0:03:09.520 --> 0:03:11.480 isn't the access to compute and that costs money. 0:03:12.760 --> 0:03:15.360 Absolutely, I mean, I think at this point we don't 0:03:15.360 --> 0:03:18.440 know the exact numbers of how large the raise could be, 0:03:18.639 --> 0:03:20.880 but we do know that there is this time crunch 0:03:20.919 --> 0:03:24.000 that everyone is thinking about when it comes to Anthropic 0:03:24.480 --> 0:03:27.040 the IPO. As we've reported, this is a company that 0:03:27.360 --> 0:03:29.880 has already been a high profile startup for the past 0:03:29.880 --> 0:03:33.239 two years. But now there's this really real ticking time 0:03:33.280 --> 0:03:36.400 crunch for an IPO as soon as October, as we've reported, 0:03:36.760 --> 0:03:39.440 and so I believe that we're going to see new 0:03:39.480 --> 0:03:43.320 capital talks continue pretty relentlessly until then. It's really just 0:03:43.360 --> 0:03:45.760 a decision of if Anthropic needs it thinks it needs 0:03:45.800 --> 0:03:48.800 to secure that capital before it makes the jump into 0:03:48.840 --> 0:03:52.040 the public markets. To your point, the new model gives 0:03:52.080 --> 0:03:55.600 a real exact pinpoint reason to where where the capital 0:03:55.600 --> 0:03:56.560 would go at this moment. 0:03:56.720 --> 0:03:59.320 Brilliant reporting has always been braised in attach and mascotiness. 0:03:59.320 --> 0:04:01.880 Thanks so much, winning us on all things Anthropic. Let's 0:04:01.880 --> 0:04:04.000 go from private back to the public markets. So you 0:04:04.040 --> 0:04:06.280 see Meta up almost two percent, broad coom is up 0:04:06.360 --> 0:04:09.200 three point six percent. That's as they expanded that multi 0:04:09.200 --> 0:04:12.720 billion dollar partnership with the chip maker to design and 0:04:12.760 --> 0:04:16.080 build custom chips to power social media giants AI ambitions. 0:04:16.080 --> 0:04:18.919 Bloombag Intelligence has said that the move along when Meta's 0:04:19.000 --> 0:04:22.559 new news spark model shows stronger confidence and in house 0:04:22.640 --> 0:04:25.320 chips for inference. Though Meta could face more pressure on 0:04:25.360 --> 0:04:28.680 spending returns without a cloud business. Many seeing BlueBag intelligence 0:04:28.839 --> 0:04:31.080 helps pen that note joins us right now and just 0:04:31.120 --> 0:04:33.479 talk to us a little bit man deep about well, 0:04:33.839 --> 0:04:36.840 the desire to custom chips, what sort of edge does 0:04:36.839 --> 0:04:39.479 it give it when it's already dependent on AMD and 0:04:39.640 --> 0:04:40.160 video two. 0:04:41.640 --> 0:04:44.279 I mean, look, the proof is what Google has done 0:04:44.360 --> 0:04:48.240 with TPUs and when you think about, you know, token efficiency, 0:04:48.320 --> 0:04:51.880 there is no better company than Google when it comes 0:04:51.920 --> 0:04:55.120 to how they have ramped up TPUs for both training 0:04:55.160 --> 0:04:58.880 and influencing workloads. And now that Meta is getting more 0:04:58.920 --> 0:05:03.200 confident about it's on model, they have redesigned the pre 0:05:03.279 --> 0:05:06.960 training stack over the last nine months. They feel they're 0:05:07.000 --> 0:05:10.560 at a point where they are looking for token efficiency 0:05:10.640 --> 0:05:12.600 and I think that's where they want to go to 0:05:12.680 --> 0:05:15.640 the custom silicon route. It's worked for them when it 0:05:15.680 --> 0:05:18.719 comes to recommendation engines and now they want to do 0:05:19.080 --> 0:05:22.559 the same on the accelerator side with Broadcom Mandy. 0:05:22.560 --> 0:05:24.520 But I want to look at the specific Meta chips 0:05:24.560 --> 0:05:29.640 themselves MTIA Meta Training Inference Accelerator. These are some images 0:05:29.720 --> 0:05:34.000 that we took when we visited the MTIA Lab last 0:05:34.040 --> 0:05:36.800 month and basically how it works is right. They work 0:05:36.839 --> 0:05:39.760 with Broadcom on the design and architecture of the silicon 0:05:40.240 --> 0:05:43.600 and Broadcom then goes to TSMC and TS and what 0:05:43.680 --> 0:05:48.440 this deal represents is expanded volume commitment on that existing technology. 0:05:49.080 --> 0:05:50.960 You can check out you know, Bloomber's coverage of the 0:05:51.000 --> 0:05:54.800 specifics of the accelerators. At the same time, Meta is 0:05:54.839 --> 0:05:58.839 a big buyer of GPUs and other accelerators from elsewhere. 0:05:59.240 --> 0:06:00.120 That part of your. 0:06:00.080 --> 0:06:03.720 Research that you say they might be inhibited by lack 0:06:03.760 --> 0:06:06.880 of a cloud computing business. What do you mean by 0:06:06.880 --> 0:06:11.280 that if there's such a great buyer of capacity, Well. 0:06:11.200 --> 0:06:12.400 So think of it this way. 0:06:12.520 --> 0:06:16.640 If you are buying in video chips in volume, you 0:06:16.760 --> 0:06:21.120 want to rent them through your cloud business because for Microsoft, 0:06:21.160 --> 0:06:24.599 Azure or Amazon or even Google for that matter, it's 0:06:24.640 --> 0:06:28.720 a big source of you know, cloud revenue, cloud infrastructure 0:06:28.760 --> 0:06:31.880 revenue that these companies are generating in their respective segments. 0:06:32.160 --> 0:06:36.120 For a Meta, if you're focused on token efficiency and 0:06:36.200 --> 0:06:39.760 you're buying the most expensive chips out there without a 0:06:39.760 --> 0:06:44.240 cloud business, that makes that ROI equation very hard. And 0:06:44.480 --> 0:06:46.479 you know, they're at a point where they feel good 0:06:46.480 --> 0:06:49.400 about their model. They have three billion users to serve 0:06:49.640 --> 0:06:51.960 in terms of inferencing, they want to do it with 0:06:52.000 --> 0:06:52.640 their own chip. 0:06:53.520 --> 0:06:57.360 Is there average chance that with excess compute it does 0:06:57.400 --> 0:06:59.520 start becoming more of a cloud company matter. 0:07:01.600 --> 0:07:04.640 I don't think so, just because of the surface area 0:07:04.720 --> 0:07:08.479 they have with the family of apps, you know, WhatsApp, Instagram, 0:07:08.560 --> 0:07:11.960 the core Blue App, three billion users. If we are 0:07:12.080 --> 0:07:16.560 talking about personalized intelligence and you know AI agents that 0:07:16.800 --> 0:07:20.920 really work for consumers, they have you know, a lot 0:07:20.960 --> 0:07:24.440 of demand for inferencing that they can use their own 0:07:24.480 --> 0:07:27.880 infrastructure force. So to my mind, they should have done 0:07:27.960 --> 0:07:30.800 this probably a few years back in terms of starting 0:07:30.840 --> 0:07:33.480 a cloud business like Oracle did you know with this 0:07:33.600 --> 0:07:35.840 dep wave, But probably it's too late now. 0:07:36.640 --> 0:07:40.280 Those MTIA chips, what they are useful currently is is 0:07:40.360 --> 0:07:44.480 ranking recommendations in timelines. That's right, really focused on inference. 0:07:44.800 --> 0:07:47.400 So the idea that they leased them out, you've got 0:07:47.400 --> 0:07:48.880 to come back and get more into that. Man Deep 0:07:48.920 --> 0:07:52.000 seeing a Bloomberg Intelligence top Research. Thank you very much. 0:07:52.280 --> 0:07:54.840 Let's continue with this sort of AI demand story. Look 0:07:54.840 --> 0:07:58.520 at shares of ASML down four point four percent. Now 0:07:58.560 --> 0:08:01.960 the supplier cutting edge hip making Machines raised its four 0:08:02.040 --> 0:08:05.000 year sales forecast, but that brought it in line with 0:08:05.080 --> 0:08:08.000 analyst estimates and attention already turning to what can be 0:08:08.040 --> 0:08:11.480 expected for next year twenty twenty seven. Let's discuss with 0:08:11.600 --> 0:08:15.520 Tammy Chew. She's the head of Tech equity research at Behremberg. 0:08:15.600 --> 0:08:20.200 So it's kind of this contradiction, right. The headline is 0:08:20.240 --> 0:08:24.520 like really positive, raising that guidance, raising that outlook. The 0:08:24.600 --> 0:08:28.080 demand is clearly there, the stock falls and so like, 0:08:28.200 --> 0:08:31.760 it seems like the expectations were high. People are already 0:08:31.760 --> 0:08:35.360 looking much further ahead in terms of that demand picture. 0:08:35.400 --> 0:08:36.280 What was your read on it? 0:08:37.720 --> 0:08:40.360 Sure, So that's a very big point. So basically I 0:08:40.360 --> 0:08:43.959 would say don't forget that. To start with, ASMI is 0:08:44.000 --> 0:08:46.760 already up nearly forty percent year to date, and semi 0:08:46.760 --> 0:08:49.760 equipment segment has been one of the strongest i would 0:08:49.760 --> 0:08:52.720 say subsegment in tech year to date in this year. 0:08:53.080 --> 0:08:55.840 So therefore I would call the set of results reported 0:08:55.840 --> 0:08:59.280 by ASMO today as solid. They raised a four year 0:08:59.320 --> 0:09:02.520 guidance by four percent. The Q one gross margin was 0:09:02.600 --> 0:09:06.040 way above constant expectation because of strow mix. At the 0:09:06.040 --> 0:09:09.679 same time, they confirm that they are increasing their capacity 0:09:10.120 --> 0:09:14.280 to about at least eighty UV tools from at least 0:09:14.320 --> 0:09:17.440 sixty UVI tools year in twenty twenty seven. So I 0:09:17.480 --> 0:09:19.640 would say the set of results has been very strong, 0:09:19.920 --> 0:09:24.040 but at the same time as the most let's say successful, 0:09:24.200 --> 0:09:29.040 our high profile tech company in Europe Asmo's always well 0:09:29.040 --> 0:09:32.720 owned and asmox location has always been elevated. So I 0:09:32.760 --> 0:09:36.280 would call today's share price movement as just lack of 0:09:36.880 --> 0:09:40.079 revision on the back of this solid set of results 0:09:40.120 --> 0:09:41.719 instead of anything which is worse to me. 0:09:42.440 --> 0:09:44.600 Can we just linger for a moment what you just said, 0:09:44.880 --> 0:09:49.160 eighty eight zero EUV machines. So a lot of people 0:09:49.200 --> 0:09:51.680 they'll say, hold on, I don't understand that this is 0:09:51.720 --> 0:09:54.960 the world's most important maker of machines in the chip 0:09:55.000 --> 0:09:58.400 manufacturing process and they're guiding that they'll make eighty of them. 0:09:58.640 --> 0:10:01.360 Of course, the context is there one hundred million dollars apiece. 0:10:01.920 --> 0:10:03.880 Just explain why that number is significant? 0:10:05.000 --> 0:10:05.240 Sure? 0:10:05.280 --> 0:10:08.960 Of course, So those tools you are correct are sold 0:10:09.000 --> 0:10:11.480 for two hundred and thirty million each, So those are 0:10:11.600 --> 0:10:14.360 probably one of the most expensive set of tool in 0:10:14.360 --> 0:10:17.240 the market or in the world. And I would say 0:10:17.280 --> 0:10:21.320 that it's important because we are moving from sixty in 0:10:21.400 --> 0:10:24.920 this year and we were at about forty few years ago. 0:10:25.280 --> 0:10:29.800 So therefore, from forty to eighty, they're already doubling that capacity. 0:10:29.880 --> 0:10:32.840 They can serve the market, which is showing that AI 0:10:32.920 --> 0:10:36.800 demand and memory demand has been super solid over the 0:10:36.840 --> 0:10:40.360 next let's say six months or two years, and going 0:10:40.440 --> 0:10:44.680 forward is likely to support the further expuilled expansion of 0:10:44.760 --> 0:10:47.760 AI market or memory market. So therefore, as a result, 0:10:47.800 --> 0:10:50.400 I would say that eighty number, it doesn't look like 0:10:50.440 --> 0:10:53.559 a big number, but don't forget those tools are two 0:10:53.679 --> 0:10:56.679 hundred and thirty million each and we were at forty 0:10:56.840 --> 0:10:57.600 two years ago. 0:10:58.520 --> 0:11:01.240 Are they in any way supply strained here? Tammy? I 0:11:01.240 --> 0:11:03.280 mean some on the straight and hope to see the 0:11:03.360 --> 0:11:07.520 number ninety thrown around. Is there a limitation factor to them? 0:11:07.600 --> 0:11:11.280 Could they be making more so? 0:11:11.360 --> 0:11:13.280 In our being, we don't think that asma will be 0:11:13.320 --> 0:11:16.520 a bottleneck. You are right that ninety was mentioned previously 0:11:17.000 --> 0:11:21.240 as the potential capacity number, but ASMO has been super 0:11:21.280 --> 0:11:25.080 careful in terms of building capacity as for customer demand, 0:11:25.280 --> 0:11:27.080 because you don't want to end up in the situation 0:11:27.160 --> 0:11:29.679 that you build ninety tools or whatever the number is, 0:11:29.840 --> 0:11:32.199 but there is no demand for that. So therefore, as 0:11:32.240 --> 0:11:34.640 a result, I would say that eighty number is a 0:11:34.679 --> 0:11:38.960 result of smo's discussion with customers and that eighty number 0:11:39.240 --> 0:11:42.720 for the time being is a reflection of asmo's agreement 0:11:42.760 --> 0:11:45.800 with customer how to serve the and the market demand. 0:11:46.480 --> 0:11:49.240 And I don't think it's a limitation in theory. In 0:11:49.240 --> 0:11:52.480 my view, Asma can actually do more. Is the customer 0:11:52.520 --> 0:11:56.080 need more and what ASMO has been needing is to 0:11:56.160 --> 0:11:59.240 hire more people expand with their supply chain. So I 0:11:59.240 --> 0:12:00.920 don't think or be the bottleneck. 0:12:01.559 --> 0:12:07.120 Let's talk about the demand constraints or fulfillment, because it 0:12:07.160 --> 0:12:09.520 feels as though it's insatiable in terms of need for 0:12:09.600 --> 0:12:12.439 chips and therefore needs for equipment. But they had exposure 0:12:12.480 --> 0:12:14.480 to China. How is that continued to unfold? 0:12:15.640 --> 0:12:19.040 So from a China perspective, China used to be forty 0:12:19.080 --> 0:12:21.760 percent of their business two years ago and now with 0:12:21.920 --> 0:12:25.200 the non China business been growing strongly, China is about 0:12:25.320 --> 0:12:28.120 sub twenty percent of the total business, and China has 0:12:28.160 --> 0:12:31.280 been stable. I would say China was slightly better than 0:12:31.320 --> 0:12:34.400 the previous expected and China will be stable in this year. 0:12:34.600 --> 0:12:37.760 But in this quarter, the four year race was purely 0:12:37.800 --> 0:12:41.000 because of non China business was stronger than expected. So 0:12:41.120 --> 0:12:43.800 going forward, I would say ASM what will be continued 0:12:44.080 --> 0:12:47.040 to be driven by this non China business, AI investment 0:12:47.400 --> 0:12:52.199 memory shortage to further continue as the case for the 0:12:52.280 --> 0:12:55.320 next few years. And China. I do have the view 0:12:55.360 --> 0:12:58.160 that China will be flatish and China is working on 0:12:58.240 --> 0:13:01.360 their localization. Of course, month for China is not likely 0:13:01.400 --> 0:13:03.880 to fall off the cliff. So therefore, as a result, 0:13:04.360 --> 0:13:06.959 in conclusion, over the next few years a SMO will 0:13:07.000 --> 0:13:10.439 be driven by both factors, which non China business will 0:13:10.600 --> 0:13:11.760 likely to be the key driver. 0:13:12.960 --> 0:13:16.079 Talking about two hundred and thirty million for for EUV, 0:13:16.280 --> 0:13:19.160 I mean if you think about twin scan exc some 0:13:19.200 --> 0:13:21.199 of the people I talked to you go sub tun animeter. 0:13:21.720 --> 0:13:23.640 They can get three hundred and fifty million for that. 0:13:23.800 --> 0:13:26.720 We'll get back to it at another day. There's a 0:13:26.720 --> 0:13:31.840 big project. It's called TERRAFAB. It's basically on paper, but 0:13:32.160 --> 0:13:35.559 it says that in the future there'll be one terrawat 0:13:35.600 --> 0:13:40.200 of compute capacity put through fabs operated by Tesla, SpaceX 0:13:40.240 --> 0:13:43.040 and XAI. That would on paper make them one of 0:13:43.080 --> 0:13:46.840 the biggest buyers of ASML's equipment. Have you and the 0:13:46.880 --> 0:13:49.760 team modeled for that being real or for you, is 0:13:49.800 --> 0:13:51.120 it still an idea. 0:13:53.040 --> 0:13:58.080 I mean, we would love to and you're absolutely right. 0:13:58.160 --> 0:14:00.240 So ASMO do have a version of machine which cost 0:14:00.280 --> 0:14:02.240 three hundred and fifty million, which is called hi Enna. 0:14:02.720 --> 0:14:06.680 That's probably we'll be seen when Terrrafab is ready to 0:14:06.720 --> 0:14:10.840 take equipment. So for the time being, we haven't put 0:14:11.040 --> 0:14:15.960 Terrorfab in our numbers, but as for our understanding, the 0:14:16.080 --> 0:14:18.880 intention is real and there has been discussion already within 0:14:18.920 --> 0:14:21.720 the supply chain of our Terrafab, but for the time being, 0:14:22.000 --> 0:14:25.240 it won't be something happening in the significant way from 0:14:25.240 --> 0:14:27.680 a number perspective over the next two years, So therefore 0:14:27.680 --> 0:14:30.760 we haven't reflected that, but I do hope that we 0:14:30.800 --> 0:14:32.400 can put that in numbers in two years. 0:14:32.440 --> 0:14:34.320 Siven, We love the numbers. Timing to you, thanks for 0:14:34.440 --> 0:14:38.040 joining us, head of Tech Equity Research at Bahreenberg. Great 0:14:38.080 --> 0:14:40.480 to have you with us. Now, we've got a key 0:14:40.480 --> 0:14:44.520 conversation coming up with the IMF Managing Director Carstilina goo 0:14:44.640 --> 0:14:47.600 Java as the IMF Spring meetings they are underway in Washington. 0:14:47.600 --> 0:14:49.640 We're going out to Francin Laquar in a minute, discipling 0:14:49.720 --> 0:15:05.240 their tech. Let's talk about where the IMF is right now. 0:15:05.280 --> 0:15:09.240 Spring meetings are upon us in Washington. Francin Lava is 0:15:09.360 --> 0:15:12.680 alongside the IMF Managing Director, Christine Millenakle, gavea please take 0:15:12.680 --> 0:15:13.680 it away, fancy. 0:15:16.200 --> 0:15:16.920 Yeah, thanks so much. 0:15:16.920 --> 0:15:19.360 I'm so delighted to be really here by the host 0:15:19.720 --> 0:15:22.320 of the whole show where everyone from around the world 0:15:22.360 --> 0:15:25.480 shows up, Christina Gregeva, thank you so much for joining us. 0:15:25.800 --> 0:15:28.800 Look the IMF and your world economic outlook. Everyone here 0:15:28.840 --> 0:15:31.360 has read it three times just to understand exactly what 0:15:31.440 --> 0:15:32.320 the dynamic is. 0:15:32.920 --> 0:15:35.400 You've down credited your forecast. But there's also a. 0:15:35.320 --> 0:15:37.840 Scenario where the world economy skims recession. 0:15:38.240 --> 0:15:43.920 How likely is that depends on the longevity of the 0:15:43.960 --> 0:15:47.680 war in the Middle East and the scale of infrastructure 0:15:48.240 --> 0:15:51.920 destruction that happens see in the countries over there, and 0:15:52.040 --> 0:15:56.920 that is the big anon. We now have ceased fire. 0:15:58.080 --> 0:16:01.840 I pray it turns into a ruble peace, and in 0:16:01.880 --> 0:16:07.040 that case we can see somewhat faster recovery. But the 0:16:07.160 --> 0:16:11.720 uncertainty has been profound, and we only need to learn 0:16:12.160 --> 0:16:16.240 to operate in an environment of high end and permanent uncertainty. 0:16:16.920 --> 0:16:19.800 And even if there's a ceasefire that stays okay and 0:16:19.880 --> 0:16:23.840 there's some kind of diplomatic resolution. Is a market discounting 0:16:24.000 --> 0:16:26.120 the real impact on the price of all and their 0:16:26.120 --> 0:16:26.720 for inflation. 0:16:27.080 --> 0:16:29.080 Well, let me just say that even if the war 0:16:29.200 --> 0:16:33.160 ends tomorrow, it would take quite some time for the 0:16:33.240 --> 0:16:40.320 recovery to kick in, not only because of the infrastructure destruction. 0:16:41.640 --> 0:16:48.080 Tankers are slow moving vessels. A tanker that leads today 0:16:48.480 --> 0:16:53.160 would take forty days to reach the Pacific islands, So 0:16:53.240 --> 0:16:56.480 we have to keep in mind that the impact is 0:16:56.520 --> 0:17:04.800 already baked in. Why are markets so optimistic? I think 0:17:05.160 --> 0:17:07.520 part of the reason is because there is a high 0:17:07.560 --> 0:17:11.560 concentration of financial assets in the United States. The US 0:17:11.600 --> 0:17:17.439 economy is doing well. It is relatively not so impacted 0:17:17.920 --> 0:17:22.200 because the US is an export and not importer, and 0:17:22.240 --> 0:17:26.280 that creates that picture of dynamism and vibrancy. But I 0:17:26.320 --> 0:17:28.600 can tell you that's not the story of the rest 0:17:28.600 --> 0:17:31.439 of the world. In the rest of the world already 0:17:31.520 --> 0:17:32.560 there is a lot of pain. 0:17:32.920 --> 0:17:35.320 So I mean, does that mean that the market is mispricing, 0:17:35.520 --> 0:17:38.560 that all this market exuberance is actually misplaced? 0:17:39.000 --> 0:17:43.240 What it means is that markets are geared by events 0:17:43.280 --> 0:17:46.359 in the largest economy that are really positive. The United 0:17:46.400 --> 0:17:52.040 States is the only large economy that has high productivity growth, 0:17:52.160 --> 0:17:56.640 and markets are seeing that technology optimism is here. Markets 0:17:56.680 --> 0:18:01.679 are rushing on it. But should they be more caution? 0:18:02.880 --> 0:18:07.439 I would argue yes, because what we are seeing in 0:18:07.560 --> 0:18:12.920 terms of supply chain disruptions is already quite significant. And 0:18:12.960 --> 0:18:16.879 with every day that passes, that anchor is not levius 0:18:17.359 --> 0:18:22.720 living for a destination, and that delays the rebounds of 0:18:22.760 --> 0:18:23.680 the world economy. 0:18:24.080 --> 0:18:26.960 How much do you worry about governments trying to overadjust 0:18:27.040 --> 0:18:30.240 if inflation expectations goes up, and actually how much of 0:18:30.520 --> 0:18:31.520 a mess would this be? 0:18:31.800 --> 0:18:35.000 And so such an important question. So look at what 0:18:35.160 --> 0:18:40.720 is happening. Short term inflation expectations have moved up, not 0:18:40.840 --> 0:18:46.119 by much, though we have increased our inflation projection from 0:18:46.200 --> 0:18:49.919 three point eight percent to four point two percent. It 0:18:50.040 --> 0:18:53.920 is up, but not dramatically and very important. Long term 0:18:54.160 --> 0:19:00.800 inflation expectations don't budge. They're well anchored made they stay 0:19:00.920 --> 0:19:05.320 this way. So it is very important that central banks 0:19:05.560 --> 0:19:12.080 act carefully. Those that are in a good position of credibility, 0:19:13.040 --> 0:19:17.520 they can take weight and see attitude. They also have 0:19:17.640 --> 0:19:20.879 to signal we are prepared to act if it is necessary. 0:19:20.960 --> 0:19:24.280 But please don't rush my word is because of twenty two. 0:19:25.400 --> 0:19:28.960 Now central banks may say let's move faster, and that 0:19:29.040 --> 0:19:32.000 could be very dangerous because it would suffocate growth. 0:19:32.280 --> 0:19:35.120 I mean, we're looking back at every IMF World Bank 0:19:35.520 --> 0:19:38.520 meetings since twenty twenty. They spent something. Twenty twenty was 0:19:38.520 --> 0:19:41.000 COVID twenty twenty two. Of course, it was the Russian 0:19:41.000 --> 0:19:43.879 invasion of Ukraine. Then you had terriffs. I mean a 0:19:44.000 --> 0:19:45.200 shock after shock. 0:19:46.000 --> 0:19:53.040 So it looks like a layer cake, shock upon shock 0:19:53.520 --> 0:19:57.879 upon shock, except it is not at all tasty. I 0:19:57.920 --> 0:20:03.760 can tell you that everybody has to absorb reality and realities. 0:20:04.000 --> 0:20:06.720 We are in a more shocked prone world. Shocks will 0:20:06.760 --> 0:20:11.120 continue to come. What does it mean for fiscal authorities? 0:20:11.359 --> 0:20:14.640 The buffers when you are in good time, don't way 0:20:14.800 --> 0:20:21.320 your resources for monetary policy, Calibrate policy carefully, be prepared 0:20:21.359 --> 0:20:27.840 to active if and when necessary, and everywhere. Countries have 0:20:27.960 --> 0:20:32.480 to be much more alert to the high uncertainty within 0:20:32.560 --> 0:20:37.119 which we operate. Agility, adaptability, this is what we have 0:20:37.240 --> 0:20:37.520 to do. 0:20:38.280 --> 0:20:40.280 How worried are you about the UK? This was one 0:20:40.320 --> 0:20:41.960 of the countries I was down graded the most. 0:20:42.280 --> 0:20:44.639 I mean they are downgraded, but also we see the 0:20:44.800 --> 0:20:48.960 UK very mature approach to policy. The UK is very 0:20:48.960 --> 0:20:55.240 careful not to go into indiscriminate spending and Bank of 0:20:55.400 --> 0:21:00.119 England is very clear around the approach they are going 0:20:59.920 --> 0:21:04.119 to take the right approach to this situation. Actually, we 0:21:04.160 --> 0:21:07.800 see the UK in terms of fiscal response as a 0:21:07.840 --> 0:21:11.560 good example for others, because many other countries are rushing 0:21:12.119 --> 0:21:15.320 to throw money they don't have out of the good 0:21:15.359 --> 0:21:19.239 intention to help people. It has to be targeted, it 0:21:19.280 --> 0:21:21.760 has to be temporary, and it has to be done 0:21:22.080 --> 0:21:25.840 within the fiscal band boundaries of countries, and the UK 0:21:26.000 --> 0:21:28.800 is doing that. UK is also a country where there 0:21:28.840 --> 0:21:33.400 is a recognition that your best defense is growth. Pro 0:21:33.520 --> 0:21:36.840 growth policies are gonna be your buffer for the future. 0:21:37.200 --> 0:21:37.879 Not easy, But. 0:21:41.720 --> 0:21:43.720 I also want to ask you about Look the US 0:21:43.960 --> 0:21:47.000 is again President Trump is going after j Powell saying, look, 0:21:47.040 --> 0:21:49.800 you'll fire family, doesn't step down from the Fed. I mean, 0:21:49.840 --> 0:21:52.280 we go back to this idea of central bank independence. 0:21:52.280 --> 0:21:55.760 How problematic is it, especially in a time where actually 0:21:55.760 --> 0:21:59.560 central banks are a very very difficult job at hand. 0:21:59.800 --> 0:22:05.640 We have decades of evidence that independence center banks are 0:22:05.640 --> 0:22:11.280 an economic asset. They serve well when they have the 0:22:11.440 --> 0:22:15.879 opportunity to make decisions and the space to make decisions responsibly. 0:22:16.480 --> 0:22:20.439 And I can tell you one very interesting fact, center 0:22:20.560 --> 0:22:28.480 banks in emerging market economies outperformed their advanced economy peers. 0:22:28.560 --> 0:22:33.240 They have built that independence strength of data dependent policy making. 0:22:33.880 --> 0:22:37.080 Again, market seems to look through a lot of these concerns. 0:22:37.359 --> 0:22:41.359 Just we're you that the function of almost keeping things 0:22:41.400 --> 0:22:43.080 and tracks working differently. 0:22:43.200 --> 0:22:45.879 On one level, it is good that the economy functions 0:22:45.960 --> 0:22:50.920 because getting into panic is not going to help. So 0:22:51.520 --> 0:22:56.199 good functioning of markets is helping. Having that, we have 0:22:56.280 --> 0:22:59.879 seen titaning of financial conditions by some but not by man. 0:23:00.560 --> 0:23:04.320 We have seen currency depreciation of some emerging markets, again 0:23:04.480 --> 0:23:10.239