WEBVTT - Bloomberg Surveillance TV: August 31st, 2026 0:00:02.430 --> 0:00:04.019 Bloomberg Audio Studios. 0:00:04.540 --> 0:00:06.710 Podcasts Radio News. 0:00:11.619 --> 0:00:15.440 This is the Bloomberg Surveillance Podcast. I'm Jonathan Farrow, along 0:00:15.480 --> 0:00:18.640 with Lisa Abramowitz and Anne-Marie Hordern. Join us each day 0:00:18.700 --> 0:00:22.180 for insight from the best in markets, economics and geopolitics. 0:00:22.500 --> 0:00:24.909 From our global headquarters in New York City, we are 0:00:24.950 --> 0:00:28.290 live on Bloomberg Television weekday mornings from 6 to 9 a.m. Eastern. 0:00:28.770 --> 0:00:31.930 Subscribe to the podcast on Apple, Spotify or anywhere else 0:00:31.970 --> 0:00:34.690 you listen. And as always, on the Bloomberg Terminal and 0:00:34.710 --> 0:00:37.870 the Bloomberg Business App. We begin this hour with stocks 0:00:37.909 --> 0:00:40.820 little changed as markets digest Fed chair Kevin Walsh's Jackson 0:00:40.860 --> 0:00:43.740 Hole message. Bob Elliott of Unlimited Funds writing, while the 0:00:43.780 --> 0:00:46.659 Walsh speech was taken as a tad hawkish, it looked 0:00:46.700 --> 0:00:51.159 more like continued policy ineptitude, relying on rhetoric rather than 0:00:51.220 --> 0:00:53.010 clear action. Bob joins us now for more. 0:00:53.030 --> 0:00:53.970 Bob, good morning. Good morning. 0:00:54.130 --> 0:00:55.709 Are you looking for a hold in September? 0:00:56.130 --> 0:01:00.050 Well, I think the real question is, is he actually 0:01:00.070 --> 0:01:03.540 going to follow through on the rhetoric that at least 0:01:03.600 --> 0:01:06.480 he outlined? And I agree with this basic idea that 0:01:06.520 --> 0:01:09.630 maybe he's boxed himself in, in terms of if you've 0:01:09.650 --> 0:01:13.690 missed your target for 60-plus months— and inflation is above 0:01:14.310 --> 0:01:17.510 what your mandate is by a lot, and the employment 0:01:17.569 --> 0:01:21.090 conditions are not so bad, it seems like they should hike. 0:01:21.550 --> 0:01:23.930 And yet, there's all of this ambiguity. 0:01:23.950 --> 0:01:26.290 The market is looking at this and saying, is this 0:01:26.330 --> 0:01:29.169 a Fed that's actually credible in delivering the tightening? Or 0:01:29.270 --> 0:01:32.220 is this, as they say out West, all hat and 0:01:32.280 --> 0:01:34.590 no cattle from the new Fed chairman? 0:01:34.860 --> 0:01:36.500 Where do you see the credibility gap in the market 0:01:36.520 --> 0:01:36.860 right now? 0:01:37.140 --> 0:01:39.740 Well, I mean, I think the basic idea is if 0:01:39.760 --> 0:01:42.980 you look at any sort of simple Taylor rule here, 0:01:43.319 --> 0:01:46.200 you'd expect interest rates on the short end to be 0:01:46.260 --> 0:01:48.420 up 100 or 200 basis points. I love the fact 0:01:48.460 --> 0:01:50.840 the Atlanta Fed puts out a whole range of different 0:01:50.860 --> 0:01:53.800 Taylor rule estimates. It's almost like they're trolling those folks 0:01:53.840 --> 0:01:56.460 in the Eccles building saying, hey, look, if you're going 0:01:56.480 --> 0:01:59.760 to actually deliver monetary policy consistent with what we've talked 0:01:59.800 --> 0:02:02.040 about that's credible, you've got to be raising rates. 0:02:02.060 --> 0:02:03.600 You've got to be raising rates quickly. 0:02:03.700 --> 0:02:06.840 And yet we're in this argument of, you know, is 0:02:06.940 --> 0:02:09.120 it 40-60 that they're going to tighten in September or 60-40? 0:02:09.160 --> 0:02:14.049 The data is unambiguous. It's meaningful hikes relatively quickly. And 0:02:14.070 --> 0:02:16.130 the odds that we get that are de minimis at 0:02:16.169 --> 0:02:18.590 this point. And so the real question is, when the 0:02:18.630 --> 0:02:22.220 market is looking at that, they're basically pushing the long 0:02:22.340 --> 0:02:25.460 end to basically penalize the Fed not doing their job 0:02:25.520 --> 0:02:25.980 on the short end. 0:02:26.200 --> 0:02:28.690 Couldn't this be a great way to kill forward guidance, though, 0:02:28.710 --> 0:02:31.450 to say, hey, market, stop parsing through every single line 0:02:31.490 --> 0:02:33.470 of my speech to try to get signal for what 0:02:33.510 --> 0:02:35.290 we are doing? Isn't that a great way to stop 0:02:35.310 --> 0:02:38.070 that from happening by not hiking in September? 0:02:38.370 --> 0:02:38.570 Sure. 0:02:38.610 --> 0:02:43.250 I mean, incompetence can be a way to kill forward guidance, right? 0:02:43.310 --> 0:02:45.150 That's certainly a way in which they could do it. 0:02:45.490 --> 0:02:48.070 But the reality is what the Fed needs is to 0:02:48.110 --> 0:02:50.570 be credible. And when you stand up there and you say, 0:02:50.830 --> 0:02:54.090 we failed at our mandate for 60-plus months— and you 0:02:54.180 --> 0:02:59.000 don't deliver based upon the data, that's the thing that's actually, 0:02:59.040 --> 0:03:01.120 the credibility is what matters a lot more than the 0:03:01.180 --> 0:03:01.720 forward guidance. 0:03:01.860 --> 0:03:05.250 And the irony, perhaps, is the 10-year yields were initially 0:03:05.290 --> 0:03:08.109 moved lower on the speech from this chair and ended 0:03:08.150 --> 0:03:08.770 the day higher. 0:03:08.830 --> 0:03:09.790 Peter Scheer points. 0:03:09.590 --> 0:03:11.990 To the large amount of IG supply that we're going 0:03:12.010 --> 0:03:14.580 to be getting from this market. Is the 30-year yield 0:03:14.900 --> 0:03:17.160 really a read on Fed credibility at this moment? 0:03:17.780 --> 0:03:20.160 Well, I think I wouldn't necessarily just look at the 0:03:20.180 --> 0:03:22.899 30-year yield in terms of Fed credibility. 0:03:22.919 --> 0:03:23.919 I'd be looking at gold. 0:03:24.000 --> 0:03:27.540 And I think gold, in general, is giving us an indication. 0:03:27.600 --> 0:03:30.060 It's reacceleration over the course of the last couple of weeks. 0:03:30.380 --> 0:03:33.910 It's giving some indication that people are really questioning not 0:03:33.970 --> 0:03:38.250 just the Fed, but the policymaking institutions across the administration, 0:03:38.270 --> 0:03:40.350 whether it be at the Treasury or at the Fed 0:03:40.650 --> 0:03:43.630 or in the administration. And it's saying, you know, despite 0:03:43.690 --> 0:03:46.430 all these efforts to try and keep rates low, gold 0:03:46.470 --> 0:03:48.090 is pushing higher. Of course, Friday it sold off a 0:03:48.130 --> 0:03:51.710 little bit. But the general trend here is for all 0:03:51.750 --> 0:03:54.110 the effort that the administration is doing to try and 0:03:54.150 --> 0:03:57.450 keep rates low, they're basically just flowing money into crypto 0:03:57.490 --> 0:03:58.170 and gold these days. 0:03:58.230 --> 0:04:00.430 Gold up 10% so far this month. At one point, 0:04:00.490 --> 0:04:02.630 up 16. We were on course for a monster month. 0:04:02.910 --> 0:04:04.120 Is the gold trade back on for you? 0:04:04.560 --> 0:04:05.060 Oh, for sure. 0:04:05.080 --> 0:04:06.760 I mean, I think if you look at this whole 0:04:06.840 --> 0:04:10.800 scope of different pressures, whether it be the continued conflict 0:04:10.840 --> 0:04:14.910 in the Middle East or the administration's lack of credibility 0:04:14.950 --> 0:04:17.690 in policymaking across a number of different areas, this is 0:04:17.730 --> 0:04:20.010 setting up an environment where hard assets. 0:04:20.570 --> 0:04:22.860 are outperforming soft assets, right. 0:04:22.960 --> 0:04:26.339 Where you see gold and oil continuing to be strong, 0:04:26.600 --> 0:04:30.180 where you see bonds and stocks soft in this sort 0:04:30.200 --> 0:04:33.420 of environment. And I think that's the complexion that we're 0:04:33.440 --> 0:04:35.710 starting to see here. You get that ball rolling a 0:04:35.730 --> 0:04:37.830 little bit, and that could really define what we're seeing 0:04:37.850 --> 0:04:38.370 in the second quarter. 0:04:38.490 --> 0:04:40.670 Is that another way of expressing a dollar short or 0:04:40.710 --> 0:04:41.549 is it something different? 0:04:42.010 --> 0:04:42.349 For sure. 0:04:42.390 --> 0:04:44.409 Because I think the challenge when you think about the 0:04:44.450 --> 0:04:47.550 dollar is a dollar against what? I mean, the dollar 0:04:47.570 --> 0:04:50.050 against the yen just keeps falling every day. Dollar against 0:04:50.070 --> 0:04:52.750 the euro. You know, that's a pretty soft economy in 0:04:52.800 --> 0:04:55.180 and of itself. And so I think investors are starting 0:04:55.200 --> 0:04:59.020 to move away from trying to express the U.S. dollar 0:04:59.120 --> 0:05:02.680 and the fiat currency trade across crosses. And they're moving 0:05:02.720 --> 0:05:04.840 more to hard assets. And that's exactly what we're seeing. 0:05:05.040 --> 0:05:05.900 What reverses that? 0:05:06.000 --> 0:05:07.929 Is there anything you see that could stop that trend 0:05:07.990 --> 0:05:10.690 from the path that we're witnessing now? 0:05:10.870 --> 0:05:11.210 Yeah, sure. 0:05:11.330 --> 0:05:14.120 A credible 200 basis points of tightening. But it ain't 0:05:14.130 --> 0:05:14.560 going to happen. 0:05:14.720 --> 0:05:17.719 But that doesn't alleviate the geopolitical risk you're talking about, 0:05:17.779 --> 0:05:20.140 about the U.S. weaponizing the U.S. 0:05:20.180 --> 0:05:20.919 Financial system? 0:05:21.560 --> 0:05:24.150 Yeah, and I think what it would do is it 0:05:24.170 --> 0:05:27.510 would basically say we're willing to take the pain necessary, 0:05:27.550 --> 0:05:30.950 we meaning the U.S. economy, and developed economies across the 0:05:30.970 --> 0:05:33.219 world are willing to take the pain and tighten the 0:05:33.260 --> 0:05:37.140 belt in response to the geopolitical pressures, the rising inflation 0:05:37.460 --> 0:05:40.440 pressures that we're seeing in oil, but we're also seeing 0:05:40.460 --> 0:05:42.400 it in other commodity areas like. 0:05:42.440 --> 0:05:44.170 Copper reaching new highs, etc. 0:05:44.830 --> 0:05:48.330 The question is, are we going to fulfill the mandate 0:05:48.370 --> 0:05:50.510 of these central banks in the U.S. 0:05:50.570 --> 0:05:52.010 And Europe and Japan? 0:05:52.470 --> 0:05:56.180 Or are we going to keep things easy and not 0:05:56.220 --> 0:05:59.480 take the pain and allow these inflationary pressures to build globally? 0:05:59.500 --> 0:06:00.700 And that's basically what we're seeing. 0:06:01.020 --> 0:06:03.140 One of the interesting things in that speech from Friday 0:06:03.180 --> 0:06:05.540 was his read on inflation and how he's using it. 0:06:05.589 --> 0:06:08.810 He brought in two new indicators for us, speed and breadth. 0:06:09.150 --> 0:06:11.190 And I wonder if you, with that in hand, would 0:06:11.210 --> 0:06:13.790 look at the inflation data differently and derive something different 0:06:13.810 --> 0:06:15.950 for this market too. If it's not just what is 0:06:15.990 --> 0:06:18.390 the headline number, it's what do the components look like 0:06:18.690 --> 0:06:20.740 and how quickly is it getting back to 2%. 0:06:20.740 --> 0:06:23.979 Yeah, I think it's a good example of how the 0:06:24.060 --> 0:06:26.800 equivocation in the speech. I think many people sort of 0:06:26.820 --> 0:06:30.300 looked at it at face value and said, looks hawkish, right? 0:06:30.320 --> 0:06:31.620 And we saw a bit of that in terms of 0:06:31.640 --> 0:06:33.580 the market pricing. But the reason why I didn't go 0:06:33.660 --> 0:06:37.080 further is because Warsh was opening the door to torturing 0:06:37.120 --> 0:06:40.060 the data to find the path that was necessary. 0:06:40.420 --> 0:06:42.840 So that he didn't necessarily have to hike. 0:06:42.860 --> 0:06:45.360 If you take a certain number of the components and 0:06:46.440 --> 0:06:49.380 the derivative of how those share of components are moving 0:06:49.460 --> 0:06:52.020 relative to the target, maybe you could figure out a 0:06:52.060 --> 0:06:54.220 way to try and make a case that we shouldn't 0:06:54.240 --> 0:06:56.700 have 100 to 200 basis points of tightening. But everyone 0:06:56.730 --> 0:06:59.850 in the market knows the reality, which is nominal growth 0:07:00.620 --> 0:07:03.870 is too high, inflation is too high for the Fed's mandate. 0:07:04.230 --> 0:07:06.670 And if he won't deliver the goods, then the market 0:07:06.730 --> 0:07:07.690 is going to figure out. 0:07:07.690 --> 0:07:09.570 Other ways to express that circumstance. 0:07:09.670 --> 0:07:13.150 So, let's talk about expressing that through equities. High nominal GDP, 5% 0:07:13.150 --> 0:07:15.870 to 6%, above target inflation, and a Fed that you 0:07:15.890 --> 0:07:18.170 believe is incredible right now. Isn't that a green light 0:07:18.190 --> 0:07:18.810 to buy stocks? 0:07:19.150 --> 0:07:21.870 Well, except for the extraordinary expectations that are already built in. 0:07:21.990 --> 0:07:26.950 I mean, we've priced in 25% earnings growth for the 0:07:27.010 --> 0:07:30.780 next two quarters, which … if we get, is an 0:07:30.860 --> 0:07:35.480 extraordinary run over the course of the last couple of 0:07:35.560 --> 0:07:37.780 quarters into the second half of the year. 0:07:37.800 --> 0:07:39.980 And so you've set a very, very high bar. 0:07:40.000 --> 0:07:42.000 And part of what we're seeing there, I think the 0:07:42.040 --> 0:07:45.700 market action is interesting, particularly in the high-flying names, where 0:07:45.720 --> 0:07:48.420 you see in this last earnings period a lot of 0:07:48.580 --> 0:07:52.360 beat and falls. And that gives you sort of an indication, 0:07:52.560 --> 0:07:56.400 even NVIDIA, that which ostensibly gave the best guidance they 0:07:56.420 --> 0:07:59.190 could plausibly get, got a pop and then a fall 0:07:59.370 --> 0:08:01.750 on the backside of it. And I think that highlights 0:08:01.790 --> 0:08:05.190 just how overextended this market is. They're basically using all 0:08:05.250 --> 0:08:08.470 sorts of tools in one way or another in terms 0:08:08.510 --> 0:08:11.450 of leverage in order to get higher and higher exposure 0:08:11.490 --> 0:08:15.570 to these assets and to stocks, the high-flying stocks. And 0:08:15.620 --> 0:08:19.160 at some point, you reach some exhaustion. And that's basically 0:08:19.180 --> 0:08:20.820 where we're at. And I think the thing that, from 0:08:20.880 --> 0:08:23.020 a stocks perspective, that few people are talking about, everyone's 0:08:23.040 --> 0:08:25.060 talking about the AI trade. Very few people are talking 0:08:25.080 --> 0:08:27.070 about what's going on with the consumer. Because the consumer, 0:08:27.490 --> 0:08:29.410 whether you like it or not, the US economy is 0:08:29.430 --> 0:08:32.470 a consumer economy, and that's what's driving it. And there, actually, 0:08:32.490 --> 0:08:35.960 you see a lot of interesting softness under the surface. 0:08:36.000 --> 0:08:38.880 I mean, the PCE numbers were pretty weak. The retail 0:08:38.900 --> 0:08:42.800 sales numbers were pretty weak. Suggests some exhaustion. And more 0:08:42.860 --> 0:08:46.179 timely numbers, whether it be TSA or Bloomberg second measure measures, 0:08:46.559 --> 0:08:49.820 suggest actually the consumer is starting to fade pretty quickly. 0:08:49.900 --> 0:08:52.260 That would be a big surprise for the equity market 0:08:52.280 --> 0:08:52.740 in the second half. 0:08:52.840 --> 0:08:54.610 It would be a major surprise going into year end. Bob, 0:08:54.620 --> 0:08:57.209 it's good to see you. Stay with us. More Bloomberg 0:08:57.230 --> 0:09:09.700 surveillance coming up after this. CJ Moose of Cannabis Gerald 0:09:09.740 --> 0:09:12.679 covers the chip sector. CJ joins us now for more. CJ, 0:09:12.700 --> 0:09:14.360 welcome to the program. Just take it from the top. 0:09:14.780 --> 0:09:17.500 How long can this cycle really keep on going? How 0:09:17.590 --> 0:09:20.570 durable is this cycle? A lot of people have their doubts. 0:09:20.929 --> 0:09:21.590 You're pushing back. 0:09:22.750 --> 0:09:24.000 Yeah, good morning. Thanks for having me. 0:09:24.620 --> 0:09:28.120 I think you're hitting on the key debate for all 0:09:28.200 --> 0:09:33.780 of semiconductors, the durability of AI infrastructure spending. If you 0:09:33.800 --> 0:09:37.420 look at the supply side, TSMC is effectively sold out 0:09:37.460 --> 0:09:39.260 until 2029. 0:09:39.260 --> 0:09:42.600 Memory, similar. And then there's the demand side, right? 0:09:42.640 --> 0:09:45.580 And I think there's some fears around open source modeling 0:09:45.660 --> 0:09:49.080 and the effectiveness and efficiency of models in China. And 0:09:49.480 --> 0:09:52.160 is that going to reduce the overall amount of spend? 0:09:52.870 --> 0:09:56.110 but I think internally at the largest frontier models, they 0:09:56.190 --> 0:10:00.630 see what their next gen models look like. and they 0:10:01.330 --> 0:10:03.910 are trying to get as much compute as fast as 0:10:03.970 --> 0:10:04.450 they can. 0:10:04.890 --> 0:10:06.650 Compute is sold out, right? 0:10:06.670 --> 0:10:08.740 We heard from NVIDIA last week that they can only 0:10:08.770 --> 0:10:13.140 meet 70% of the demand that they have for calendar 27. 0:10:13.140 --> 0:10:16.620 They obviously have visibility into 28 and 29. 0:10:16.620 --> 0:10:18.300 And so our view. 0:10:18.220 --> 0:10:21.860 Is that, you know, this spending will continue until 2030 minimum. 0:10:22.500 --> 0:10:25.040 But that is the core debate. You know, if you're 0:10:25.520 --> 0:10:28.719 thinking that 28 is call at the peak, then you're 0:10:28.760 --> 0:10:31.819 probably going to put a lower multiple on the entire group. 0:10:31.940 --> 0:10:32.220 Right. 0:10:32.300 --> 0:10:35.810 But if you think this can extend into 2030 and beyond, then, 0:10:36.110 --> 0:10:37.989 you know, these stocks are incredibly cheap. 0:10:38.630 --> 0:10:41.530 How important are NVIDIA's moves in funding various efforts in 0:10:41.550 --> 0:10:45.310 this ecosystem, CJ, to helping that supply issue? It's something 0:10:45.350 --> 0:10:47.450 they talked about on the call that the CFO did. 0:10:48.090 --> 0:10:50.069 Is this the lens at which we should view their 0:10:50.130 --> 0:10:53.530 investments of trying to ease some of the shortages and 0:10:53.570 --> 0:10:55.530 bottlenecks that are rife within the industry? 0:10:56.720 --> 0:10:57.559 Yeah, it's a great question. 0:10:57.580 --> 0:11:00.210 I think there's, you know, a couple of points, you know, 0:11:00.370 --> 0:11:03.250 specific to MediaTek. You know, Rick Tsai used to be 0:11:03.270 --> 0:11:07.750 the CEO of TSMC. He's great friends with Jensen. MediaTek 0:11:07.809 --> 0:11:14.109 is essentially NVIDIA's physical AI partner. And so this combination 0:11:14.170 --> 0:11:17.890 or this investment, you know, makes perfect sense to me. 0:11:18.570 --> 0:11:22.150 You know, I think the earlier comments around getting customers 0:11:22.200 --> 0:11:25.560 to buy into NVIDIA's full system rack is, even if 0:11:25.600 --> 0:11:28.020 they have to give up some of the custom silicon business, 0:11:28.360 --> 0:11:30.740 you know, is a strategy for Jensen. So I think, 0:11:31.140 --> 0:11:34.540 you know, when you look at these investments in isolation, 0:11:34.980 --> 0:11:37.100 this one makes perfect sense to me. They are their 0:11:37.160 --> 0:11:40.930 physical partner. More broadly speaking, you know, I think that 0:11:41.570 --> 0:11:44.630 Jensen is doing two things with the investments. Number one, 0:11:44.670 --> 0:11:47.810 he's securing his supply chain. You know, and I think 0:11:48.030 --> 0:11:54.620 it is multi-year contracts plus, you know, endeavoring to bring 0:11:55.080 --> 0:11:59.580 financial institutions to support the investment so that that becomes 0:11:59.640 --> 0:12:02.400 another competitive advantage for them. And then the other is 0:12:02.460 --> 0:12:05.760 just the simple notion that, you know, Anthropic and OpenAI 0:12:06.350 --> 0:12:09.189 have limited balance sheets. You know, they're competing against the 0:12:09.250 --> 0:12:11.390 largest hyperscalers on Earth. 0:12:12.170 --> 0:12:14.390 They have run out of compute. They have. 0:12:15.230 --> 0:12:18.710 visibility to demand for more compute to support their models. 0:12:19.510 --> 0:12:23.550 And so in that backdrop, they need to find funding elsewhere. 0:12:23.600 --> 0:12:26.010 And that's where NVIDIA has been willing to do it. 0:12:26.059 --> 0:12:29.620 So I agree more with Colette from NVIDIA that this 0:12:29.720 --> 0:12:33.400 is not your typical circular financing. This is more of 0:12:34.260 --> 0:12:38.520 a demand problem in terms of just not enough supply. 0:12:39.460 --> 0:12:43.580 And they're helping to make that match up better. 0:12:43.960 --> 0:12:46.380 CJ, the other part of this ecosystem that NVIDIA has 0:12:46.400 --> 0:12:50.420 been very supportive, be it vocally or with dollars, is 0:12:50.620 --> 0:12:53.460 open source modeling. They bought Hugging Face shortly after their 0:12:53.540 --> 0:12:56.929 earnings announcement. Jensen Huang himself has been out writing letters 0:12:57.330 --> 0:13:00.610 advocating for the ability of American companies to be able 0:13:00.650 --> 0:13:04.270 to use open sourcing. How important is it for NVIDIA 0:13:04.370 --> 0:13:07.550 and their continued growth that that access remains turned on, 0:13:07.590 --> 0:13:11.710 that there are options of cheaper models for corporate global 0:13:11.730 --> 0:13:12.870 corporations to use? 0:13:14.260 --> 0:13:16.730 Yeah, no, I think, you know, the worst case scenario 0:13:16.770 --> 0:13:19.349 for an NVIDIA would be the world or at least 0:13:19.370 --> 0:13:24.010 the United States and non-China, you know, gravitates to a 0:13:24.890 --> 0:13:27.290 open AI anthropic only frontier model world. 0:13:27.700 --> 0:13:28.030 Right. 0:13:28.429 --> 0:13:31.620 And that's where they get such great scale benefit that 0:13:31.660 --> 0:13:34.880 they will, you know, look to produce their own silicon 0:13:34.920 --> 0:13:37.040 and try to use less and less of NVIDIA. 0:13:37.840 --> 0:13:39.790 So if you're NVIDIA, what do you do? 0:13:39.840 --> 0:13:44.709 You go full stack, which means you're going to compete 0:13:44.750 --> 0:13:47.930 at every level of the AI cake that they've described. 0:13:47.970 --> 0:13:50.790 And as part of it, it is providing the models. 0:13:51.550 --> 0:13:54.650 If you look at their investments... with the Neo clouds, 0:13:55.190 --> 0:13:59.089 you could argue that, you know, NVIDIA is a virtual hyperscaler. 0:13:59.340 --> 0:14:02.619 They are absolutely competing with the likes of, you know, Amazon, 0:14:03.800 --> 0:14:06.860 you know, Google, et cetera. And so here again, on 0:14:06.900 --> 0:14:10.620 the model side, you know, they're going to actively work 0:14:10.660 --> 0:14:13.080 there and, you know, if they can kind of push 0:14:13.520 --> 0:14:17.800 their CUDA software and then these kind of smaller enterprise 0:14:18.320 --> 0:14:21.140 industry focused models, you you know, that's a real win 0:14:21.180 --> 0:14:23.900 for them in terms of, you know, driving, you know, 0:14:23.990 --> 0:14:28.210 adoption of their platform. And really, you know, going to 0:14:28.270 --> 0:14:31.270 first principles, that's what Jensen wants. He wants the world 0:14:31.430 --> 0:14:32.670 on NVIDIA's platform. 0:14:33.010 --> 0:14:35.450 CJ, before you go, Danny asked a version of this 0:14:35.510 --> 0:14:37.790 question just last week. I'd love your reaction to it. 0:14:38.190 --> 0:14:40.890 Is there a bit of a disconnect between the cost 0:14:40.930 --> 0:14:43.750 of building out the ecosystem right now and the loss 0:14:43.790 --> 0:14:46.160 of pricing power for the frontier models? 0:14:46.180 --> 0:14:51.030 It's a great question. Um, You know, I think. 0:14:52.700 --> 0:14:55.200 I think the truth today is that the frontier models 0:14:55.260 --> 0:15:00.100 probably do about 30 percent of token generation, yet earn 0:15:00.120 --> 0:15:03.330 about 90 percent of the economic profits. You know, I 0:15:03.390 --> 0:15:08.290 think true real intelligence will garner the lion's share of 0:15:08.310 --> 0:15:11.760 the economic profits. And so as They continue to push 0:15:11.800 --> 0:15:14.920 forward great models. You know, I think they'll continue to 0:15:14.960 --> 0:15:18.600 continue to do extraordinarily well. But there is absolutely a 0:15:18.660 --> 0:15:20.729 need for other models. 0:15:20.790 --> 0:15:21.030 Right. 0:15:21.090 --> 0:15:23.030 You know, if you're if you're just going to ask, 0:15:23.070 --> 0:15:26.210 you know, this year, where does Thanksgiving fall? 0:15:26.990 --> 0:15:28.480 You don't need a great model for that. 0:15:28.500 --> 0:15:30.460 And so in the future, you know, we're going to 0:15:30.640 --> 0:15:35.500 optimize that. to the question to each, you know, specific 0:15:35.540 --> 0:15:38.420 model that is best for that and offers the lowest 0:15:38.460 --> 0:15:40.800 cost per token. And so I absolutely think we're in 0:15:40.840 --> 0:15:44.040 a world of, you know, closed frontier plus open, you know, 0:15:44.080 --> 0:15:44.860 will be our future. 0:15:45.260 --> 0:15:48.080 CJ, final question, I promise. There's a near unhealthy obsession. 0:15:48.120 --> 0:15:49.979 Some people might say there is an unhealthy obsession on 0:15:50.000 --> 0:15:52.920 this program over monetary policy. And the prospect of higher 0:15:52.960 --> 0:15:55.280 interest rates, even on the radar of some of these 0:15:55.340 --> 0:15:58.240 companies investing in this build out and borrowing money in 0:15:58.260 --> 0:16:03.490 capital markets the way they are, You know, not I guess. 0:16:03.350 --> 0:16:07.650 Not typically with equity investors, but the companies, I would 0:16:07.670 --> 0:16:10.730 think 100 percent. I mean, in the last few weeks, 0:16:10.930 --> 0:16:14.630 we've seen CDS spreads push out for NVIDIA, Broadcom, others. 0:16:15.590 --> 0:16:19.980 Cost of financing is absolutely critical for, you know, investing 0:16:20.000 --> 0:16:22.800 in these data centers. That's part of the reason why 0:16:23.440 --> 0:16:29.960 NVIDIA created that $ 500 million platform with BlackRock, Blackstone, etc. Absolutely, 0:16:30.000 --> 0:16:35.020 cost of financing is incredibly important to the NPVs of 0:16:35.080 --> 0:16:38.780 these investments. If interest rates do continue to push higher, 0:16:39.880 --> 0:16:42.800 that will limit the magnitude of investments. 0:16:43.370 --> 0:16:44.370 I certainly hope they don't. 0:16:44.770 --> 0:16:48.230 Stay with us. More Bloomberg surveillance coming up after this. 0:16:57.530 --> 0:17:00.320 At Santa Cruz, President Donald Trump announcing a historic deal 0:17:00.360 --> 0:17:02.800 with Venezuela, giving the U.S. majority control of more than 0:17:02.800 --> 0:17:06.679 65 billion barrels of proven supplies in the country. Officials 0:17:06.720 --> 0:17:09.649 saying the agreement would create the world's second largest private 0:17:09.710 --> 0:17:12.910 oil company by reserves. Paul Sankey of Sankey Research calling 0:17:12.930 --> 0:17:15.939 the deal, quote, Easier said than done. Paul joins us 0:17:15.960 --> 0:17:17.440 now for more. Paul, good morning. Good to see you, buddy. 0:17:17.619 --> 0:17:20.460 I've got to say, I found it difficult to internalize 0:17:20.960 --> 0:17:23.120 this message coming from the White House. What is going 0:17:23.160 --> 0:17:24.270 on here? What do we know about it? 0:17:24.530 --> 0:17:26.330 Well, I think a lot of this is just fiddling 0:17:26.369 --> 0:17:28.190 while Rome burns. You know, what we've got here is 0:17:28.230 --> 0:17:30.930 a $ 4 gallon gasoline price and a $ 5. 0:17:30.690 --> 0:17:31.070 50 diesel. 0:17:32.690 --> 0:17:37.100 So you're up at $ 250 a barrel diesel, similarly, nearly 0:17:37.100 --> 0:17:40.520 200 gasoline a barrel. And we're talking about Venezuelan oil 0:17:40.540 --> 0:17:43.359 reserves when we really don't need Venezuelan oil reserves. And 0:17:43.780 --> 0:17:46.280 starting at the top, this is, for example, contrary to 0:17:46.300 --> 0:17:51.430 the Venezuelan constitution. These are really, really big issues to address. 0:17:51.850 --> 0:17:54.870 And an announcement between Delce Rodriguez and President Trump, which 0:17:54.910 --> 0:17:57.890 don't match, by the way, they're saying different things. You're 0:17:57.930 --> 0:18:00.169 just like, yeah, whatever. We don't need more crude at 0:18:00.210 --> 0:18:02.510 this point. That's not the issue. We don't really, in 0:18:02.550 --> 0:18:05.290 my view, need an SPR because we're in such a 0:18:05.330 --> 0:18:06.969 good position from the point of view of our own 0:18:07.030 --> 0:18:10.020 oil and gas industry. So, I just think it's, you know, 0:18:10.040 --> 0:18:11.389 a bit of a sideshow, quite frankly. 0:18:11.540 --> 0:18:12.500 You think we need refiners? 0:18:13.100 --> 0:18:15.550 Yeah, we're short refining and that's the big issue. And, 0:18:15.640 --> 0:18:18.860 you know, there's really almost nothing you can do about that. 0:18:19.800 --> 0:18:22.340 The headlines around Hormuz are similar. You know, what you're 0:18:22.359 --> 0:18:24.900 talking about here is 8 million, 10 million, 5 million 0:18:24.920 --> 0:18:27.200 barrels a day of crude. And the issue is whether 0:18:27.220 --> 0:18:28.980 or not we can get products out. And if you 0:18:29.000 --> 0:18:31.260 look at how much product is coming out of the Hormuz, 0:18:31.320 --> 0:18:35.189 it's actually not what increases when deliveries of oil increase 0:18:35.210 --> 0:18:38.410 through the Hormuz. So that doesn't address the issue either. 0:18:38.730 --> 0:18:40.189 And from the point of view of the stuff you 0:18:40.210 --> 0:18:43.570 guys just talking about, the wash, the interest rates, everything else, 0:18:44.010 --> 0:18:45.869 this diesel price is a colossal problem. 0:18:46.050 --> 0:18:46.910 You know, it really is. 0:18:46.950 --> 0:18:48.980 And I just don't hear anyone really talking about it 0:18:49.040 --> 0:18:51.420 in terms of making announcements about Venezuela. 0:18:51.440 --> 0:18:51.520 Yeah. 0:18:52.000 --> 0:18:54.260 What is industry? What role do they have to play? 0:18:54.400 --> 0:18:56.540 If you're an investor, if you're Chevron right now looking 0:18:56.580 --> 0:18:59.200 at the actions in Venezuela, how do you act? 0:18:59.859 --> 0:19:01.830 Well, I mean, this is good for Chevron, right? They 0:19:01.850 --> 0:19:05.210 never left Venezuela. They are the dominant U.S. 0:19:05.250 --> 0:19:05.550 Player. 0:19:05.570 --> 0:19:08.250 They don't have the legal dispute that you have with 0:19:08.310 --> 0:19:11.750 Exxon and Conoco against the Venezuelans. And so Chevron's in 0:19:11.810 --> 0:19:13.989 very good shape here. And you've seen their volumes from 0:19:14.010 --> 0:19:16.980 Venezuela grow rapidly. The other issue here, which makes it 0:19:17.500 --> 0:19:20.640 somewhat moot in terms of the impact on global oil markets, 0:19:20.720 --> 0:19:24.060 is basically every barrel of Venezuelan growth, which is growing 0:19:24.100 --> 0:19:27.439 quite rapidly in terms of incremental supply from where they 0:19:27.480 --> 0:19:28.300 were at the lows. 0:19:28.359 --> 0:19:28.980 So, you're probably at. 0:19:28.980 --> 0:19:31.990 About 1.2, 1.3 million barrels a day coming out of 0:19:32.010 --> 0:19:34.170 Venezuela right now. You can probably get to 2 million 0:19:34.190 --> 0:19:37.929 barrels a day. That essentially just competes with Canadian crude. So, 0:19:37.950 --> 0:19:41.570