WEBVTT - JPMorgan's Jay Barry on the Big Selloff in Bonds 0:00:10.119 --> 0:00:13.440 Hello, and welcome to another episode of the All Thoughts Podcast. 0:00:13.560 --> 0:00:14.880 I'm Tracy Alloway. 0:00:14.560 --> 0:00:16.400 And I'm Joe Whysnal So Joe. 0:00:16.440 --> 0:00:20.800 The big story in markets up until relatively recently has 0:00:20.880 --> 0:00:24.760 been the bond sell off, Like quite a dramatic sell 0:00:24.760 --> 0:00:26.720 off across the fixed income space. 0:00:26.880 --> 0:00:29.000 Yeah, really over like I guess this is the last 0:00:29.040 --> 0:00:32.560 couple months. So the yield today, well, we can sort 0:00:32.560 --> 0:00:35.680 of talk about the bond market up until today, but 0:00:35.800 --> 0:00:37.479 you know, we had been like in the threes and 0:00:37.600 --> 0:00:39.919 the fours and then we got us almost besides four 0:00:39.920 --> 0:00:42.800 point nine percent on the tenure, I've had a little 0:00:42.800 --> 0:00:45.600 bit of a pullback. We're recording this October tenth, but 0:00:45.720 --> 0:00:48.440 we are very high by any recent standard. 0:00:48.640 --> 0:00:50.519 Yeah, it is not lost on me that we are 0:00:50.560 --> 0:00:53.480 recording our bond sell off episode on the day when 0:00:53.520 --> 0:00:56.360 treasuries are recording their best one day performance since I 0:00:56.360 --> 0:00:59.480 think March of last year. But you know, we're trying, 0:00:59.520 --> 0:01:02.160 We're trying. But you're right. The recent bond market itself, 0:01:02.240 --> 0:01:04.080 it was one of those times when you see a 0:01:04.120 --> 0:01:07.679 lot of superlatives around, a lot of like highest yields 0:01:07.720 --> 0:01:10.440 since two thousand and seven, a lot of talk of 0:01:10.520 --> 0:01:14.200 standard deviation moves when we try to start calculating, you know, 0:01:14.240 --> 0:01:17.000 how many trading days it would normal trading days it 0:01:17.000 --> 0:01:20.080 would take to get moves of this size. But the 0:01:20.240 --> 0:01:24.120 interesting thing about this whole dynamic is it's not really 0:01:24.360 --> 0:01:27.959 clear what the proximate cause of the sell off is. 0:01:28.360 --> 0:01:30.800 So you have a lot of people blaming you know, 0:01:30.840 --> 0:01:33.920 the recent FOMC meeting when they release the dot plot 0:01:34.040 --> 0:01:37.880 showing higher for longer. You have some people blaming oil. 0:01:38.080 --> 0:01:41.440 You have people talking about supply. Some people are looking 0:01:41.480 --> 0:01:44.640 at more technical aspects. You have some people blaming the 0:01:44.720 --> 0:01:48.240 term premium, which I find absolutely hilarious because to me, 0:01:48.360 --> 0:01:51.200 a higher term premium is like a symptom of the 0:01:51.240 --> 0:01:54.400 sell off, not the actual driver of it. But anyway, 0:01:54.440 --> 0:01:57.200 all of this means we really need to try to 0:01:57.480 --> 0:02:00.640 get into the weeds of what is happening here and 0:02:00.680 --> 0:02:02.320 what it might mean for the wider market. 0:02:02.560 --> 0:02:05.880 Right because even if okay, maybe last Friday, right after 0:02:05.920 --> 0:02:08.120 their jobs report, when we got that very brief spike, 0:02:08.160 --> 0:02:10.200 maybe that was the peak. Who knows. But I do 0:02:10.240 --> 0:02:12.960 think it's important to try to disambiguate these things because, 0:02:13.200 --> 0:02:15.000 as you said, like I guess it's one of those 0:02:15.000 --> 0:02:17.520 things when there's a lot of explanations for something, none 0:02:17.520 --> 0:02:20.000 of them are like very satisfying, Like a lot of 0:02:20.080 --> 0:02:23.000 deficit talk lately, but it's like what, like, Yeah, what 0:02:23.160 --> 0:02:24.960 did we all do? I think we knew that the 0:02:25.000 --> 0:02:27.880 deficit was really big for a long time. Is that 0:02:27.960 --> 0:02:30.480 recent FED meeting? But again, okay, like the twenty twenty 0:02:30.520 --> 0:02:32.280 four dods came up a little bit, but does that 0:02:32.320 --> 0:02:36.000 really outrageous four point nine percent on the tenure? Nothing 0:02:36.080 --> 0:02:40.000 is quite satisfying as to what happened and why we're 0:02:40.040 --> 0:02:41.160 at this sort of like new. 0:02:41.080 --> 0:02:44.720 Level exactly well put. So today I am very pleased 0:02:44.720 --> 0:02:46.840 to say we do, in fact have the perfect guest. 0:02:46.919 --> 0:02:49.440 We're going to be speaking to a bond analyst. I've 0:02:49.440 --> 0:02:52.040 been a fan of his research for a long time. 0:02:52.120 --> 0:02:54.120 He used to work with one of our other favorite 0:02:54.120 --> 0:02:57.040 All Thoughts guests, Josh Younger. We're going to be speaking 0:02:57.120 --> 0:02:59.760 with Jay Berry. He is the co head of US 0:02:59.800 --> 0:03:02.720 and stret strategy at JP Morgan. We're going to be 0:03:02.720 --> 0:03:05.600 trying to get a handle on what might be driving 0:03:05.880 --> 0:03:08.840 the moves in bonds, but also how you go about 0:03:09.080 --> 0:03:12.880 as an analyst trying to disaggregate all these different factors. 0:03:12.880 --> 0:03:15.400 So I'm very excited. Jay, Thank you so much for 0:03:15.440 --> 0:03:16.360 coming on our thoughts. 0:03:16.520 --> 0:03:18.919 Tracy, thanks so much for having me really appreciat being here. 0:03:19.400 --> 0:03:21.520 So I guess maybe just to begin with, I mean 0:03:21.960 --> 0:03:25.520 the simple question what is driving the sella, We'll start 0:03:25.560 --> 0:03:28.200 there and then we can try to dig into different pieces. 0:03:28.320 --> 0:03:30.000 Well, I mean, Joj just said there's a whole host 0:03:30.000 --> 0:03:32.320 of reasons that have been bantered about and that maybe 0:03:32.320 --> 0:03:34.280 none of them are all satisfying. But I really think 0:03:34.280 --> 0:03:36.880 it's the confluence of a number of different factors here. 0:03:37.520 --> 0:03:39.880 So I mean an aggregate, Now we're off the highs 0:03:39.920 --> 0:03:41.560 and yield, as you said, but we're still about one 0:03:41.600 --> 0:03:43.400 hundred basis points higher than we were in the middle 0:03:43.440 --> 0:03:46.160 of the summer. And to me, it's like, what's changed 0:03:46.280 --> 0:03:49.040 over that period? And I think the early part of 0:03:49.040 --> 0:03:51.160 the move you could say was definitively a story about 0:03:51.160 --> 0:03:53.600 the US economy and fundamentals, because I look back at 0:03:53.600 --> 0:03:56.119 where we were very early in the summer and where 0:03:56.120 --> 0:03:59.119 we are right now, and over that period, we increased 0:03:59.120 --> 0:04:01.960 our second half growth expectations at JP Morgan by about 0:04:01.960 --> 0:04:04.760 two and a half percentage points from like the mildest 0:04:04.800 --> 0:04:06.600 of recessions. If you can call it that to looking 0:04:06.600 --> 0:04:08.680 for above trend growth the second half of this year, 0:04:09.120 --> 0:04:11.320 and that's like a meaningful driver I think of the 0:04:11.320 --> 0:04:13.480 move and long term rates because it's given us more 0:04:13.520 --> 0:04:15.680 confidence that the recession may be a bit further off. 0:04:15.720 --> 0:04:18.640 But it's also helped sort of anchor FED expectations at 0:04:18.720 --> 0:04:21.240 higher levels as well. So, you know, we haven't really 0:04:21.279 --> 0:04:23.400 changed the expected peak in the FED funds rate since 0:04:23.400 --> 0:04:25.160 the early part of the summer, but we pushed out 0:04:25.160 --> 0:04:27.839 the timing and we've pushed it out from you know, 0:04:27.880 --> 0:04:29.799 at that time thinking it was going to happen around now, 0:04:30.080 --> 0:04:31.599 to sort of pushing it out to the early part 0:04:31.600 --> 0:04:34.359 of next year. But then the more powerful influence is 0:04:34.360 --> 0:04:38.640 that we've disinverted that money market curb a bit, and 0:04:38.680 --> 0:04:41.000 that you know, early in the summer, we were pricing 0:04:41.040 --> 0:04:43.359 about one hundred and fifty bases points of cuts for 0:04:43.440 --> 0:04:45.719 twenty twenty four. Now we're pricing in still cuts, but 0:04:45.760 --> 0:04:48.440 we've taken out about forty percent of that. So I 0:04:48.440 --> 0:04:51.160 think the first part was fundamental, but then I think 0:04:51.160 --> 0:04:54.200 the drivers began to shift, and more recently since September, 0:04:54.600 --> 0:04:57.279 you know, FED and growth expectations have been pretty stable, 0:04:57.640 --> 0:04:59.800 but inflation expectation has been rising, and we like to 0:05:00.120 --> 0:05:02.640 get five year ahead, five year inflation expectations from the 0:05:02.640 --> 0:05:06.000 tips market, from the inflation link treasury market, and those 0:05:06.040 --> 0:05:08.680 break even expectations have gone higher by about twenty five 0:05:08.720 --> 0:05:10.400 basis points over the past six weeks. 0:05:10.480 --> 0:05:13.440 Wait, this is the five year five year forward break even? 0:05:13.560 --> 0:05:15.320 Is that the exactly the thing that you like to 0:05:15.320 --> 0:05:18.600 look at. And I'll bring that up on my terminal 0:05:18.880 --> 0:05:22.719 while you're talking about it, So I understand, Okay, growth 0:05:22.720 --> 0:05:25.800 has picked up or versus expectation, certainly compared to the 0:05:25.800 --> 0:05:27.800 beginning of the year, when, as you say, like almost 0:05:27.800 --> 0:05:30.800 everyone expected recession. Now we're at above trend growth. Maybe 0:05:30.800 --> 0:05:33.680 it don't moderate, et cetera. What is the link between 0:05:33.880 --> 0:05:36.520 that and thirty year yields or the long end of 0:05:36.560 --> 0:05:38.360 the curve, Because of course I get it the short end. 0:05:38.400 --> 0:05:41.320 It's very easy to draw a very bright line between 0:05:41.680 --> 0:05:44.279 short term rate expectations and the two year bond or whatever. 0:05:44.360 --> 0:05:46.440 But then when you get further out, it's like, Okay, 0:05:46.440 --> 0:05:48.919 why does the growth picking up in Q four of 0:05:48.960 --> 0:05:51.440 twenty twenty three effect that people are going to pay 0:05:51.480 --> 0:05:53.719 for a bond that matures in the twenty fifties or whatever. 0:05:53.920 --> 0:05:56.359 No, I think there's still some sensitivity of policy that 0:05:56.440 --> 0:05:59.680 sort of reverberates through the term structure, and the sensitivity 0:05:59.680 --> 0:06:02.240 may decline the further act to go, but it's still there. 0:06:02.320 --> 0:06:05.320 So just for a number, you know, every one hundred 0:06:05.360 --> 0:06:08.719 basis point change in FED policy expectations three to six 0:06:08.760 --> 0:06:11.800 month forward tends to sort of change long term meals 0:06:11.800 --> 0:06:14.919 by about forty to forty five basis points. And similarly, 0:06:14.960 --> 0:06:17.960 like every change in year ahead growth expectations tends to 0:06:17.960 --> 0:06:20.240 move long term meals by five to ten basis points. 0:06:20.279 --> 0:06:22.720 So yeah, the further you go out the term structure, 0:06:22.960 --> 0:06:26.120 the more I think idiosyncratic it becomes. But there still 0:06:26.160 --> 0:06:29.200 is sensitivity to what's happening to the underlying economy and 0:06:29.360 --> 0:06:31.760 what's happening with FED policy. So I think it's still there, 0:06:32.200 --> 0:06:34.680 just with a lower sensitivity than at the very short end. 0:06:34.920 --> 0:06:38.520 How do you actually go about, as a strategist sort 0:06:38.520 --> 0:06:43.320 of decomposing the different moves into different factors or drivers. 0:06:43.360 --> 0:06:45.880 And I know that you have at JP Morgan, for instance, 0:06:46.160 --> 0:06:50.440 this fair value model of where you think Treasury should 0:06:50.480 --> 0:06:52.360 be trading. And I think last week you were saying 0:06:52.400 --> 0:06:55.960 that you thought that yields had overshot fair value, which 0:06:56.160 --> 0:07:00.000 you know, I'm guessing you didn't expect what happened over 0:07:00.120 --> 0:07:02.920 the weekend in Israel to actually happen, But it seems 0:07:02.960 --> 0:07:05.479 to have been quite precient because we have seen yields 0:07:05.520 --> 0:07:07.520 come down a little bit. But how are you actually 0:07:07.600 --> 0:07:10.640 analyzing the different drivers of a move in rates? 0:07:11.040 --> 0:07:13.560 Yeah, so you've talked about the fair value model. I 0:07:13.560 --> 0:07:15.400 think that's a key input to what we do because 0:07:15.440 --> 0:07:18.880 we try to identify empirically what have been the largest 0:07:18.920 --> 0:07:21.440 drivers of yields over time, and we can look back 0:07:21.480 --> 0:07:24.240 over windows of five or ten or fifteen years, and 0:07:24.280 --> 0:07:26.040 we've got a host of factors that are sort of 0:07:26.080 --> 0:07:30.760 always consistently in the framework. And we talked about FED policy, growth, 0:07:30.760 --> 0:07:34.520 and inflation expectations being the three key drivers, like the triumvirate, 0:07:34.600 --> 0:07:37.280 so to speak, and then other factors which at various 0:07:37.320 --> 0:07:39.480 points over the last you know, ten to fifteen years, 0:07:39.480 --> 0:07:41.960 have been important and less important. I mean, there was 0:07:42.000 --> 0:07:45.520 a time when with policy rates at zero and negative 0:07:45.600 --> 0:07:48.360 territory globally, we had the share of the bond universe 0:07:48.400 --> 0:07:51.280 that was trading it a negative yield globally because policy 0:07:51.320 --> 0:07:54.120 rates being anchored at very low levels helped anchor US 0:07:54.200 --> 0:07:57.160 rates lower and that was important but less important right now. 0:07:57.480 --> 0:07:59.200 So that starts, and you know, if we have a 0:07:59.400 --> 0:08:02.320 sort of center universe about where we think fed policy, 0:08:02.360 --> 0:08:04.920 growth and inflation are headed, that's a starting point. And 0:08:04.920 --> 0:08:07.160 you're right, like when we adjusted for those factors, there 0:08:07.200 --> 0:08:09.000 was a point last week where it looked like we 0:08:09.000 --> 0:08:11.960 were trading about thirty five or forty basis points too high, 0:08:11.960 --> 0:08:14.280 where you know, you're talking about standard deviations. Before that 0:08:14.480 --> 0:08:17.080 was something like a two and a half standard deviation 0:08:17.240 --> 0:08:19.800 move relative to fair value in our framework, and one 0:08:19.840 --> 0:08:22.320 which we hadn't seen since this time last fall, after 0:08:22.360 --> 0:08:23.440 the uk LDI. 0:08:23.240 --> 0:08:26.360 Crist Right, which was the other big bond sell off. 0:08:26.640 --> 0:08:28.640 And really since the spring of twenty twenty two. So 0:08:28.720 --> 0:08:30.840 I think, you know, we take notice of that because 0:08:30.880 --> 0:08:32.760 it's hard to say, you know, I think we've all 0:08:32.760 --> 0:08:34.440 been pretty humble to the fact that the economy has 0:08:34.480 --> 0:08:36.320 been resilient. It's been tough to call, but we have 0:08:36.400 --> 0:08:38.360 to have something that sort of centers like where should 0:08:38.440 --> 0:08:40.400 rates be and how far have they gone? And this 0:08:40.520 --> 0:08:42.040 was at least a flag to us that they'd gone 0:08:42.040 --> 0:08:42.440 too far. 0:08:43.559 --> 0:08:47.160 You mentioned sort of factors driving the US Treasury price 0:08:47.200 --> 0:08:49.920 and how those have evolved over time. Can you spell 0:08:49.920 --> 0:08:51.680 that out in a little bit more detail, Like I 0:08:51.679 --> 0:08:54.480 would be very curious to hear more about what is 0:08:54.559 --> 0:08:59.000 driving moves now versus say, maybe pre twenty twenty. 0:08:59.640 --> 0:09:02.040 Yeah, I mean, I think you know pre twenty twenty. 0:09:02.040 --> 0:09:04.400 We just briefly talked about one of those factors, low 0:09:04.440 --> 0:09:07.880 and negative policy rates elsewhere meant that even as the 0:09:08.000 --> 0:09:11.840 US was increasing rates during the twenty fifteen through twenty 0:09:11.880 --> 0:09:14.120 eighteen cycle, that was something that helped anchor long term 0:09:14.200 --> 0:09:16.720 yields at lower levels. And you could see the influence 0:09:16.760 --> 0:09:19.840 there when you talk about the hedge dyel pickup for 0:09:19.960 --> 0:09:23.280 US bonds for versus most foreign currency pairs, and that 0:09:23.320 --> 0:09:26.280 has since of course eroded because basically ever made every 0:09:26.320 --> 0:09:29.600 major develop market central bank has been increasing policy rates 0:09:29.600 --> 0:09:31.679 at a rate. And the last of which that's out 0:09:31.679 --> 0:09:33.840 there the Bank of Japan, we think at some point 0:09:33.880 --> 0:09:37.280 will completely lose its YCC band and well at some 0:09:37.280 --> 0:09:40.640 point exit negative interest rate policy next year. So that 0:09:40.679 --> 0:09:43.480 was an important factor which we now don't have as 0:09:43.480 --> 0:09:44.480 a factor in the model. 0:09:45.320 --> 0:09:47.600 The Bank of Japan. It was like about a month 0:09:47.640 --> 0:09:49.400 and a half ago or two months ago, like sometime 0:09:49.440 --> 0:09:52.120 in August. I forget that because I have to admit, 0:09:52.200 --> 0:09:54.800 like all these Bank of Japan had lines about whether 0:09:54.800 --> 0:09:56.520 they're going to like keep the ten years er or whatever, 0:09:56.520 --> 0:09:57.840 like they all sort of blur. But there was like 0:09:57.880 --> 0:10:00.439 some news that happened. It was something the Bank of 0:10:00.520 --> 0:10:02.439 Japan that like changed the entire tenor of the market 0:10:02.440 --> 0:10:04.360 all at once. Can you remind me of what I'm 0:10:04.400 --> 0:10:05.280 like forgetting here. 0:10:05.320 --> 0:10:08.800 Yeah, Joe. In late July, the Bank of Japan basically 0:10:08.840 --> 0:10:12.280 allowed jgbs in the tenure sector to try trade even 0:10:12.320 --> 0:10:16.360 wider around it's sort of plus or minus fifty basis 0:10:16.360 --> 0:10:18.840 point target and effectively kind of gave you notice that 0:10:18.880 --> 0:10:22.640 at some point it was getting closer to completely removing 0:10:22.720 --> 0:10:23.440 that Y series. 0:10:23.480 --> 0:10:26.000 They're having their highest inflation in years too. 0:10:25.840 --> 0:10:28.760 Right exactly, Okay, exactly, So, I mean I think on 0:10:28.800 --> 0:10:30.880 a partial basis, you can argue that was a catalyst 0:10:30.880 --> 0:10:31.240 as well. 0:10:31.840 --> 0:10:34.439 Can you talk about when you talk about these foreign buyers, 0:10:34.800 --> 0:10:36.720 as you say this sort of like price in sensitive 0:10:37.000 --> 0:10:40.600 demand from a foreign sector, the disappearance of these price 0:10:40.640 --> 0:10:43.680 in sensitive foreign buyers, how much does that affect rates 0:10:43.800 --> 0:10:46.640 versus say, just like rates volatility. 0:10:46.400 --> 0:10:48.440 By the way. This is why I really wanted to 0:10:48.440 --> 0:10:51.440 get Jay on for this podcast, because he wrote a 0:10:51.480 --> 0:10:55.000 great note about a year ago basically talking about the 0:10:55.040 --> 0:10:57.760 retreat of price and sensitive buyers in the form of 0:10:58.080 --> 0:11:00.480 foreign central banks and the FED as well as it 0:11:00.559 --> 0:11:02.680 was winding down its balance sheet. I think I wrote 0:11:02.679 --> 0:11:05.200 it up under the headline JP Morgan is worried about 0:11:05.240 --> 0:11:07.240 who's going to buy all the bonds and at the 0:11:07.280 --> 0:11:11.040 time it got a lot of pushback. But fast forward 0:11:11.080 --> 0:11:14.600 a year and again it seems like you were broadly 0:11:14.679 --> 0:11:16.000 directionally correct. 0:11:16.960 --> 0:11:19.199 No, I think it's an important driver over a longer 0:11:19.240 --> 0:11:21.160 period of time. Like it's hard to say, and Tracy, 0:11:21.160 --> 0:11:23.559 you said this before, like whether it's the proximate cause 0:11:23.640 --> 0:11:26.200 of the selloff, but I think in the background, it's 0:11:26.200 --> 0:11:29.240 something that's contributing to what's going on because to us 0:11:29.280 --> 0:11:31.839 at JP Morgan, you know, we look at three sets 0:11:31.880 --> 0:11:34.960 of buyers who have been the main price in sensitive 0:11:34.960 --> 0:11:36.680 sources of demand for the better part of the past 0:11:36.679 --> 0:11:39.360 two decades at various points. And you talk about the 0:11:39.400 --> 0:11:42.520 foreign demand story, I mean, we know that FX reserves 0:11:42.520 --> 0:11:44.960 peaked about seven or eight years ago. The dollars share 0:11:44.960 --> 0:11:47.120 of those reserves have been coming down. But there's a 0:11:47.160 --> 0:11:49.280 point in time at the beginning of the century where 0:11:49.559 --> 0:11:52.240 FX reserves were growing so rapidly and the share of 0:11:52.240 --> 0:11:54.360 those reserves held in dollars were so rapidly that the 0:11:54.679 --> 0:11:56.760 deposit of that savings into the US, I think was 0:11:56.800 --> 0:11:59.120 something that kept long term rates low. And your remember 0:11:59.200 --> 0:12:01.839 chair green Span talking about this and the conundrum in 0:12:01.880 --> 0:12:04.160 two thousand and five about why long term rates were 0:12:04.160 --> 0:12:07.240 not rising even as the FED was tightening. So that's 0:12:07.280 --> 0:12:09.560 a key driver right there, and we look at it. 0:12:09.920 --> 0:12:13.200 FX reserves we don't really expect to grow appreciably from here, 0:12:13.360 --> 0:12:16.160 and we're not in the d dollarization camp, but the 0:12:16.240 --> 0:12:18.240 dollar share of those reserves have been on a downward 0:12:18.280 --> 0:12:21.600 trend as central banks globally have been diversifying. So it's 0:12:21.640 --> 0:12:23.760 just saying that if that was a tailwind for rates 0:12:23.840 --> 0:12:25.360 for the better part of the first half of this, 0:12:25.679 --> 0:12:29.280 you know, last twenty years, it's not there the second one. 0:12:29.640 --> 0:12:32.160 And this was more local or the US banks where 0:12:32.280 --> 0:12:34.480 they bought about three quarters of a trillion of treasuries 0:12:34.520 --> 0:12:37.640 over twenty twenty and twenty twenty one when supply was heavy, 0:12:38.160 --> 0:12:40.679 largely due to the fact that deposit growth at strip 0:12:40.760 --> 0:12:43.440 loan growth. And now we know deposits have stopped coming 0:12:43.480 --> 0:12:45.400 down like they did in the spring, but they're not growing. 0:12:45.679 --> 0:12:47.680 And I think one would think even as deposit growth 0:12:47.720 --> 0:12:51.079 picks up, that banks, after what's happened here, might generally 0:12:51.120 --> 0:12:53.800 speaking bias their purchase is shorter along the yield curve, 0:12:53.800 --> 0:12:56.880 which with less duration risk. And then the final piece 0:12:56.880 --> 0:12:58.080 of the puzzle is the FED. And I think we 0:12:58.160 --> 0:13:00.000 lose this that even though we're coming to the end 0:13:00.120 --> 0:13:02.200 to the FED policy rate tightening cycle, or we think 0:13:02.200 --> 0:13:05.800 it's actually concluded, balance sheet policies operating in the background, 0:13:05.840 --> 0:13:08.200 And just as the boj JO was really important at 0:13:08.240 --> 0:13:10.600 the end of July, I think chair pals comments at 0:13:10.600 --> 0:13:13.720 the July press conference were is equally important because a 0:13:13.760 --> 0:13:16.760 reporter asked him about whether the FED could continue to 0:13:16.800 --> 0:13:19.920 do QT while it actually lowers interest rates as inflation 0:13:20.000 --> 0:13:22.160 comes down next year, and he made the point that 0:13:22.200 --> 0:13:25.359 you'd be normalizing both the policy levers may be in opposition, 0:13:25.720 --> 0:13:28.360 but you're normalizing the balance sheet as you're normalizing rates. 0:13:28.720 --> 0:13:31.800 And I think the extended runway for QT matters because 0:13:32.200 --> 0:13:34.760 we found over a longer period of time the FED 0:13:34.840 --> 0:13:38.240 stock of holdings matters for yield levels and as that 0:13:38.360 --> 0:13:41.000 unwind that should slowly keep long term rates anchored at 0:13:41.040 --> 0:13:42.600 higher levels and re steep and yielders. 0:13:59.360 --> 0:14:02.840 So you have these longer term factors I guess happening 0:14:02.920 --> 0:14:06.000 in the background the retreat of these three groups of 0:14:06.600 --> 0:14:10.240 price insensitive buyers. Just going back to some of the 0:14:10.400 --> 0:14:13.839 short term catalysts here. You know, you mentioned term premium 0:14:14.240 --> 0:14:16.880 in the intro. Term premium is one of those concepts. 0:14:16.920 --> 0:14:19.200 I feel like it gets bandied around quite a lot. 0:14:19.440 --> 0:14:22.680 Not everyone quite understands what it means, but also there 0:14:22.760 --> 0:14:24.200 is no consensus, like. 0:14:24.200 --> 0:14:27.200 Myself, but you're not. I'm about to learn something that 0:14:27.200 --> 0:14:27.960 I've always been want to. 0:14:28.080 --> 0:14:29.960 But see you're missing out, Joe, because you should just 0:14:30.000 --> 0:14:32.960 start saying term premium. Just blame everything on the term 0:14:32.960 --> 0:14:35.080 premium and just use it as a scapegoat for any 0:14:35.280 --> 0:14:37.640 like any move that you don't agree with or that 0:14:37.720 --> 0:14:39.520 you are on the wrong side of. That's how most 0:14:39.520 --> 0:14:42.240 people seem to use it. But okay, maybe just to 0:14:42.280 --> 0:14:45.040 begin with, what is the term premium and what has 0:14:45.160 --> 0:14:47.000 happened to it in recent weeks? 0:14:47.240 --> 0:14:49.320 Yeah, so the term premium in you're right, is nebulous 0:14:49.440 --> 0:14:52.520 term I think is the extra compensation required for investors 0:14:52.520 --> 0:14:54.200 to buy longer duration assets. 0:14:54.840 --> 0:14:57.120 That is a perfect definition that I'm pretty sure I 0:14:57.160 --> 0:14:59.200 have used in my own copy before. 0:15:00.120 --> 0:15:04.120 Just made sure I googled it before. But I think 0:15:04.160 --> 0:15:05.680 there's a number of ways to look at it, and 0:15:05.680 --> 0:15:07.400 I think we can start and just talk about our 0:15:07.440 --> 0:15:09.560 fair value framework, and I think we can say everything 0:15:09.600 --> 0:15:13.320 outside of these fundamental drivers, one might possibly attribute to 0:15:13.400 --> 0:15:16.160 term premium that in the absence of being able to 0:15:16.200 --> 0:15:19.760 explain with growth and inflation and FED policy expectations, that 0:15:19.760 --> 0:15:22.760 that's a driver. There. There's also a series of very 0:15:22.760 --> 0:15:25.880 widely watched academic models. There's the ACM model from the 0:15:25.880 --> 0:15:28.960 New York Fed, the kim Wright model that the Federal 0:15:28.960 --> 0:15:31.720 Reserve Board in DC watches, which are a series of 0:15:31.800 --> 0:15:34.680 I think no arbitrage term structure models which are kind 0:15:34.680 --> 0:15:37.880 of mean reverting in nature, and those were sitting relatively 0:15:37.880 --> 0:15:40.720 low until recently, and they actually would attribute most of 0:15:40.720 --> 0:15:43.280 the sell off over the past six weeks to term premium. 0:15:43.720 --> 0:15:45.120 I think we've done a paper on this, and we 0:15:45.160 --> 0:15:47.320 think that there's some idiosyncrasies with the way that these 0:15:47.520 --> 0:15:50.520 models are constructed because they are sampled over a period 0:15:50.560 --> 0:15:52.720 of declining rates that they attribute a lot more to 0:15:52.880 --> 0:15:56.400 changes in policy expectations than term premium. So it turns 0:15:56.440 --> 0:15:58.720 they can be less sort of influential or less I 0:15:58.720 --> 0:16:02.320 think insightful. They're still very valuable, just less inciseful at 0:16:02.360 --> 0:16:05.160 these turns. And then there's finally more empirical ways to 0:16:05.160 --> 0:16:07.400 measure it too, because there's survey based measures of where 0:16:07.760 --> 0:16:12.120 economists expect policy rates to be, like the survey professional forecasters, yeah, 0:16:12.160 --> 0:16:14.240 and the survey of primary dealers, where you can observe 0:16:14.280 --> 0:16:16.640 where economists think policy rates will be over the next 0:16:16.640 --> 0:16:18.160 five to ten years, and you can compare them to 0:16:18.160 --> 0:16:20.080 ten year rates to get a sense of the extra 0:16:20.120 --> 0:16:21.400 compensation that's required. 0:16:21.880 --> 0:16:23.680 Right, So this is the key, right, because the basic 0:16:23.760 --> 0:16:26.200 idea is a long rate is just a series of 0:16:26.240 --> 0:16:29.320 overnight rates, and so that gap, if you have some 0:16:29.560 --> 0:16:31.880 estimate of where the overnight rates are going to be 0:16:31.920 --> 0:16:33.360 over the next ten years, then you look at the 0:16:33.400 --> 0:16:36.440 yeld them. Theoretically, that gap is the turn. 0:16:36.320 --> 0:16:36.920 The term premier. 0:16:37.120 --> 0:16:39.920 Exactly Can you talk about sentiment? And you know that 0:16:40.080 --> 0:16:43.520 strikes me tracy that we were recording this basically literally 0:16:43.600 --> 0:16:46.240 a month after we interviewed the bond King Bill grows, 0:16:46.640 --> 0:16:49.400 in which he said he doesn't own any bonds, and 0:16:49.440 --> 0:16:51.920 so it really strikes me that is like, well, people 0:16:52.040 --> 0:16:54.800 hate bonds right now, like people really hate them, and 0:16:54.840 --> 0:16:57.680 everyone like in that month since we talked about first 0:16:57.680 --> 0:16:59.880 of all, when we had that conversation, the ten years 0:17:00.080 --> 0:17:02.920 closer to four point two percent, So very timely call. 0:17:03.000 --> 0:17:05.840 But you know, in that month, I do not recall 0:17:05.880 --> 0:17:08.600 there's so much talk about the deficit. There's so much 0:17:08.640 --> 0:17:12.000 talk about what at all of these dates, higher inflation 0:17:12.119 --> 0:17:14.960 higher for longer we can't get it under control? Like 0:17:16.080 --> 0:17:18.960 how do you measure sentiment and how much does that 0:17:19.080 --> 0:17:20.200 drive some of these moves. 0:17:20.320 --> 0:17:21.960 I'm glad you asked that, Joe, because I think it 0:17:21.960 --> 0:17:25.080 can be really influential over shorter periods of time, say 0:17:25.080 --> 0:17:27.560 four to six weeks. So there's a host of metrics 0:17:27.600 --> 0:17:31.200 that we like to watch from the CFTC's data on 0:17:31.560 --> 0:17:34.439 sort of speculative positioning in interest rate futures, some more 0:17:34.440 --> 0:17:37.359 empirical models that sort of track the performance of hedge 0:17:37.359 --> 0:17:40.520 funds and asset managers. But my favorite ends very close 0:17:40.520 --> 0:17:42.120 to my heart because it's been something I've been working 0:17:42.119 --> 0:17:44.480 with for like more than two decades, is our weekly 0:17:44.560 --> 0:17:46.840 JP Morgan Treasury Client Survey. It's a bit of a 0:17:46.880 --> 0:17:49.960 misnomer because it's really our duration survey of the aggregate 0:17:49.960 --> 0:17:53.440 exposure of our rates franchise, and every week we ask 0:17:53.520 --> 0:17:56.320 the same number of clients in our franchise whether they're long, 0:17:56.480 --> 0:18:00.320 neutral or short duration, either outright or relative to bench mark. 0:18:00.359 --> 0:18:04.320 And we've found that when that measure sort of moves 0:18:04.440 --> 0:18:07.119 very sharply away from average levels, it can have a 0:18:07.200 --> 0:18:09.520 mean revert effect on yield the opposite direction. So you 0:18:09.520 --> 0:18:12.240 talk about sentiment, I think everyone through the spring and 0:18:12.280 --> 0:18:14.439 summer was trying to handicap when the FED would be 0:18:14.440 --> 0:18:18.320 done raising rates, thinking the next move is going on hold, 0:18:18.359 --> 0:18:21.280 which would be the precursor to rates moving lower. And 0:18:21.359 --> 0:18:23.880 our survey back in July and August was as long 0:18:23.920 --> 0:18:25.919 as it had been in over a decade, and it 0:18:25.960 --> 0:18:27.760 gave you a signal that over the next five to 0:18:27.840 --> 0:18:29.919