The market’s next test: Rates, AI, and elections

BRIAN HESS:

And hello everyone. Thanks for joining us and welcome to the Natixis quarterly macro webinar. I'm Brian Hess, a portfolio manager on our multi-asset team. And with me today are Jack Janaseiwicz and Garrett Melson. Jack is multi-asset portfolio manager and lead portfolio strategist and member of the Natixis Investment Committee.

We have a good lineup today of topics. So we're going to cover the AI CapEx cycle, the outlook for US growth, volatility in global bond markets, recent central bank rate hikes, oil, inflation, and a look at US midterm elections. So we have a nice varied topic of items. It's a-- quite an interesting time to be investing.

First up, though, is AI. Everyone's talking about the AI trade. We witnessed the melt-up in AI stock prices back in June-- or during the second quarter, really, we had a melt-up in prices. And there was a blow-off top in June. Since then, AI stocks have been in an orderly consolidation, for the most part. But while the stock prices took a breather over the summer, the AI CapEx cycle most certainly did not. And Jack, you have some interesting charts on AI investment trends and the race towards profitability. I thought we could start there. What are you seeing on that front?

JACK JANASIEWICZ: Yeah. Thanks, Brian. And this is, I think, central to the thesis of the concerns around whether AI is a longer-term trend or if it's something that's more of a bubble that's looking to pop.

And so we've been talking about this quite extensively in our podcasts, as well as in some of these previous webinars, and as well as some of the conversations we're having with our clients-- so probably worth giving a little bit of an update. We're getting a little bit more clarity on this front. And so I think it's worth walking through some of these.

This first slide I want to share right here is just-- is the simple story of just looking at the big-picture concern here. You're looking at forecasted CapEx spend for what we call the hyperscalers-- so the combination of Microsoft, Google, Amazon, Meta, and Oracle. And in that dotted line moving up on the right side, you're looking at the CapEx spend as a percentage of GDP, just to give you a sense just how big some of these numbers are.

So right now through 2026, expectations for total CapEx right now for these five names-- a little bit under $800 billion. You can see that continues to grow.

As we push out closer to 2030, expectations are north of $1.2 trillion. And that moves from something closer to almost 5% of GDP-- so some staggering numbers. And again, the question for investors continues to be, will this start to manifest in some sort of payoff in that ROI question?

So a couple of charts here to walk through to give us some-- I think a glimpse into what we're seeing regarding the potential payoff structure, if you will-- in this first one, we're just simply looking at, well-- is a proxy for demand. We're looking at these order backlogs. And we're just comparing the growth rate in those backlogs relative to the growth rate in CapEx spend.

And you can see first quarter and second quarter backlog for these hyperscalers really starting to take off. And when we think about backlog, a couple things to worth noting here-- these are multiyear customer commitments where these commitments are basically all related to infrastructure, AI infrastructure investment.

And so Microsoft, for example, has told us that their incremental demand is now coming from non-frontier AI customers-- so a little bit more of a diversified base. Google is talking about gaining from robust enterprise adoption and, really, its internal TPUs that they're pushing towards. And Amazon is saying that their increases are coming from enterprise clients locking in those scarce compute resources.

So again, we're seeing signs that demand is starting to pick up. And if you take that one step further, if we look at cloud revenue growth and compare that to CapEx growth-- seeing the same thing. We're starting to see the monetization of those cloud revenues, which for a while, there was a concern going forward that people like Meta, for example, lending out or leasing out that extra capacity that they have on the cloud front-- a lot of concerns that maybe that was a sign of excess capacity.

And the flip side of that is simply these guys are trying to monetize what they have and keep it in spare reserves, should they need it. But the point being here is as we move out towards the beginning of 2027, you see those growth rates starting to really converge. And I think that's a key point going forward.

But again, the bottom line here is the gap between that CapEx spend and that cloud revenue and revenue in general. So one of the things I'm trying to show here is simply, yeah, they're both increasing that CapEx. The absolute number is still pretty robust, but those cloud revenue numbers still not keeping up. So the question then becomes, when can we start to finally see that gap close?

And a couple of things I think worth highlighting before we get into this chart right now is I think what's-- we've seen more recently on the back of the announcement from Meta and its Muse app-- because this is a pretty good indication of signs that monetization is starting to pick up from usage. And again, that Muse app is an app designed to act, really, as an AI agent that can perform tasks like shopping, managing communications, booking events, personal content discovery, and so on and so forth.

This is basically a higher value add than simple chatbot interactions. This creates the potential path for monetization of this backdrop through subscriptions, transactions, commerce, advertising.

And keep in mind, Muse is also behind the scenes, generating continual inference workloads. And what does that mean? Well, every user interaction is going to be consuming some sort of compute resources. And these are all running on Meta's AI infrastructure. And when the announcement hit for the Muse, when that hit the tape, you saw a bunch of the chipmakers basically spike on the back of that.

And obviously, with all of these people potentially using Muse and basically having their own personal storage up on the cloud to keep their own personal information segregated, if you will, that just shows an increasing need for incrementally more CPUs. And so we saw those chipmakers bounce on the back of that.

So the bottom line for all of this, though, is that these workloads are now finally becoming real businesses. And I think this does help to justify some of that AI infrastructure spend, where we've had a lot of questions.

So if we think about the cycle right now, Meta spends all this money on AI CapEx. They start to build better models. A result of those better models, this comes out with this AI agent-- in this case, Muse. And then you start to get broader Muse adoption, which leads to more usage, more data and revenue. And really, this justifies, I think, at the end of the day, all this CapEx.

So you can see that this sort of feedback loop is exactly what investors are looking for from these kind of companies that have been spending tens of billions of dollars on AI infrastructure. And if you look at Meta's stock price, it's certainly rewarded as such.

So I think this is one of the first large-scale tests of whether the industry can actually convert AI CapEx into durable application layer revenues. And we're seeing some positive signs there-- so good sign going forward.

Now, shifting back to this last couple of charts here, one of the things that we continue to hear about, though, is financing that CapEx spend. And you hear about all these companies where they're getting-- operating cash flow heading negative.

Well, again, keep in mind a lot of these companies-- very, very strong businesses outside of this AI spend. And so as a result, they're basically taking the revenues from that side of the business and using it to expand out. And you can see here as we move into 2026, we're basically at operation cash flow flat. And then we start to dip into the negative territory and we start to think out to 2027.

But that starts to roll over. And that's because we start to see the CapEx spend, which is basically the numbers here on the top, the growth rate starting to slow down.

So as that slows and the other business line revenues continue to increase in earnest, you start to see that cash flow pick up. And this is not even including the potential revenues coming from all of these AI infrastructure spends that potentially could start to monetize.

So given the projections that we're seeing going forward, this might actually see the worst of that negative free-- or the negative operating cash flow. In 2027, it starts to flip back positive as we push forward to 2028-- so again, moving in the right direction here.

And then the last piece-- we get a lot of questions about the balance sheets of these companies because of all the debt that we keep hearing about being issued. And I just simply went out and looked at the interest coverage ratio for a lot of these hyperscalers and compared them to the median investment-grade bond issuer.

And you can see here-- and again, Oracle is going to be the outlier for all of these, but the point being here you can see these interest coverage ratios for all these hyperscalers ex Oracle stand well above the median investment-grade issuer. And if you even extend that out to some of the other leverage ratios, or in this case, we're looking at debt to EBITDA-- same thing-- still looking at debt well below EBITDA levels for the typical company within the investment-grade space.

So as a result, again, keep in mind where these companies were coming from before they started to issue all this debt. They're coming in from quite strong levels, and as a result, adding a little bit of debt to their balance sheet. And in most cases, this is a capital structure decision going forward. They're still at least, maybe in worst case, in-line with the averages that we're seeing on the investment-grade space.

So again, yeah, we're seeing a lot of issuance coming to the marketplace. I think drawing that analogy where we're starting to see the balance sheets deteriorate on the back of this is a little bit of a stretch. These guys are some of the best companies in the world that have the strongest balance sheets in the world. Taking on a little incremental borrowing isn't going to do much damage to them.

So I think summary-- we are starting to see signs of monetization kicking in. The question is, well, it can persist, and do we see that accelerate?

BRIAN HESS: Thanks very much, Jack. So yes, that move towards profitability certainly should be good for supporting future investment trends. And we've already seen the AI investment trends having been hugely supportive to the US economy over the past couple of years now that we're faced with persistently high oil prices, rapidly rising interest rates, and central banks tightening monetary policy around the world.

So Garrett, I want to turn to you now. And I'm wondering, will this AI investment cycle be enough to help the economy weather the new headwinds, or are we maybe losing momentum beneath the surface? What are you seeing on that front?

GARRETT MELSON: So a couple of things to unpack there-- it really does feel like it's AI, and then everything else. And I think the consensus viewpoint has really come around to the fact that the bulk of that growth impulse is coming from AI.

That's not to say that the rest of the economy is doing terribly. But followers of our work will recognize the term the "ho-hum economy." We coined that at the beginning of the year basically to highlight the fact that growth is good. It's just not great.

And so coming into this year, there's a lot of optimism for strong growth, somewhere-- 2 and 1/2% to 3%. Our view is basically that we wouldn't outperform those lofty expectations. We'd probably come in closer to slightly above trend-- say, low twos. And that seems to be about where we're tracking right now.

And through all the oscillations this year, the shock on the energy front, tightening policy on the monetary front, I think the economy as a whole has really persevered through that in resilient fashion. And certainly, the consumer has.

But I think our view is still largely the same-- is that as you look into the back end of this year and into next year, growth is probably going to be fine. But it's not looking like it's going to be great, at least over the next year or so. Fortunately, that's good enough to continue supporting decent revenue growth and translating to the bottom line for equities in here-- so maybe not the most bullish of outlooks from an economic perspective, but overall not a terrible place to be here, considering some of the shocks that we're facing.

So a couple of things to just walk through here, starting with the AI backdrop, is the fact that if you decompose the economy, again, that idea of AI versus everything else, that's what we're doing on this first slide here. So AI CapEx-- this is pretty broadly defined.

So there is a little bit more that bleeds into this than is necessarily purely AI investment. But information processing equipment expenditure, software, and R&D-- if you broadly define that as AI CapEx, that's growing at about a 13% clip year-over-year.

Take a look at the rest of the domestic final demand backdrop. So we're trying to strip out some of the noise from government expenditures and net exports.

We all know a lot of the AI investment is sourced from abroad. So there are some drags from the GDP accounting perspective. But exclude that and just look at real domestic final demand, and that's only growing about 1 and 1/2% here.

So you have 8% of the economy that's growing gangbusters at 13%, while the rest of the economy is just muddling along here at a fairly low pace here-- again, not huge downside risk, but just not doing a whole lot in terms of a positive growth impulse. And if you put this into context in terms of how we generally think about the US economy, we are still very much a consumption-driven economy.

That still is generally 65% to 70% of the economy. But you can see that the moderation and the pace of growth of consumption has been cooling off and ultimately leading to less of an upside contribution to GDP growth in real terms.

We're looking at this just over the last six quarters. So basically, read that as, what is the average contribution to GDP growth coming from consumer spending every quarter over the last six quarters? You're sitting at right now looking at about 1.4%.

Take a look at where AI has moved. And that's over half of that pace at about 80 basis points here. So the magnitude of that investment is really pretty significant here and translating to a pretty meaningful growth impulse. But what matters more so is moving forward, what do we see in terms of that investment? And I think it's-- a couple of things to keep in mind here.

AI CapEx-- if you listen to corporate management teams, it certainly doesn't sound like that investment is letting up anytime soon. If you really think what the motivations are behind that investment, it's really bimodal.

You're either investing because there are these huge potential returns to be yielding at some point in the future or AI, to the extent that you fall behind your peers, poses an existential threat to your firm as a going concern. And so when those are the two driving forces behind the investment decisions, well, a little bit of monetary tightening here and there is really not going to matter when those profits are so large or the downside is you are no longer in business here.

So the bottom line is we can argue about what the exact level of the investment is moving forward. But you are certainly seeing a lot of signs that investment is going to be robust. And right now, estimates for '27 are tracking well over $1 trillion.

But when we think about growth, growth is a flow measure. It is not a measure of stock. And so that $1 trillion is basically more of a stock measure. What really matters is that growth rate. And I've called those out here in the red numbers above each bar.

And you can see based off current consensus estimates, 2026 is looking like that peak year in terms of CapEx growth. So we're running at about 87% year-over-year. That basically halves to about a 37% pace for 2027.

So now you go back to that last chart. If we're getting 80 basis points contributed to growth on average coming from AI CapEx, well, slash that in half, be a little bit generous, and you're probably looking at now a contribution of about 40 basis points moving forward.

Now, put that into the context of the rest of the economy, and you have a mixed bag here. You have a slowing impulse to growth from CapEx expenditure-- basically morphs from being that growth driver to more of a robust floor under activity. And that just means we're more vulnerable and more reliant on sources of growth outside of AI CapEx.

So where does that stand right now? Well, there's a couple key pillars to focus on. The housing market-- don't need to spend a whole lot of time on that. I think we're well aware that that basically remains in stasis.

The recent backup in yields only ensures that that's going to remain in the penalty box for some time in here-- so not a material drag to growth, per se. It's already basically on ice. And marginal increases in rates doesn't really do much to drive that activity even lower. So you're basically going to just chop sideways here for the foreseeable future.

Outside of the housing market, one of the areas or bright spots is manufacturing. We saw that front and center yesterday with the PMI prints surging higher. But when you actually take a look at activity in the manufacturing sector, it's a little bit more of a mixed bag, as well, here.

PMIs don't necessarily translate very accurately to actual output levels. And you can see that in the divergence between manufacturing production. It's only growing at about 1% clip over the past year despite this pickup in PMIs.

But I think more importantly is part of that acceleration in PMIs, I think, is more so a function of this short-lived inventory restocking cycle. And that's what you can see highlighted on this chart.

Basically, if you look at the PMI survey, that sub-index that tracks customer inventory levels, that's the light blue line. And we've inverted that. So as that line moves higher, that suggests that customer inventories are getting too tight. And ultimately, what do you need to do? You need to rebuild those inventories and produce more product to meet that demand.

And so what does that translate to? It translates to a growth impulse from inventory investment, which is the purple line. If we plug in where the Atlanta Fed GDP now estimate is tracking, that's about a 2% growth impulse from inventory investments for Q3. Well, you can see how that pushes up that four-quarter moving average pretty meaningfully in here.

But the key here is, is this a cyclical re-acceleration that has real legs to it, or is it more of this cyclical-- or excuse me, inventory restocking cycle that once you see inventory levels pick up, which you can already see-- we're starting to see that tick higher in that customer inventories index-- ultimately, that just means that you have a little bit of a dead spot in terms of further production from there.

So I think this remains a pretty key question. The door is open for more of a material and durable re-acceleration in manufacturing. But I think the jury is still very much out on that. And a lot of the commentary that we hear still suggests that inventory restocking is the bigger marginal driver in the near term here.

But if you look at some of the other headwinds out there-- I think we focused a lot on the monetary policy side over the last couple of weeks, of course. But what's playing out on the fiscal side is pretty meaningful as well. And part of the reason for the resilience and growth this year, and certainly for the consumer, has been that positive fiscal impulse as a result of tax cuts from the Big Beautiful Bill.

But we've been talking about this for, it feels like, well over a year at this point, which is we're past that positive fiscal impulse stage, which you can see here in the early part of 2026. And now as you look at these bars, moving out of the balance of this year and into next year, you're looking at about a 40 basis point drag on growth from fiscal spending and other measures through the balance of Q3 and Q4. And that accelerates to about an 80 basis point quarterly drag over 2027.

So now you're actually getting into those pay-fors that want to fund those tax cuts. And ultimately, that just means that you see incrementally larger headwinds to growth. It's not necessarily a death knell, but certainly, a lot of this coming from reduced benefit programs, SNAP spending as well.

It's not just on the federal government front. But you can see that playing out in state and local government budgets that are compressing at the margin as well. And that pressures some spending lower as they have to keep budgets balanced. And so some of those headwinds are just somewhat growing outside of the AI complex here.

And the last thing just to keep in mind is the consumer-- again, still a consumer-driven economy. The consumer has been resilient. But again, part of that resilience was increased tax refunds.

But the other part of it is the fact that you do see consumers spending more out of savings, not necessarily drawing down savings, but compressing that savings rate. And you can see that reflected, basically, in this divergence that's opened up between nominal consumption in the purple line and nominal disposable income, excluding transfer payments, in light blue.

You're basically running a 2X pace on consumption versus income growth over the last six months. So that's helpful to support growth in the near term. But there is some cost to that in the sense that it's generally an unsustainable source of spending.

Yes, we're talking quite a bit about wealth effects. And the fact that financial markets continue to perform well in here does help to support maybe some dissaving from consumers. And that would argue for a lower savings rate equilibrium.

But at the end of the day, you do tend to see some convergence here between spending and incomes here. And to us, this suggests that you're probably due for a little bit of cooling off in the pace of consumption, maybe somewhere from that 3% clip that we saw in Q2 down to a more durable pace of growth around 2%.

So you put that all together. Trend growth is probably still right around that 2% to 1% pace. But you're basically having to clip off that-- about half of that CapEx impulse. So shave off 30, 40 basis points, and you're probably looking at a trend level of growth for the next year or so that might be a little bit closer to a high 1% handle than what we've seen this year.

BRIAN HESS: Thanks a lot, Garrett. I'm looking at your chart here, or table, that shows that 40 basis point potential drag as AI CapEx slows. And I'm looking at the potential fiscal drag, another 40 basis points.

Now it's adding up to 80 basis points at the same time that, with what's going on in interest rates, it's hard to imagine the housing market recovering. And the wage growth chart that you showed is a bit sobering with respect to support for consumption.

So definitely, the signs seem to be in place that we could be at risk for a bit of a slowdown. And I think something that we would need to look at as maybe like a serious risk, but a lower probability outcome would be if-- imagine AI CapEx investment, just instead of slowing, actually contracted somewhat.

Now, Jack's chart spoke to the potential profitability coming online and the momentum behind the spending. So I don't think that's a high risk right now.

But as you start to get out further into the out years, like looking at your chart of 2028-2029, it's showing very little marginal growth in that spending. So it would be very easy to imagine it flipping over and that 80 basis points of support we've gotten going away, which would be quite a change from the last couple of years-- so certainly something to pay attention to.

Now, I mentioned that rising interest rates have been a new headwind in the economy. And we're seeing just yesterday the 10-year yield in the US Treasury market pushed through 5%, the first time it's done that in many, many years. And so I was hoping you could break down for us what's been happening in the Treasury market and what you think has been driving this upside pressure on yields.

GARRETT MELSON: Yeah, of course.

BRIAN HESS: Hold on. Before you start it, though, it's-- this is a macro webinar, where we're always interested in a market view, too. So if you have any kind of inclination if this is a safe place to start thinking about adding duration, we'd love you to get to that, too. Thank you.

GARRETT MELSON: Well, maybe we'll just cut right to the chase on that front. I don't know if it's necessarily a safe place to add duration in here, at least when you see moves like what we did yesterday. But it does feel like you're getting to the point where it's that time to hold your nose and start dipping your toes in-- out the curve a little bit.

And we'll walk through a couple reasons as to why. But to us, the skew is starting to set up a little bit more favorably in terms of adding some duration exposure within portfolios here.

So a couple things to just keep in mind-- first and foremost, it does feel, in many ways, that we've rewound the clock back to 2023. It's Q3. We've had this fundamental repricing in terms of the rate backdrop that started in the late spring, early summer. And now it's starting to transform into something of a buyers' strike at the long end of the curve, with a lot of people unwilling to stick their necks out in front of that duration buzzsaw.

But a couple of things to just put that move into context here-- it's not just a US phenomenon. And I think that really matters here. We've seen some rate moves where it's concentrated within the US and you don't see that confirmation abroad. But that's very much not the case this time.

If you take a look at our moves and rates going back to the end of February, right here, you can see the US isn't even at the top of that list. We're somewhere down in the middle of the pack even though 10-year yields have blown out almost 100 basis points and 30-years are out close to 70 basis points.

You've seen some greater moves coming out of Europe and, obviously, the repricing in Japan. So this is very much a global bear flattening here. And I think one thing that that points to is there's a lot of debate as to whether the bond vigilantes are back.

This, I think, would argue for the fact that-- the simple fact that we're seeing this play out on a global basis across countries with very different fiscal stances tells you, basically, all you need to know. It's really not necessarily a fiscal repricing here, and certainly not just a US dynamic here.

What's really driving it, though, is a repricing, basically, of the policy path. So you think about the backdrop this year, where you've had stronger than expected growth after the shock early this year.

We've, obviously, had disappointing news on the inflation front. And so you put those two together, and you have a stronger nominal growth backdrop. And ultimately, that's simply just translating to tighter expected policy.

So when you break down the move in 10-year yields, do so in the normal fashion, between break-evens, your inflation expectations, and real yields. It's basically all been driven by real yields. The break-evens have gone nowhere.

And if you do that a slightly different way and just break down the 10-year yield into your-- basically, your average expected short rate over that 10-year window and then that term premium, that compensation for taking on that duration risk, again, it's the same story, where term premium has gone basically nowhere. And it's all about repricing that path.

And just to put that into context, the magnitude of that repricing, this is the SOFR curve. So that Secured Overnight Financing Rate-- you can basically use this as a proxy for where the market thinks policy rates will be at various points out in the future.

So we have this going out to 2031-- obviously, a lot of uncertainty that far out. But the big story is what's playing out over the next year, year and a half. And you can see at the end of February, again, a very different environment. We're pricing in three cuts, an imminent job-- white-collar jobpocalypse from AI-- so probably got a little bit over our skis in terms of pricing some of those cuts.

But you take a look at the delta, the change in some of these contracts. The June contract for next year-- you've moved about 200 basis points-- so moving from those cuts to outright pricing in a tightening cycle up to almost 5% policy rates by the middle of next year-- so a very, very sharp repricing here.

And that ultimately drives a lot of the move at the front end of the curve. But it's been so sharp and so persistent-- you see that we stabilize at those elevated levels-- that that actually starts to leak out the curve and places upside pressure on 10s and 30s in here.

Some of the other debates and reasons for this move-- it's funny. Anytime you get a sharp move in rates, basically, the market comes out and just throws out any narrative it can at the wall to see what sticks. And so the bond vigilantes is the one you can always reliably count on to come back in here.

But one of the ones that we've heard a lot more lately has been this idea of a loss of Fed credibility. So that aligns with the incoming Fed chair, Warsh, and this idea that he was actually now a dove and would find any way possible to be cutting rates, even though the inflation backdrop and the growth backdrop is arguing for stable to even tightening policy here.

But if you take a look at inflation break-evens, again, I mentioned they really have not moved anything in terms of this repricing, well, basically across all these tenors. That five-year window-- so the five-year window starting five years from today-- that's basically the Fed's preferred market-based measure of inflation expectations. That's hardly budged. And the same story holds for 10-year and 30-year break-evens.

So if this is really a story about a loss of credibility, you would expect to see a lot more pressure in terms of inflation expectations leaking higher because it was basically pricing in the fact that the Fed would not tighten real policy rates. And that's just simply not the case.

You've seen the exact opposite, markets pricing in that higher real yield and basically reflecting the fact that central banks would do their job to fight inflation and tighten policy as a result-- so not a whole lot to suggest that this one holds a significant amount of water. Maybe it added to it some of that buyer strike dynamic, just given some concerns there, but not a whole lot of support in the empirical evidence.

But the other one that's a little bit more interesting and ties back, again, to the AI theme is this idea that this surge in AI debt issuance is actually crowding out Treasury markets. And I have to laugh a little bit because when you're talking about crowding out, it's usually the opposite direction. You have too much Treasury issuance, and that ultimately crowds out private investment. And so you're flipping that narrative on its head.

But again, when you really look at the empirical data, I don't think there's a huge amount of support for this. Is it a marginal driver of some upside pressure, especially at the long end? Perhaps. Maybe you could argue 10, 15 basis points. That might even be a stretch.

If you just put things into context here, what really matters from a supply perspective is not the gross numbers that we hear about-- so $250 billion from the hyperscalers this year, $1.5 trillion from total corporates on a non-financial corporate basis. What matters is the net issuance.

So exclude those maturities. Exclude those calls. You're actually looking at right now, running on a net basis, around $450 billion of supply. And that's all across the curve. So it's still a large sum. It's still above what we've seen over the last couple of years, but much smaller than some of the headline numbers that you hear on a gross basis.

Of that, net tech issuance-- so hyperscalers, and then other IG and high-yield issuers-- it's actually running at just $130 billion. So again, it's higher than prior years. It's not really an unmanageable number, though.

And so what we've done on this chart is just scale total net non-financial corporate issuance to the size of the marketable US Treasury market here-- so just to give you a context of that incoming supply versus the outstanding stock of treasuries. And the story is generally the same here. 2026 is certainly running ahead of the pace that we've seen for the last five years. It's obviously well below what we saw in 2020.

And keep in mind, that was an era of just a huge amount of deals getting done for corporates to build war chests through the pandemic era. So take that one with a little-- with a grain of salt here.

But the bottom line is that delta is really not all that significant here. You're running at about 1.7% of total outstanding treasuries of new net supply. Last couple of years, you're running somewhere around 1% to 1 and 1/4%. So you're talking about a 50-60 basis point delta here.

Yes, it is more supply hitting the market. But to pin all of on this idea that AI issuance is really crowding out the Treasury market, I think, is a bridge too far in our book here-- so maybe a marginal driver, but not really a key driver.

The bigger driver, in our view, is probably the fact that, one, if you want to make a case that AI is driving it, just stick to the simple argument. AI is translating to a much stronger growth impulse now.

That is pressuring up prices in certain areas, namely information processing equipment. And ultimately, that's translating to higher nominal growth. And that's what's translating to higher yields. And that's what you see highlighted on this last slide.

Nominal growth generally drives nominal yields. That's the fundamental driver. You can certainly have technical oscillations. And I think that is part of the story now when you look at positioning, very, very crowded in terms of net short rates basically across tenors and across global sovereign bond markets.

But at the end of the day, there is some fundamental support for why we've seen a repricing in 10s. The bigger story is, moving forward, are we durably shifting into a regime of nominal growth in the 6 and 1/2% to 7% range?

Well, after all that we've just talked through, I do think that there's a window, and who knows when this hits, where some of that CapEx starts translating into real productivity gains and real higher growth. But between now and then, I think the risk is you actually do see a little bit of a soft patch in growth, especially over the next 12 to 18 months in here.

And so from where we see things, a lot of the data still seems more consistent with a nominal growth environment, somewhere closer to about 5%. And so, yes, we're still somewhat trapped in this negative feedback loop.

And so that's why I say maybe it's not a safe place to be adding duration, just given that direction and path of least resistance. But there's nothing like a whiff of slower growth and softer inflation to really spur a renewed bid into duration, especially when the positioning is as crowded and extended as it is.

And so if you do start to see some of that volatility at the long end settling down and some signs of that softer growth, I don't think it's going to take a whole lot to get some incremental bid coming into the long end, especially when there's as much value in real yields as we find today. 10s are close to 2.80%, 30-year real yields close to 3.15%-- so a lot of value there for the taking for some of these more price-sensitive buyers here.

And so at the end of the day, I think we're a lot closer to the end of this repricing than the beginning. And from our perspective, that sets up a skew that says, hold your nose and start adding a little duration in here.

BRIAN HESS: Thank you, Garrett. That was a refreshingly direct answer to my question. I like that. And I think you made the strong case for what could potentially unlock the value that's been building up in treasuries. So for the buyer with a medium-term horizon, it's probably an interesting place to start thinking about adding some risk there.

Now, Jack, Garrett made a pretty clear argument for how the market has repriced the Fed rate path. And that's been what's putting upward pressure on yields.

Now, coming into-- or I guess all year, really, there's been this fear that perhaps the next Fed chairman might not be able to raise interest rates due to a lack of independence. That argument or that question was settled last week, when the Fed raises rates for the first time since 2023.

So I thought maybe we could spend a little bit of time now talking about why central banks are feeling the need to raise interest rates. And anything that might change their perspective about that would be worth hearing about.

JACK JANASIEWICZ: Yeah. And it's interesting, too. Keep in mind that the Fed is a unique central bank in the global scheme because it carries a dual mandate. You've got price stability and full employment. But the bulk of other major central banks around the world really only have that price stability mandate.

And so the energy component there starts to really have a significant impact-- and so a little bit of a different trajectory, I think, going forward because a lot of these central banks are going to focus exclusively on the inflation backdrop, whereas the US has to look a little bit at both sides, the labor mandate and the full employment mandate-- and the price stability mandate. But when we start to think about what's going on with regard to the global backdrop, it still comes down to inflation. And that's the bottom line here.

And so just looking at a chart of the major regions around the world, you can see here that the core inflation rates are certainly well above that-- well, I shouldn't say well above-- but above that 2% target. And so from that perspective, not shockingly, we should expect to continue to see the idea that global central banks need to start to continue to tighten and trying to get that level back down to where that target rate is.

But what's interesting if we start to look at what's priced in, again, looking at a lot of these short-term interest rate contracts-- and a lot of these only go out to a summer of next year. So that's, hence, why I've basically pulled the numbers here through July of next year.

But on average, you're looking at somewhere between three and four rate hikes still coming in the pipelines from a lot of these global central banks. And that's where this gets a little bit interesting.

And Garrett has outlined some of the things that I'm going to walk through at the margin here. But part of our argument here might be simply that maybe we're getting a little bit too hawkish at a global level. And maybe we're getting a little ahead of ourselves because if we think about what's pushing yields higher, we're certainly hearing a number of different catalysts.

You're hearing about, obviously, the inflation concerns. But as Garrett pointed out, those inflation expectations remain fairly well anchored. And that's true across the globe.

Certainly, energy prices are becoming a really big concern. When you look at the correlation between Brent oil prices and the US Treasury 10-year yield, it's basically at the highest level on record.

So you can certainly see how that's directly impacting the rate backdrop. But from a Fed perspective, it's the second-order effects of those energy costs being pushed through and starting to infect the broader economy.

And so think of things like airfares. That's going to be one of the most sensitive portions of the marketplace relative to the energy backdrop because jet fuel is such a big expense for the airline companies. And when we look at those line items within the inflation basket, airfares certainly remain elevated-- so the second-order impacts certainly putting upward pressure on yields and upward pressure on the need for Fed hiking in here.

Garrett, again, mentioned the expected future path of rates going forward-- so that central bank reaction function certainly being priced in here. And you're also hearing concerns about deficits and non-discretionary spending and then, again, supply.

Garrett touched on the corporate side. But don't forget the sovereign side, as well, financing from the deficits not just here in the United States, but abroad as well. You've got the defense spending, the rearmament. You've got some of the infrastructure bills that are being potentially passed in Europe-- so plenty of fiscal spending abroad that's going to require financing as well.

So you put all of these things together, and you can certainly make the case as to why you could certainly see central banks having to continue to tighten rates going forward. The three things, though, that I think are a little more interesting from our perspective-- and this is what I wanted to touch on for this portion of my presentation.

It's certainly the repricing of rates going forward. But two other things that I think are worth mentioning here-- and one, it's the savings glut that's really disappeared.

Throughout much of the early 2000s and 2010s, we had this massive savings glut out there. And so what do I mean by savings glut? Well, you're looking at, for example, a lot of emerging market central banks started to hoard reserves. They were trying to protect their balance sheets, protect their currencies, protect their ability to make those payments on the sovereign debt backdrop.

And as a result, central bank reserves began to skyrocket. And what happens with those reserves? Well, those get recycled back into safe-haven assets-- i.e., Treasury bonds. So that was certainly making its way back into the bond market.

You also had these massive gains with regard to commodity price windfalls-- so a lot of petrodollar savings being recycled back into the bond markets. At the same time, you weren't really seeing massive amounts of investment.

So rather than seeing some of this money being put to work, you're actually seeing it saved. And so that lack of investment starts to weigh on that backdrop and really increase that savings glut.

And if you just want to get a little bit more specific for what was going on here in the US, you're coming out of the ashes of the 2008 global financial crisis. And so you had a lot of these major banks that were certainly beefing up their cash reserves. So you started to see cash on the balance sheets starting to really pick up aggressively there.

Still, again, in the United States-- somewhat of an economic uncertainty backdrop, which was, I think, weighing on investment here. Tech was really starting to take off at that point. And that, obviously, was requiring less physical capital, which meant you didn't have to borrow as much-- so again, more savings there.

And maybe the last piece is really the upper income earners were really benefiting from the economy at that time. And they have a much higher propensity to save versus spend. So you put all these things together, and you have this massive savings glut.

Well, if you fast-forward to where we are today, that's basically now working in reverse. And so that savings glut backdrop that was very supportive for fixed-income assets is fading away. And I think that's putting a little bit of upward pressure on the overall backdrop for rates.

The last piece is really this idea of normalization of monetary policy. And this is somewhat of a busy chart here. So I'll walk through this. But that red line is looking at, basically, the Fed funds rate. The purple and the blue line are just looking at the 10-year and the 30-year relative to the Fed funds rate.

So we're trying to get somewhat of a snapshot of what the yield curve looks anchored to where Fed funds are. And I think when we start to think about normalization of monetary policy, you're thinking about, really, the lower zero bound.

We've spent quite a bit of time over the last couple of years coming out of these crises where the Fed pushed rates, basically, to 0%. And I think that had somewhat distorted the backdrop for the overall bond market.

And so what we were trying to do here is look at periods where we're really not in the zero lower bound, as well as these excess backdrops where you have Fed funds north of 6%-- so what we would consider to probably be more of a normal backdrop. So looking at periods between 1% and 6% on Fed funds-- and that's basically the area here that is not being shaded by gray.

So we do have periods in the run-up to the dot-com pop-- so the mid to late '90s, coming out of the dot-com. So basically, from that run all the way up to the GFC, and then a little bit right before the COVID backdrop, and then more recently, obviously, the last two or three years—

So when we look at that spread between where the 10-year is relative to the Fed funds rate, where the 30-year bond is relative to the Fed funds rate, historically speaking, when you're in what we would consider to be more normal monetary policy backdrop, those relationships stand at, basically, those same levels that we're sitting at today.

So again, on average, the 10-year relative to Fed funds was right around 109 basis points. Well, when I pulled this number the other day, where are we today? At about 107. Looking at where the 30-year is relative to Fed funds, the average in those more normalized monetary policy backdrops-- 149 basis points. Where are we today? 141.

So the point here is simply there's also this concept of monetary policy basically normalizing. And when we think about that normalizing, this is basically where we should be from a nominal rates perspective. It's really not all that abnormal to see rates at this level.

And I think people are missing some of that backdrop because we've been, basically, at these periods of Fed intervention where you're looking at depressed yields, depressed borrowing rates. And the bigger backdrop here, that savings glut, is gone.

So again, 10-year pushing 5%-- that's probably about where we should be, given this backdrop. And so I don't think this is all that abnormal and all that scary, so to speak.

BRIAN HESS: Jack, so I guess another way of describing what you're showing in this chart is to say that with the Fed funds level where it's at today, it would be unreasonable to expect 300 basis points of curve steepness, for example?

JACK JANASIEWICZ: Correct.

BRIAN HESS: And the only way we would get 300 basis points of curve steepness would be if the Fed ended up cutting rates very significantly?

JACK JANASIEWICZ: Yep, and that's obviously a bad thing for the economy and the stock market, I would say.

BRIAN HESS: But a good thing for the bond market, a good thing for long-term bonds. So to the extent that the Fed's cutting rates and the curve is steepening, it probably is steepening in the context of rates falling everywhere, whereas for long-term rates to be pushed up significantly higher, it probably requires a central bank to take rates up significantly higher, which I'm not sure we see the setup for that. And I think that's what we want to go to next with Garrett, where he's going to talk about inflation.

So Garrett, can you run through some inflation charts, because we've spent a lot of time on this macro webinar talking about the bond market, basically, in one way or another? I think to round it out, we should probably spend some time on inflation. And maybe you can opine why you think the rates market might be overestimating the need for additional hikes, which is more of that argument for why it could be a good time to think about duration. So Garrett, over to you.

GARRETT MELSON: Absolutely. Well, a lot of time, I just talk about the bond market. But it does feel like that's really been the marginal driver for risk assets over the last couple months in here.

Hopefully, we're starting to move away from that and we can get back into more of a theme, a narrative-driven market for equities. But as far as the rates backdrop goes from an inflation perspective, I think Jack already laid it out, which is the sense that from our view of the pricing, right now, markets look to be pretty hawked up-- in our view, too hawked up.

That's not to say that there aren't good reasons for the Fed and markets to be hawkish. But I think when you actually start to peel back some of the layers of the inflation outlook, it is actually marginally more encouraging than what I think the consensus narrative is right now, and one that doesn't necessarily necessitate the-- an outlook that's as hawkish as we have priced into markets right now.

So a couple of things to keep in mind-- the first one is the dynamics of the current outlook right now are pretty unique in the sense that what we're seeing is a really significant inversion of the historical relationship between core CPI and the Fed's preferred measure of core PCE. And historically, because of how they're constructed, some of the weighting and scope and methodology differences, core CPI tends to run about 30-40 basis points above core PCE.

Right now, you flip that over, and we're basically running 90, almost 100 basis points higher on core PCE than core CPI. So it is a very sharp reversal here of that historical trend.

Some of that's the unique weightings. As I mentioned, CapEx is having a much larger impact on core PCE just because of the weighting to information and processing equipment. That's about 30 basis points of that spread alone.

So there are some good reasons for that. But I think the key here is that core PCE is somewhat overstating the degree of deterioration in the inflation outlook. And so I think it's a little bit more reflective of some of these idiosyncratic shocks and supply shocks that is somewhat overstating the deterioration of that outlook here.

The next thing to keep in mind, though, is that in line with some of these idiosyncratic shocks is the fact that, yes, we've had some firmer prints. But what you've actually seen for much of the last couple of years and over the last couple months here for this year, as well, is the fact that inflation breadth is actually narrowing. What I mean by that is the share, the weight of those components that have an annualized growth rate over 2% on a monthly basis-- that's slowly, but surely moving its way lower here.

So inflation is very much a process. And the Fed's worried about broad price pressures. Right now, it seems like they're much more responding to the fact that they just can't look through the fifth or sixth in the series of negative supply shocks when you're running at-- as levels above their target as we are right now.

And so ultimately, they do have to tighten to try to bring better balance into play here. But it doesn't look like a broad inflation issue. It looks like more of a narrow one, which you could argue is going to improve on its own as you move through the next couple of months and quarters here. So I think this is something to keep in mind. It's not the broad issues that we saw back in 2021 and 2022.

The other thing that certainly has faded away from the market's focus is tariffs. But that hasn't faded out of the inflation prints and data just yet.

When you actually decompose the tariff contributions here, there's a couple different regional Fed banks that have done research on this. But this one is coming from the St. Louis Fed. And what they found is that you basically saw this last impulse of tariff pass-through come through in January and February.

So February is the peak-- peaked out right around a 60 basis point contribution in terms of where inflation is running relative to the target. That was about half of that overshoot at the beginning of the year. Now you can see some of these other factors have really picked up the flag here, as you've seen a little bit of a moderation.

But the story here is that you're still running at about 40 basis points contribution to excess inflation coming from tariffs. And assuming that we stay at a status quo, that's 40 basis points of that year-over-year print that's going to roll off over the next 12 months here. So that's some of the easy work to get the core PCE back down closer to the target here.

But the last one that I think is really telling is when you look through all the noise of the supply shocks and get a better read on underlying inflation, everything's really starting to move in the right direction here. This is the New York Fed's multivariate core trend.

It's a mouthful, basically, to say that they're trying to strip out some of these idiosyncratic and supply shocks, short-lived shocks, and get a better sense of where underlying inflation trends are. And you can see it's actually done a pretty good job of marking inflection points in core PCE, which we've laid out in blue.

And so it's not just one pocket or one component of the inflation basket. It's all three of the main categories the Fed looks at-- core goods, housing, and then super core services in purple here. They're all starting to dive lower and diverging pretty materially from what we're actually seeing in the core PCE print-- so again, more evidence to suggest core PCE might be somewhat overstating the deterioration in the inflation outlook, but also suggesting that it's probably some of those idiosyncratic supply shocks that are having much more of an impact here. And that probably just is a matter of time to work out and ultimately resolve on its own without the need for tighter policy here.

So when you think about market pricing, this is the probability distribution that we always like to look at. That curve has shifted significantly from the left to the right here-- and so some good reasons, given the inflation data we've seen this year, to be more hawkish. But the modal outcome or base case for the market pricing, especially after yesterday's move, probably is around four to five hikes in total from where we started this cycle.

In our view, I think the data probably suggests that if growth is moderating at the margin, you get some help from some of these disinflationary trends that are underneath the surface kicking back into gear here. This is probably more likely to be a short-lived recalibration of policy-- so basically, walk back those three insurance cuts from last year-- as opposed to a really durable and longer tightening cycle where you're looking at five, potentially, more hikes, which is, honestly, what the market is discounting in here-- so another reason to say, hey, this is probably another feather in your cap to say, maybe start thinking about adding in some duration here.

BRIAN HESS: Garrett, we got a client question asking about potential revisions to the PCE data at the end of the month. Do you think that could be a big enough shift that it would change market sentiment and perhaps catalyze a different thinking towards inflation?

GARRETT MELSON: That is a great question and a point that I wanted to make. If you look at those revisions-- we get them annually in September-- I think a lot of the estimates are somewhere around 20 to 40 basis points-- so kind of a wide spread. But ultimately, that's 20 to 40 basis points of that year-over-year core PCE number is going to be revised lower. So it means that that gap is somewhat overstated right now.

I would say it's probably not going to materially change market sentiment too much just in the sense that most people are pretty aware of that incoming revision and the fact that it's going to take down those year-over-year prints. So I think to me, it matters a lot more about what some of those underlying trends are. But it certainly doesn't hurt to help get those numbers a little bit closer back to target here.

BRIAN HESS: Thank you very much for that. Appreciate it. You spent some time earlier opining on whether 6% to 6 and 1/2% nominal GDP growth is sustainable. And I think you've made a case for why growth could slow, which is one input there. And you made a case for why inflation could slow.

So between those two inputs, it certainly makes a good argument for why 6 and 1/2% nominal GDP growth is a temporary thing, probably tied to the AI CapEx cycle that's been so powerful, which is where we started this macro webinar. So just bringing it full circle, to the extent that that slows, other drivers of the economy slow, inflation slows, we get downward revisions, you can paint a scenario where we no longer have this big divergence between nominal GDP and long-term interest rates.

GARRETT MELSON: Absolutely.

BRIAN HESS: So I think with respect to the bond market, let's leave it there. And we're running a bit tight on time. But we do want to spend our last segment on something completely different.

We have an election coming up in the US in a couple of months. And we've been getting more and more client questions about it. So we thought maybe we'd highlight some interesting data about market performance around the election cycle, plus some other tidbits Jack put together for us. So Jack, this one's all yours.

JACK JANASIEWICZ: And I'll make this real quick so you can make sure we cover all this stuff. But everybody is aware of the seasonality for September. And we've written about this in some of our writings, where it really is the trend coming into September if the market is heavy, it tends to trade poorly in September. But if the market's trading pretty well and in good shape, it tends to do well.

And so it's a-- really a-- we modeled that by just simply saying, hey, if we came into September above the 200-day moving average on the S&P 500, that was considered to be a positive backdrop. And lo and behold, September tends to be a good month when it comes in above the 200-day. When it's below the 200-day, it was a weak month.

But interestingly enough, usually, the last two weeks of September tend to be soft, and then you start to turn. And we're just looking here at the median performance in midterm years.

You can see that-- the red vertical line that I have in there. That's the midterm date. And you can see we roughly bottom about three weeks before midterms. And then the market tends to just run into the end of the year.

So from a historical seasonality perspective, the last two weeks of September-- pretty soft. And then we start to pick up as we head into the election backdrop. So keep that in the back of your mind when you're thinking about positioning going forward here.

When we start to think about the potential outcomes for the election-- these are just the Polymarket odds-- you can certainly see here that the Democrats in control of both the House and the Senate-- the odds are now pushing up closer to 66%. So the market's continuing to price in the Democrat sweep of Congress.

So what are the potential ramifications if we do get a Democrat-controlled Congress? Well, a couple of things worth highlighting here-- on the health front, on the health care front, you're probably going to see the ACA tax credits extended.

From a consumer discretionary perspective, I wouldn't be shocked if we continue to see a little bit more support for the lower end consumer. So think of things like affordability being highlighted here and some targeted fiscal programs to help support the-- that lower income portion of the lower K-- of the K-shaped economy, so to speak.

On the industrial front, I think there's general agreement that defense spending is going to need to persist. We need to replenish our armaments. But I think technology pushing, really, the frontier-- so some of these next-gen defense backdrops probably are going to face some headwinds.

Financials-- you probably get some headwinds from capital markets with-- maybe a little bit of the deregulation still lingers. And then obviously, on the energy front, permitting reform probably doesn't really get much traction in here as well.

So I think in general, you're probably going to see deregulation, those sort of backdrops fading, and a little bit more oversight with some potential headwinds on that front. But despite all of this, keep in mind, though, as we come into the end of the year and into next year, we are in the third year of a presidential cycle.

And presidential cycles-- that third year tends to be very strong. And so you can see here on average, the market goes up about 19%. And the hit rate on that is 83%-- so very strong numbers when you think, again, for that next year for-- from a seasonality perspective, regardless of what happens with Congress.

And then to wrap up, we like to say it's profits over politics here. And so on the left side, you can see the number of years where you have a positive return based on who is in office.

Obviously, the Democrats here have a slightly better hit rate than the Republicans do. But-- the overarching factors here on the right side, where if you have a positive earnings per share growth backdrop, I think that tends to override that political backdrop. And that tends to be the ultimate driving force here.

So again, it's profits over politics. Keep that in mind as we roll into next year. And again, what we've highlighted is broad slowing. But slow and slowing are two different things.

Recession odds are still pretty low. And when we look at earnings estimates for next year, you're looking at mid double digits. The economy-- or if the stock market continues to just grow at whatever the pace of earnings growth is, that's still a pretty solid return for next year.

BRIAN HESS: Thanks a lot, Jack. That was a great way to wrap things up. Hopefully, we've given our viewers plenty to think about after today's webinar.

Thanks so much for our viewers for-- to our viewers for attending this quarter. We wish everyone a great rest of the year. We'll see you later in the year. And I hope to see you all next quarter for another one. Thanks so much.