Pulling back the layers: What's driving markets

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

Now today we're going to talk about the US economy, the impact of AI on the stock market and the economy. We're going to discuss the Fed and its new chairman and close with some key US stock market topics like factor, performance, and supply risk. So it should be a good discussion. We have a wide variety of items on the agenda. 

Now on to today's topics. So Garrett, let's get right into it. Let's get going with the US economic update. As we approach the midpoint of the year, how's the economy looking, and what are some of the key variables you're tracking to help guide your forecast for the rest of 2026? 

GARRETT MELSON: All right, well, thanks for the intro there, Brian. Honestly, I would say our outlook hasn't changed a whole lot from where we stood at the beginning of the year. We coined this phrase, "the ho-hum economy" at the top of the year. And that was really referring to our expectations that growth would be fine, but it was likely going to be coming in below what were at the time some pretty lofty expectations from the consensus. 

We'd heard calls anywhere from 2.5 to 3% real growth for 2026, and a lot of that optimism really was resting on this fiscal impulse from tax refunds. Obviously, the energy surge has really eroded that fiscal impulse. And so we were looking through that. We expected the effects to be a little bit more modest. But the bigger story was, what was happening in the rest of the economy. 

And I think for the most part, a lot of those conditions really remain basically where they stood at the beginning of the year. I would say the nuance is that conditions do appear to have stabilized to a degree, and so the distribution of outcomes is somewhat shifted to the right. But generally when we look at conditions, I think that ho-hum sentiment largely persists today. 

And so I'll walk through a couple slides just to give you some context in terms of how we're thinking about the economy, how we're thinking about how conditions unfold for the balance of the year. And you'll probably recognize a lot of these themes. It's boiling down the economy to the building blocks. 

So if you think about the key sources of growth, first one to keep an eye on is the housing market. As the adage goes, housing is the economy. But that really hasn't been the case over the last couple of years, and I think we all know the story there. Mortgage rates continue to be elevated. Affordability continues to be an issue. 

And so while we have this strong demand, that demand continues to be somewhat sidelined in here. And so we're starting to see inventories improve a little bit. You're not seeing the same kind of drag on new construction for home-builders where they were just working through that inventory or those homes that were underway and really holding back on starts. 

But we're still in this limbo where conditions, as you can see in the light blue line here-- that housing market index is basically a sentiment indicator for the big home-builders. It's still bobbing around the bottom in the lows that we've plumbed out over the last couple of years. And when you look at a couple of the leading indicators-- so starts in purple. They've had a couple false starts here. No pun intended. But you've seen that converge back down to the more reliable leading indicator, which is permits. 

So when you think about the contribution to growth coming from the housing market, there's a couple of different components, but it does start with that construction activity, and you're just not seeing a whole lot of activity picking up there. So on the whole, housing really remains mired in the stasis it's been in for the last couple of years. 

The other side of the investment coin is obviously what you're seeing in the nonresidential space. We'll touch on the AI buildout in a moment here. That continues to be running gangbusters here. So certainly a bright spot within the economy. But outside of the AI complex, you're really not seeing a whole lot of any activity to write home about. 

We keep a close eye on CapEx intentions from the regional Fed manufacturing surveys, so going out and asking those purchasing managers what they're thinking about in terms of CapEx plans over the next six months. If you aggregate those into a composite, that gives you the light blue line here. 

So we're certainly working our way out of the basement here, but we're basically back to average levels here. And that does tend to correlate pretty closely with non-AI CapEx contribution to growth. So excluding the AI components, you're just not seeing a whole lot of optimism coming from business leaders or seeing it in the real hard data in terms of this boost to CapEx that was supposed to come on the back of tax policy. 

We've said this time and time again. You can make tax policy more conducive to investment, but you can't force businesses to invest. And that's really going to be more of a function of the demand backdrop and growth expectations more than tax policy. And clearly, we're just not seeing the type of surge and demand and pickup in growth that really is driving a huge pickup in CapEx outside of the AI complex. 

Another bright spot in addition to the AI complex would be manufacturing. This is getting a lot of attention. There's a lot of hope here that we're starting to see a re-acceleration within the manufacturing economy here. You certainly have seen activity pick up. You've seen employment within the manufacturing complex pick up as well. 

And the reason for that is-- think back to last year. We had the tariff shock. A lot of companies frontrunning those tariffs in anticipation of trying to get lower-cost inventories over to the States here. And they've basically been unwinding those inventories and selling out of those inventories since. But we've gotten to a point where inventories are just simply too low relative to what is a resilient demand. 

And you can see that really highlighted in this next chart here, the light blue line. Take a look at that manufacturing PMI survey and specifically look at what businesses are saying their customer inventory levels are. That line is inverted. 

So the idea being here, as that light blue line moves lower, then inventories are very high. What does that mean? That means companies don't have to produce more inventory, and it's that production that gets booked into GDP growth. And that's what you're looking at for the purple line. 

Where are we today? Well, you can see that inventories have been steadily slipping here, and we've gotten to a point where you're starting to see activity pick back up. Companies are actually ramping up production to rebuild those inventories. But you're actually starting to see a little bit of a reversal there. So there might be a little bit more juice to squeeze here, and I think the additional impulse is probably coming from the energy shock. 

Same idea of pulling forward demand, trying to get ahead of rising costs. And I think those two in unison are what's really driving this pickup in manufacturing activity. So it's certainly additive to growth, but I think it's likely to be much more of a short-term phenomenon than any sort of a structural or long-run, cyclical recovery in the manufacturing complex. 

That brings us to the labor market. And I think this is really the key story here. This is where you're hearing a lot about re-acceleration. And we'll certainly concede-- we've been cautious on the labor market for some time now, but it's very clear that you have seen a stabilization in activity here. And what you're looking at is just nonfarm payrolls and private payrolls in the light blue and purple lines, respectively. And that's a three-month average, so giving you a little bit of a sense of the trends here. 

But I've overlaid that with the unemployment rate in green. And the key theme here is you've seen a pickup in jobs creation, and that is absolutely welcome. But you're not really seeing a huge rollover in the unemployment rate. Yes, we're well off the highs that we ticked back in November of last year. But you go back to basically the lows in job creation in August of last year, and you're basically unchanged on the unemployment rate. 

So what does that suggest? Again, think about the distribution here. I'd say conditions have stabilized. You're no longer in this linear cooling in the labor market. But it still, to us, looks like conditions are soft but stable. And that certainly is an upgrade from where we started the year. But I think there's another point to take away here, which is if we are actually seeing this level of payrolls growth, and yet the unemployment rate is basically moving sideways, well, we talked a whole lot about that break-even rate of employment growth. 

We had this negative supply shock from immigration policy last year, and so essentially the bar for what constituted a good jobs report came down pretty materially. So you didn't really need to add a whole lot of jobs because the labor force wasn't growing, and so you could actually see the unemployment rate stay stable despite not creating many jobs. Now you flip that on its head. If the labor market is seeing a huge amount of jobs creation, and the unemployment rate is going sideways, that argues that maybe you've actually seen a pretty significant shift higher in that break-even of jobs creation here to keep that unemployment rate stable here. 

So I think there's a couple questions here. The conditions have improved, but to us, I don't think there's enough here to decisively say that the risks are now skewed to the upside. And there's a few more reasons for that that I think are probably the bigger takeaways here. The first one is, hiring intensity has still remained anemic here. You're looking at job openings from the JOLTS survey. 

We got a big spike there, but to us, that looks somewhat anomalous. It was driven by one industry, professional business services. And the bigger story is, look at the relationship with the Indeed data. This has been a very reliable coincident indicator with JOLTS. And you're not seeing any confirmation of that uptick. If anything, it's continuing to grind lower here. 

So you're certainly not moving out of this no-hire, no-fire environment. You're just not seeing any sort of uptick in job demand. And I think that's going to be a requisite to really declare that you're seeing a renewed tightening in labor markets. 

But the really biggest piece of evidence running counter to this reacceleration is the fact that across nearly every measure of wage growth you look at-- you don't have to call out each one of these lines. Just take a look at the trends. They're all moving down and to the right. Now yes, wage growth is somewhat of a lagging indicator.

So you should start to see some tightening play out before that wage growth starts to inflect higher. But there's very little sign of that trend of cooling stabilizing just yet. So the clearest indications of a tightening labor market-- that tends to be wage growth. And that still suggests to us that there is slack out there. 

So that leaves you with really an economy where, if you boil it down to that last building block for the consumer being the big driver of economic growth, well, you're still somewhat faced with this dual squeeze on the consumer. In the sense that upper incomes are still doing just fine. The wealth effects are helping support income. That's great. That puts in a high floor for consumption. But what really drives that marginal rate of growth is going to be lower-income consumers. 

And we've talked ad nauseam about the K-shaped economy here, but I think this is a good way to encapsulate that. Take a look at nominal income, wage and salary disbursements in purple. You're actually now sitting below the level of both the Fed funds rate as well as the headline PCE. So what does that suggest? Not only a real income squeeze because costs are rising and inflation is above wage growth. But it also suggests that you're starting to see some tightening with respect to financial conditions via debt-servicing costs. 

Incomes are growing slower than debt-servicing costs. So that tends to squeeze incomes as well. So it's not necessarily a dire backdrop for consumption. But I think it just leaves us with where we started the year, which is to say that there's probably going to be some continued moderation in consumption, even when you allow for the fact that energy shock's probably fading. You're starting to see some of that impact fade in reverse. Potentially some of that might just go back into rebuilding buffers and rebuilding savings here. 

So a lot to unpack, but I think we're largely back where we started the year. AI is certainly resilient. That's helping to support growth. But a lot of the other engines are still basically on ice, and that leaves you with a ho-hum economy that's probably somewhere around 2%. That's fine. It's not something to write home about, but it's certainly not falling out of bed. And ultimately, I think that's still a decent backdrop for risk assets looking through the balance of the year. 

BRIAN HESS: Thanks, Garrett. So you painted a mixed picture of the economy, with some pluses and some minuses. And I think that's true as we drill down through the layers. One thing that hasn't been missed this year, though, has been the supremacy of AI as a thematic-- driving markets, for sure, but even the economy to an extent. The technology sector has dominated all other areas of the stock market this year. 

So Jack, I think we should focus on technology stocks for a minute, if we can. Many people think a bubble might be forming. I'm curious if we could get your thoughts on the fundamentals underlying that sector of the market. 

JACK JANASIEWICZ: Yeah, and I think it makes some sense here to talk through some of the earnings backdrop first. By certain estimates, depending on how you're categorizing the individual names, AI and AI-related stocks could account for somewhere between 45% and 55% of the S&P 500 today. 

And you can tier them out. You've got the core AI leaders, which I would consider to be the mega-cap hyperscalers, so things like Apple, Google, Microsoft, Meta, and NVIDIA. And then you have all the ecosystem players underneath them that are basically leveraged to that supply-chain backdrop. So you've got the utilities and data centers there, semiconductors and optical software and analytics, certainly a pretty large and diverse ecosystem. 

And I think it makes a lot of sense to actually talk about that earnings backdrop because they actually matter quite a bit. And so this first chart that we're looking at here-- and I think it's worth highlighting the pace of acceleration in earnings has been quite impressive, to say the least. 

That purple line down the bottom that reflects the 2025 backdrop-- pretty flat. You can see the green line there. That's 2026. The gray line, 2027. And the blue, which is the next 12 months-- basically an average of '26 and '27. And note that we really start to take off and accelerate once you get past the end of 2025. So pretty impressive numbers pushing straight up there. 

And if I actually put some numbers on these lines, so to speak, the 2026 estimates are up almost 19% since the start of this year. For 2027, you're up north of 27%. And this is all happening while multiples are contracting. That forward PE for the tech space has actually dropped 8% since the start of the year. So pretty impressive. 

If we also look at the earnings backdrop, then you're getting a huge dichotomy in terms of the split. The big contributions to the overall headline S&P 500 number, really, I think are being driven by the tech space. And this bar chart, I think, is trying to hammer that same point home. 

We're looking at the earnings estimates for the S&P 500 X tech. That's the purple bars-- and then for the overall S&P 500 tech in blue. And you see a pretty significant gap here. The S&P 500 X tech has been pretty consistent, rising roughly about 11% to 12%. That first quarter '27 number is a base effect coming off of a elevated first quarter of '26 number here. 

But tech runs up and then starts to come back down, and it's still a pretty healthy premium to the broad market when you go all the way up to the second half of 2027. So I think maybe the question that begs answering here is, are these estimates right? Do we believe in them? Can we hit those numbers? 

And historically speaking, outside of the onset of recessions or emerging from a recession, analysts have been pretty good at getting these earnings estimate calls right. So that gives us a little bit of confidence here that maybe we actually are looking at something that continues to play out as such. 

This begs a little bit of a question, too, to shift gears now and talk about the potential for the bubble. And I think one of the things that we get asked quite a bit about, obviously, is, are we in a bubble when you start to think about the tech backdrop. And so earnings have certainly been eye-watering. I think it makes some sense to look at the valuation side of the story here. 

So what you're looking at here on the x-axis-- these are the forward net margins for tech. On the y-axis, the forward PE ratio. So you're simply plotting the relationship between these two characteristics going all the way back to 2005. And that gray line in the middle is the historical regression relationship with the plus or minus 1 standard deviation bands around that. 

And the thought process here is simple. As margins expand, earnings prospects increase. And that likely leads to an increase in return on equity. And return on equity and margins are key quality factors that investors want. As these qualities continue to improve, investors tend to be willing to pay a premium for these characteristics, as they provide stability and quality. 

So this really should come as no surprise that these lead higher multiples in the long term. And where we are today, given this historical relationship, you can argue that tech actually still looks fairly cheap relative to these fundamentals. And another way, I think, to look at this-- and I think this is somewhat interesting-- is that the market is now pricing semiconductors, for example, as a non-cyclical industry with a competitive moat. 

This implies that these margins might actually be structural, not necessarily cyclical. And why does this matter? Well, when a cyclical business increases margins, its valuation multiples either remain flat or even start to fall as the market prices in the natural mean-reverting nature of that backdrop. And when you think of a less cyclical company, and it's increasing margins, that market is rewarding it with a higher multiple, reflecting evidence of a competitive moat. 

So when the S&P 500 index becomes less cyclical-- and you can think of things like price-to-sales multiples or forward PEs, for example-- those should be rising more aggressively when margins are expanding. And with margins at an all-time high and the S&P 500 index really less cyclical than ever before-- and think of roughly 50% of that index we would consider maybe mature technology-- we should expect the index to have a record multiple valuation. 

So as we expect the S&P's composite to shift towards less cyclical sectors over the years and think of us moving towards tech and away from financials and energy, investors should be seeing those index multiples creeping higher as well. So I think to summarize this whole backdrop, as the composition of the S&P 500 becomes less cyclical and margins expand, higher multiples, I think, are completely justified. 

But if we want to move back to the old-school metrics here and just look at simple forward PEs, tech is still trading a bit higher than its 10-year average. That's what you're seeing here in that red dotted line. And it's still below the five-year average, which is that purple line here. So can we make the case that valuations are stretched from a forward PE perspective? It all comes back down to, beauty is in the eye of the beholder, I guess. 

But to us, tech does not look expensive. To call it a bubble, we'd like to see valuations getting much more stretched, much more ahead of itself. And we're just not seeing that. And so I think the last thing-- to really zoom out here and move back from not just the tech space, but to the overall S&P 500 backdrop-- earnings are really doing the bulk of the heavy lifting here. 

So if we look from left to right, assuming price returns are basically the sum of earnings per share growth and PE multiple growth, and if we start to look at decomposing those year-to-date returns here with that methodology, I think a few things are popping out that are interesting. The first one here-- five of the sectors within the S&P 500 have an earnings growth greater than 10%. I took the liberty of rounding discretionaries up to 10%. 

 

But the bottom line again here-- maybe the question to ask isn't necessarily if this is a bubble, but rather, how sustainable is this earnings cycle and that growth backdrop? A bubble, to me, implies a correction in prices that can be pretty dramatic and severe. A simple cooling of earnings-- and I'm stressing the cooling here, not a sudden stop-- this implies simply a slower level of price appreciation. So big difference here with very different implications for performance going forward. 

BRIAN HESS: So Jack, it sounds like you're comfortable with the idea that technology can continue to provide support to the market or be a leading sector with respect to the S&P 500. Is that right? 

JACK JANASIEWICZ: Yeah, 100% agree with that. Right. 

BRIAN HESS: That's good. Now, while AI has been a huge driver of stock market gains over the past couple of years, its influence hasn't ended there. And I think it'd be nice to see the other side of the coin, which is how it's affected the actual real economy. So if you could discuss this artificial intelligence CapEx boom and how that has spread throughout the economy, creating growth that's somewhat historically unprecedented, I think that'd be quite interesting for our viewers. And again, you talked about this a little bit with the stocks, but any comments on the sustainability of this CapEx cycle would be interesting, too. 

JACK JANASIEWICZ: Yeah, and I think this really ties back to what Garrett walked us through and really the key drivers here of the economic growth backdrop. And you're going to find a lot of this does rely on the overall AI CapEx story here. What you're looking at here, these are our contributions to real GDP growth here in the United States. 

So we're just looking at total spending on information processing equipment and software. We're processing that as part of the AI CapEx boom. And you can see this current spike here and compare that to what we saw back at the end of the 1990s. We're basically through those same levels. 

What's interesting, though, if I look at the last five quarters for real GDP growth and take that average, which is what we're using for a base metric here, that's been coming in just inside of 2%. Well, where are we today on that CapEx contribution to that growth rate? You're just inside of 1%. So almost half of what we're seeing for growth coming from the US economy is related to that AI CapEx story. 

And just how important is this CapEx spend? Well, you can see here the contribution from AI related to that CapEx spend is almost as strong as consumer spending. And we all know that adage where consumption drives roughly 70% of the growth backdrop here in the United States. That's a big driver. You can see this massive convergence here. So I think that puts some perspective around this. 

And implicitly, when we combine both AI CapEx and the consumption backdrop, you're basically accounting for almost all of that GDP growth over the last couple of quarters. So again, very, very important. I think another thing to think about here as well is that the AI spend and the appreciation that we've seen with regard to the stock market driven by the tech backdrop, it filters into that wealth effect. And I think that wealth effect also is very important. 

So if we start to see a rolling-over, not only are you going to see it with regards to slowing on that CapEx and revenue front, but the wealth effect could easily take a hit on the back of this. And so what you're looking at here, this is our attempt to show you how important it is. That blue line, that's the savings rate. The purple line, that's the household net worth to disposable income ratio inverted. 

And you can see here the savings rate's falling while the net worth is actually rising. And so if you think about consumption, which has remained pretty firm for the US consumer over the last couple of quarters, it's been pretty impressive. So I think it's possible that consumers here have been tapping stock market gains to finance some of those purchases more recently, meaning that the drop in the savings rate that a lot of people might be pointing to as somewhat of an alarming feature of consumption going forward, maybe we're overestimating the impact there because we're seeing that wealth effect actually act as a substitute here. 

So I think at the margin, that's something important in this last chart to go through. You're just simply looking at the household equity exposure. It's sitting at record highs. And so I think that gives us a sense as to how big and important this AI theme is to the overall economy. And it actually plays its way all the way back to the potential for the consumer being supported as well going forward here. 

BRIAN HESS: It sounds like a lot hinges on this AI theme, so let's hope your comments about the sustainability of it and the valuations and the earnings expectations are all right because otherwise there could be some collateral damage, it sounds like. And I think who's probably worried about the potential for collateral damage if there's some slowing in this segment of the market is the Federal Reserve, where just this week, we witnessed Fed Chairman Warsh's inaugural meeting. 

We had a more streamlined statement, a bit of a more guarded press conference. So there's a new era at the Fed, and I think it'd be interesting to spend a few minutes talking about that. So Garrett, what are your thoughts on the new chairman, this week's meeting, and the potential push-and-pull of market pricing for what the Fed's expected to do, versus your thoughts on what's likely to actually happen? 

GARRETT MELSON: Sure. So there's a lot to unpack there. Maybe we'll just start with a couple quick thoughts in terms of Warsh's inaugural presser. You alluded to it. Some pretty big changes in terms of the statement being shortened. I'd say also what was notable was the removal of the balance of risks from the statement as well. 

 

Warsh straight-out confirmed that he withheld this dot. And again, that's very much in line with what he's been alluding to. He wants to move away from forward guidance, and so that's certainly one way to do it. And I'd say honestly, the rest of the press conference aligned with that as well-- a whole lot of dancing around and not actually answering any sort of questions about where policy is likely headed here. 

And so I think that raises a couple of things. First, what we did hear about was the five task forces that Warsh is going to be laying out here. And so that does imply that we will be seeing some pretty material changes here coming down the pipe. Maybe that's setting up for Jackson Hole at the end of the summer. Maybe it's a little bit further out. But expect to certainly see some continued changes in terms of communication, some of the data that they're using, as well as the inflation framework here. So a lot that's still somewhat unanswered. 

But I think the bigger story, and certainly what had some implications for the reaction yesterday, is they certainly are moving away from this idea of forward guidance. And that's totally OK to do. We certainly have seen markets do just fine without that forward guidance hand-holding here. But I think what was missing and what markets latched onto-- not just the hawkish shift in the dotplot and the expectations, but it was the fact that we didn't really get any sort of discussion about the reaction function. And that's, I think, really the key here. 

So in the absence of any sort of update from Warsh about what his reaction function looks like, I think the default for the market really should be to fall back on the rest of the committee. So again, look at the dotplot. Very conditional. Warsh downplayed that, much as the way that Powell has over the years as well. So don't take that as certainty here. 

But I think what's going to be more useful, probably, in the near-term is-- your guiding light, I think, has generally been Governor Waller over the last couple of years. He's been at the leading edge for messaging. He's probably a little bit more reflective of where the median voice sits. And I think that's probably going to continue to be a pretty strong voice if you're not going to be getting a whole lot out of the Fed Chair here, as he wants to ratchet back on that communications here. 

So a lot that remains up in the air here. But I think the key that markets are latching onto is no forward guidance. And that ultimately opens up the door to potentially some more volatility because policy might be adjusting a little bit more rapidly at inflection points, and you don't have that hand-holding to drive that inertia in policymaking from the Fed here. So something that we'll have to keep an eye on as we move forward here. 

But in terms of how I think policy plays out for the balance of the year, a couple thoughts. And we'll touch on some of the moves in rates markets as well in this. The first one that I think is helpful-- obviously, a lot of questions about Fed credibility. I think Warsh addressed that pretty well during his presser. But there certainly doesn't seem to be much concern about Fed credibility when you look at rates markets. 

Just looking at what's driving the repricing in yields out the curve-- so this is your 10-year decomposition of the 10-year yield. You have your nominal yield in light blue. You have your real yields in purple, and then the break-even, which is the market-implied inflation expectations in green. Well, pretty clear that there's no concern around the inflation mandate. You've seen this roll all the way back to where we started at before the war broke out in Iran.

 

So that curve that the short-rate market is pricing out here as you go out to the end of 2031-- and obviously a lot of uncertainty out the curve, but you're looking at anywhere from, say, 40 basis points to over 120 basis point shift in market expectations. So it's that repricing of the Fed's policy path that's actually bleeding its way out the curve and pushing upward pressure on the rates backdrop here. So not really any concern about Fed credibility or desire to get inflation back to target here. 

But I think the big story is thinking about what the Fed is likely to do versus what's actually priced into the market. And I think there is an opportunity there. It might be a little bit early. But when you think about the Fed's ability to address inflation, they have one fairly blunt tool, which is rates, and that really affects aggregate demand. 

Well, if aggregate demand isn't really the key driving force behind inflation, then it's not really going to do a whole lot here. And I think the risk is, when you decompose core PCE inflation, the Fed's preferred target, you can see that acyclical components are really what's driving that reacceleration here. What does acyclical mean? Basically those are portions of the inflation basket that are less sensitive to overall economic conditions, less sensitive to aggregate demand. And they're more a byproduct of industry-specific, idiosyncratic factors here. 

So that suggests that policy is somewhat ill-suited to address those inflation risks. So the risk is here, you're dealing with the unwind of two supply shocks. Still dealing with the effects from tariffs last year that will continue to roll off this year. So that's somewhat of a reason to be optimistic on the inflation backdrop. And then on top of that, you're dealing with what is now-- you're starting to see some reversals in an energy shock. Again, another supply shock that's fairly short-lived that you really can't address with Fed policy. 

So I think it suggests that the Fed's going to be a little bit more measured. And I think this speaks to why we think the bar is still very high to shift from a neutral stance to maybe an even modestly tightening bias to actually implementing those hikes. That's a whole different ballgame. And I think this is one of the reasons to push back against why the Fed will actually follow through on hikes. 

Another reason is-- you talk about AI. It's having effects in equities markets. It's also having effects within the economy, but not just on the growth side, but in inflation as well. This is a little bit of a busy chart, but it's important to know what measures you're talking about. And we've talked a couple times over the past few years about this idea of a wedge between the two inflation measures, CPI and PCE. 

Through the disinflationary process, post the pandemic, you had this wedge where core PCE was lower than core CPI. And that's the historical norm. But core CPI was overstating the degree of inflation stickiness. Now that's completely flipped on its head. So the average gap or wedge is core CPI running about 30 basis points above PCE. You're now seeing core PCE running almost 60 basis points above core CPI. 

So the Fed's preferred measure is somewhat overstating inflation stickiness. And part of that is this purple bar, which is computer accessories and software. So you can see that AI effect bleeding into a category that basically has a 0 weighting in CPI, but it's about a 1.25% weight within PCE. It's gone up about 60% annualized over the last six months here. So you can see the effect, adding about 17 basis points to that wedge. 

Why does this matter? Well, I think there's reason to think that PCE is somewhat overstating inflation stickiness right now. But it does make the Fed's job and communication a little bit challenging. So they might not be hiking, but this wouldn't be a reason to push back against any expectations of cuts in the near future here. It's simply going to be too hard from where inflation currently is and specifically relative to where their preferred measure of inflation is. 

The last thing to just keep in mind is, think about that distribution of outcomes once again. That's how we like to think about market pricing and potential outcomes moving forward. It's not just about that spot pricing. And right here, you're looking at spot pricing for rate hikes. So you've moved again from an environment where you’re pricing in April-- policy rate that was going to be at least 25 basis points lower. 

Now you're looking at one that's going to be about 25 basis points higher here. But that's a function of the entire curve. So you can see how we basically clipped that entire left tail. We've fattened out that right tail, and the entire distribution has shifted to the right. 

I think the key theme over the summer months is going to be, the markets are probably going to price in a tightening cycle here. The Fed doesn't usually just hike once and done. They tend to hike a couple of times here. So a reasonable base case is probably unwinding the insurance cuts from last year, so that's three hikes that you see. Market's already pricing in a full hike by October of this year. And so you can see just how hawkish the markets have gotten here. 

So to us, I think the risk is, markets price in that rate hike cycle that ultimately doesn't come to fruition, one, because inflation trends under the surface are somewhat more encouraging. Two, because that bar is just so high to actually shift into that tightening bias here. And as a result, I think you're probably going to see that the market actually might do the Fed's bidding in here by tightening policy to a degree where the Fed can hold steady and ultimately doesn't need to follow through on those hikes that are priced. And ultimately, that just sets an opportunity up for investors, maybe later this year, to lean into duration. It's still a little bit early on that front, but something to keep on the back burner. 

BRIAN HESS: So it sounds like you're not expecting a big move one way or the other from the Fed. Sure, they might hike once or twice, but even that's questionable, and that's the upside of what you'd be expecting. And then in terms of cuts, they seem like a distant prospect for the time being as well. Is that a fair assessment? 

GARRETT MELSON: Yeah, I think that's fair. And I think what would get us to flip our view, in terms of actually following through on hikes-- it boils down to the labor market. If we really do start to see clear and convincing signs of tightening there-- and that would really come down to the wage growth measures starting to re-accelerate. Until that happens, I think it's a market that prices in a tightening cycle that ultimately the Fed doesn't have to deliver on. 

 

BRIAN HESS: Yeah, I guess you can think about it as maybe creating an opportunity to buy some less expensive insurance if we get a move higher and longer-term rates, since you're not expecting it to be a runaway move higher. But there's still a good chance that if the economy were to roll over aggressively against our forecast, that that long end would provide a hedge. So maybe the opportunity to purchase some cheaper insurance. 

GARRETT MELSON: Exactly. 

BRIAN HESS: And not only are you a fixed-income guru, but you've also been doing a lot of work on the stock market, particularly with factors. And so let's talk about that next. And what have you learned from this research that might be interesting for our viewers today? 

GARRETT MELSON: Sure. So I'll keep this one fairly brief, but I think our followers will know well that we do quite a bit of work and focus quite a bit in terms of market structure. I think that has a pretty important bearing on markets. And increasingly, some of these systematic strategies, trend followers, vol-targeting strategies, they're a very important driver of the marginal bid and uptrends, and that can help dampen volatility. But they can also exacerbate reversals here. 

And so I think that just speaks to the fact that fundamentals really do matter. Everything I walked through is important. But technicals can sometimes be more important in some of these near-term moves here. So a couple of things I just wanted to point out, in terms of some of the moves we've seen and some of the things to keep on the radar as we move into the summer months. 

And the first one is the momentum market. So Jack walked through the fundamental support and case for the tech sector. But increasingly, that's probably the one area of the market that has gotten very crowded here. And so tech and the AI trade has basically become this huge momentum trade.

And there's nothing like prices moving higher to draw in investors. People chase returns. And so any area or pocket of the market that starts going up, you tend to see this momentum effect where that starts to draw in more assets, and that just feeds on itself. But increasingly what we've seen is that momentum factor-- so basically stocks that are going higher continue going higher-- it's also become wrapped up in the beta trade. 

And so what you're looking at here is long/short baskets of the beta and momentum factors. So just trying to exclusively zero in on those stocks that have high beta or high momentum characteristics and looking at their rolling returns over a 65-day window, so basically a three-month rolling trading period. And what you can see here on the percentile rank here, you're basically in that top percentile of historical performance. 

So they've gone vertical and parabolic, basically, out of the March lows. And you've gotten to a point where basically you haven't ever seen returns be stronger than they've been over the last three months in the recorded history for these factor baskets here. So you've got to a point where that sentiment and that positioning was pretty crowded. 

And that doesn't mean that you need to necessarily immediately revert, but it does give you the conditions where, if you start to get some sort of catalyst, you can see that reversal set in, and it can be pretty sharp. And I think that's exactly what we got just about two Fridays ago on June 5th in the wake of the payrolls report. A strong upside surprise on the payrolls print, and that triggered a pretty sharp reversal here. 

But I think what's really encouraging is that you're seeing that rotation and unwind, but it's not a liquidation. It's just a rotation. So what I did here is just boil down S&P 500 into deciles, so 10 different groups based off of their return year-to-date through the beginning of June, basically, on June 2nd, which was the high in the NASDAQ 100. 

And pretty obvious when you do so, you're going to have a linear or stack ranking here, highest basket having returns over 100% on average, the lowest basket negative 30%, 31% on average. But the really interesting part of this is, now take those baskets forward through basically that momentum unwind and then the recovery and stabilization that we've seen since. All of your biggest winners through that unwind were then your biggest losers, but you didn't really see all that selling pressure dip into the rest of the market. 

So you actually saw this dynamic where money stayed within the market but just rotated into relative laggards here. And that's been the theme we've seen over the last couple of years. The AI trade is just so large, so robust, it just sucks the oxygen out of the rest of the market. And so when the AI trade is rocking, it just is going up basically alone and leaves everything else behind. 

But when it takes a little bit of a breather, you start to see those assets rotate back into the laggards, and that's exactly what we've got. And then actually since, you've seen a little bit of stabilization, obviously, on the optimism of a resolution to the conflict in Iran. Well, now you're back off to the races with those momentum leaders taking the reins again here. So that's an encouraging sign that actually helps to keep downside somewhat limited, to the extent that you're not seeing unwinds actually moving into cash, they’rejust rotating to laggards. And I think a big reason for that-- getting back to the market structure-- is the fact that there's this dispersion trade that is somewhat under-discussed, under the surface. Everybody talks about the momentum trade, but there's a lot of trade volume within dispersion trades, which is basically trying to monetize the volatility differences between the index and then the constituents of that index. 

So the common way to implement it is, go long the volatility on individual names and then short-index vol. And what does that do? That just pushes down correlations. And that's what you're looking at, the blue line here, is the S&P 500 implied correlation index. 

So you're basically down at the lows on record here. 

And the key here is, when do you see really sharp unwinds that start to spiral? It's when you get a tick higher in both volatility as well as correlation. So big risk-off moves, correlations start to move up towards 1, and then volatility spikes. And so what happens is you get the unwind of the dispersion trade. What do you have to do there to unwind the trade? Well, now you have to sell the index, the underlying vol, on those constituents, and you're buying back volatility on the index. 

So now you're pushing up the VIX. What does that do to those systematic strategies? Well, vol goes higher, those vol-targeting strategies have to degross mechanically. And so you can see how the dispersion trade can unlock another wave of gross and net de-risking for a lot of these systematic strategies. So there's a reason I touch on that market structure. I think it's really intertwined into a lot of these technical factors and some of these moves here. 

The last thing to keep in mind, though, is you didn't get that move. You got a tick-up on VIX, but you didn't really get that move in correlation. And I think that's really reflective of that dispersion trade's alive and well, and that's helping to keep that rotation market intact. And to that point, on those systematic strategies here, these are just one of those cohorts here. This is the vol-targeting cohort. 

You can take a look at the S&P 500 in purple. You have realized volatility in the light purple here. And then you have essentially the equity exposure for these vol-targeting strategies in the light blue. So yes, you saw a little bit of a reduction in that equity exposure. 

But again, because you didn't have that rising correlation to a material degree on top of the VIX backing up, it keeps this somewhat rotational market, which ultimately, I think, just keeps any sort of these unwinds and selloffs from really morphing into anything more of a waterfall decline in here. So something to keep an eye on. But those tend to happen with big macro shifts. And for now, conditions seem to be OK, and that rotational market should keep things, in terms of the downside, supported in terms of any sort of pullbacks in here. 

BRIAN HESS: Thanks, Garrett. And a few of those charts were rather complicated with a lot going on. So anybody who's curious to take some time on their own to look at them and analyze what's really being illustrated there, the slides from our presentation today will be available. It takes us a few days to get compliance approval. And thanks, Garrett, for that interesting, few different perspectives on looking at stock market drivers. 

Now, Jack, we're going to stick with the stock market for our last segment. And you have some really interesting charts as well, which I'm eager to get to. But before we do that, we've got one client question. And it's a question about the summary of economic projections, which the new Chairman, Warsh, has made some comments that maybe he doesn't think there's a need to continue with that practice. 

And so the question has to do with whether or not we think the SEP is likely to stick around, and maybe if there would be any cost to eliminating it. What are your thoughts, either one of you? But Jack, you're up next, if you want to take it. 

JACK JANASIEWICZ: Yeah, it's interesting because Garrett and I were just talking about this this morning. And quite frankly, I think we're in the camp that if they got rid of it, there'd be no harm, no foul. I think we get more questions and confusion that comes from the SEP than it's worth, and so I think the market puts a little bit too much stock in what I consider to be basically 18 forecasts or guesstimates. And I think the market treats it too much as Bible in here. So we wouldn't be, I think, upset if they replaced the dots. But at the same time, I think having some discussions after some of these announcements would also be better suited to replace them. 

BRIAN HESS: And it does seem like the practice is at risk if there's a decent chance it's going away, I would say, right? 

JACK JANASIEWICZ: Yeah, I think that's right. 

BRIAN HESS: All right, well now let's go back to the stock market and Jack, wrap us up. We've got some comments on supply risk mainly, after SpaceX's impressive debut this week, and also some bellwethers you're watching to guide for that sustainability of this AI mega-bull market we've been dealing with. Anything that we can look at to give us a sense for the trajectory and if fundamentals are still moving in the right direction, I think, is critical to watch right now and helpful. So let's dive into these last few charts. 

JACK JANASIEWICZ: Yeah, a lot of comments that we're receiving, trying to draw some comparisons between the late 1990s, 2000s, and what we're seeing with regard to some of these big, monster IPOs that are coming and obviously, SpaceX being the first one to market here. But I just went back. I pulled a couple of the bigger IPOs that have happened in history and tried to mark them, where they were with regard to the overall S&P 500. 

It looks like my chart here for the 2000s might have drifted to the right a little bit because some of those dots are hanging out in the middle of nowhere, but I think you get the point here. It's a mixed bag. I think you've seen IPOs come, and the markets continued to grind higher. We've had some IPOs where the markets actually sold off and head lower. I think at the end of the day, a lot of this is really a function of the economic backdrop when these guys come to market here. 

So I'm not necessarily sure that some of these IPOs call market tops. I'm a little bit skeptical of that sort of analysis. We've talked about the bubble backdrop. One of the things that we were paying attention to with the SpaceX launch was, how would it trade in the first day. Because I think if you go back in history and look-- and I'm specifically pointing here to this middle bar, 1999-- the mean first-day return for IPOs that launched during that year was 71%. 

And so when you start to think about that excess speculation, that excess froth in the market, this is maybe one way we would try to measure what some of these recent IPOs might look like. And if we look at what happened for that first day of launch for the SpaceX IPO, it was up just a little over 19%. So in line with those numbers between '97 and '98, which I think we're a little bit early in terms of making that call for a market top. So had we seen a huge pop, something akin to what we saw in '99, I'd be a little bit more concerned on that backdrop, but we just didn't see that. 

What's also interesting-- and a hat tip to our friends over at Renmark. They put this together. We thought it was a great chart to follow because if you go back and look at some of the select 20 IPOs, if you will, since 1995, and what their 12-month returns look like, all we simply did was equally weight those returns and how did they do over the subsequent 365 days. 

And what's interesting-- you can see here in the purple line-- the median return basically was down for the full year for these new IPO launches. Obviously the average is a little bit different, so you had a few really successful IPOs. But at the end of the day here, I think what it's telling us is, historically speaking, sometimes these IPOs actually trade down, and some of the drawdowns that we've seen from a historical perspective can actually exceed 50% here. 

So if you're waiting, maybe, for a chance to get your shares of SpaceX, maybe you'll get them at a cheaper price at some point over the course of the year. History probably stands on your side for that backdrop. So a little bit of an interesting sequence in terms of how historicals have played out for IPOs. 

But I think the bigger concern that we might point to is really when these restricted or these shares unlock-- SpaceX really only floated something a little less than 5% of that total market cap-- you still have the remaining balance that's set to unlock it. And a little bit of a unique backdrop, the structure, in terms of that unlock schedule. And I've tried to map it out here. But the point being is, we move to the back half of the year. That's when a lot of these unlocked shares can come and be redeemed. 

And so when you start talking about the potential for Anthropic, for OpenAI, we're hearing Meta talk of potentially issuing, Google issuing, you start to lump all these things together, and then you also start to factor in the unlock shares coming to market. And again, this is something closer to the fourth quarter of this year. It'll be interesting whether that supply will be easily absorbed. 

And again, I go back to the idea that the economic backdrop matters here. So if the backdrop is supportive for risk-taking, the market probably ends up taking it down pretty well. But if it's a little bit struggling with some uncertainty for whatever reason, that can create a little bit of indigestion. So one of the things we're flagging for the back half of the year is this unlock schedule coming from SpaceX. 

But again, getting back to that market top, it's really the number-one question that I'm getting from clients over the last couple of weeks. And so a couple of thoughts here, I think, worth offering up on the new issue market. First of all, I don't think we would consider the IPO activity right now extreme, and I don't think we're really expecting it to reach extremes. 

 

And it may be we get a little bit of exaggeration with this because these are very high-profile names coming to market. But again, just put some numbers into context here. From an estimated capital raise, we're thinking in estimates roughly about $700 billion for this year. To put some context behind this, that's roughly about 1% of the total Russell 3000 market cap. 

And so if we go back from the period of 2015 to 2019, when you have a total issuance worth roughly 1% of the market, that's roughly in line with that period that we're talking about there. So 1% of the Russell 3000 is really not out of the ordinary, so from an absolute size perspective, nothing that massive. And the last piece maybe people tend to forget, these estimated gross buybacks, that's north of $1 trillion this year that's expected. So this should also help to offset some of that supply. 

But as I keep alluding to here, timing does matter to everything. We do flag the risks for further supply set to come later this year with the likes of OpenAI and Anthropic as well as the unlocked shares that we were just talking about. But again, as we push farther into the year, the balance of supply and demand does become a little bit more challenged, so it's something worth watching here. But again, it's the market backdrop that I think makes a big difference. 

So what are we paying attention to here? What are some of the flags that potentially could derail that AI trade? Well, the first one here, it's really contingent on these CapEx numbers. And so paying attention to the forward guidance coming from these hyperscalers, I think, is critical. And when we look at the estimates going all the way out to 2030, they still remain fairly robust. If we start to see these numbers rolling over, that certainly is going to send up some red flags. But for now, paying attention to what these hyperscalers are announcing for planned investment so far continues to be very robust. 

The last one here is-- and we're going to simply go around the Horn-- these are all just trying to find real-time indicators from the market and what's being priced in. So if we start from the top left there, looking at GPU rental prices, obviously if those prices start to come down, that questions maybe some of the demand for GPUs going forward. One thing to think about, though, is as some of these chips move down the line, in terms of their maturity, it wouldn't be surprising if they roll off and just become replaced with other ones. So obviously we would then start to replace what we're paying attention to here. But bottom line, GPU rental prices holding up just fine. 

The bottom left, we're looking at semiconductor exports coming from Korea, and also looking at basically what their export prices are as a function of what they are being reported in the PPI for Korea. And again, still moving up and to the right. So we're still seeing plenty of exports for semis happening here and prices continuing to still grind higher. 

Bottom right, the same thing here. I'm looking again at additional factors coming from the Korean export market with regard to flash memory, semiconductors, and DRAM, all of them pushing higher. So if we start to see any softening in these numbers, that throws up some red flags for us as well. And then the top right, the silicon data large language model expenditure index, this is something new that's popped out. This is basically a benchmark that measures the use-weighted average price that the market's paying per millions of tokens used. And I think this actually acts as a pretty good proxy for the market's marginal willingness-to-pay for AI capabilities. 

So the key here, this actually, I think, is a pretty good reflection of user behavior. When the enterprises shift from cheap open-weight models to premium closed-source models, this index should actually be going up. But when we start to see users substitute expensive reasoning models with cost-effective alternatives, that index falls. 

So what is it telling us right now? Maybe we're seeing a little bit of a shift underneath the surface, where the demand for AI is becoming a little bit more price-sensitive. We've heard all these stories about the massive run-ups people have had in terms of how much they've spent, not knowing that they've spent this kind of money on tokens. OpenAI is now talking about price cuts for tokens. 

So again, the bottom line here, firms don't want to cut prices aggressively when they're seeing exponential demand, so we're seeing a little bit of softening there. It throws up some questions to look out. So the worry here-- are we really at the first stages of maybe seeing AI being commoditized? The technology is definitely generating significant amounts of revenue here that require plenty of compute, but competition and pricing pressures are also now starting to constrain some of this. And how much of that top-line revenue actually reaches the bottom line? 

Big question. That certainly might keep a little bit of a lid on margins going forward here. But the bigger impact-- are we starting to see that market bifurcated? What do I mean by that? Well, maybe we're only going to have a handful of upside frontier names really becoming the key drivers here. 

While those use cases are certainly narrowing to those who can effectively monetize those kind of issues, those players are few and far in between, where the rest of us are simply leveraging cheaper, simpler, and more efficient models good enough for what we need to do on an everyday basis to enhance our productivity. So in a sense, a two-tiered ecosystem starts to arrive, where those who can turn expensive inferencing into a profitable venture, good for them, and then those who need and use simpler versions to execute everyday tasks, that's what's left here. 

But I think regardless of the backdrop we're talking about here, we are seeing no signs of slowing in this space, at least not yet, and just seeing shifting crosscurrents beneath the surface. So it's certainly something worth following here. It's an important theme, like we just walked through, for just the economy. Tech for the broader markets. There's certainly plenty of things that are pivoting, and really the crutch of what the US economy is being built upon. So definitely worth paying attention to these sort of signals, and we're trying to keep our best to stay on top of all this stuff. 

BRIAN HESS: Thanks, Jack. You've definitely left us with plenty to think about and some concrete metrics to track. So I think, if nothing else, we're going to take away a lot of ways to think about the developments within AI and analyze potential winners and losers, like you were just going through, and some of those data points you highlighted will be important to track. So thank you for running through all that. 

So I do want to thank Jack and Garrett for their insightful commentary and their charts, but also need to thank everyone for joining the webinar. It's great to have you here. We appreciate you coming out to listen to what we have to say. In the meantime, we wish everyone well, and we hope to see you next quarter. So thanks very much.