Intro
Jon Turek, founder of JST Advisors, reads Kevin Warsh's first ten weeks at the Fed, explains why 30-year Treasury yields are at multi-year highs, and works through what the Treasury's decision to buy back bonds when nothing was broken does to the rates, dollar and AI capex setup. Bilal Hafeez presses him on wages, Europe and the Strait of Hormuz.
Guest: Jon Turek, founder and CEO of JST Advisors, previously a portfolio manager at Brevan Howard and an analyst at Moore Capital
Host: Bilal Hafeez, head of strategy at Macro Hive
Published: 28 August 2026 on Macro Hive Conversations · recorded 20 August 2026
Episode page | 51 min
Key Takeaways
Warsh may be an owl rather than a hawk or a dove, and he may not control the committee
"I kind of been, you know, pushing this idea that maybe Warsh is more of an owl"
"he doesn't really sort of have a strong enough pull over the committees"
Warsh solves for political optimization, which makes the regional bank presidents the thing to watch
Waller and Powell are the board's intellectual leadership; Logan, Hammack and Kashkari are the strong voices outside it
The Treasury bought back bonds with nothing visibly broken, and the context was the whole story
"the Treasury seemingly decided that they didn't really like the number on the screen in terms of the level"
Swap spreads had tightened on the month and there was no on-the-run problem
Doing that invites the market to test the level rather than respect it
"you end up making the market a little bit more of a floor than a ceiling"
If Washington keeps the financing open, the AI build-out gets more durable and terminal rates can price higher
"you could actually price terminal rates higher in the US because the chances of something, you know, cutting off this positive growth impulse has actually gone down"
Both tails of the AI capex distribution have gotten fatter, not just the right one
"capex busts are painful and this is turning out to be the biggest capex cycle that we've seen"
Hyperscalers are spending out of fear, so a higher cost of capital does not stop them
"why would 50 basis points in Google's 10-year or 30-year bond is not changing the calculus for them"
Yields are high because a global investment call is outrunning global savings
"the free cash flow is now gone because they're spending so much and that means they have to come to market to get the money"
The dollar's downside is now the bigger tail, because the buyback path leads back toward Fed balance sheet growth
"I don't know if that's technically QE but that sounds like QE to me"
The ECB's reaction function is asymmetric and hawkish, which puts a floor under European yields
Germany is running a 4% budget deficit that gets mechanically bigger as growth fails to show up
A stuck oil market is worse for inflation than a spike, because it lingers
For anyone entering markets now, the friction on building and sharing expertise has never been lower
Warsh Came In Promising Cuts and Delivered Inflation Credibility
Hafeez opens by putting the market's confusion to him: Warsh arrived as a rockstar, and after his most recent meeting people decided he might not be one, largely because he will not give forward guidance
Turek's starting point is that the initial conditions were all dovish. The president has been an advocate of lower rates, Warsh appeared to get the job on the promise of lower rates, and his confirmation testimony leaned into alternate inflation measures and the supply-side potential of AI
Then the first press conference went the other way — Turek describes it as the meeting where the Fed "kind of had this first press conference which was no inflation tolerance"
The pricing swing was violent. Going into June, September carried eight or nine basis points of hikes; leaving June, July carried eight or nine basis points and "September almost priced a full hike"
The July hike never came, and Turek treats that as the tell. Warsh had the option, the market was in that lane, and three members dissented; the minutes released the night before this conversation said "several participants favored" a hike
His read of the sequence is blunt: "the sort of inflation credibility message that he tried to assuage markets of in June is not earnest and it's not authentic"
The Owl Analogy, and Why Warsh Is Both
Turek's frame comes from Christine Lagarde's own introduction at the ECB, where she said "she's not a hawk, she's not a dove, she's an owl"
"I kind of been, you know, pushing this idea that maybe Warsh is more of an owl," he says, adding that "he doesn't really sort of have a strong enough pull over the committees" yet
Asked which side Warsh actually sits on, Turek's answer is both: "I think he really does believe in sort of the productivity gains to come and that is going to be a structural disinflationary force"
At the same time, "I also think he's uncomfortable with the level that you know core PCE at 3.3%" — and Turek thinks Warsh badly does not want to be the chair who repeats 2021
The caveat Turek keeps attaching: "he's only been in the job for you know 9 10 weeks"
Who to Watch on the Committee Now
Hafeez asks the practical consequence: does the owl reading mean watching every other influential member and working out who moves him
Turek says the idea that Waller becomes less influential because his style is antithetical to Warsh's is not true, and that paradoxically such people become more important
The Lagarde precedent is his evidence — first Philip Lane's remarks were the leading indicator for how Lagarde would sound, and more recently it has been Isabel Schnabel, whose message seems to have her ear
The board of governors is thin, in Turek's description: "you have Waller and Powell as sort of the intellectual leadership" and not much else, which is why "I think the regional bank presidents do become increasingly important"
He names the strong voices: Lorie Logan in Dallas, Beth Hammack in Cleveland, Neel Kashkari in Minneapolis
"I think he thinks in political optimization," Turek says, rather than the model conviction of a Bernanke or the gut of a Powell or a Greenspan
The Style Problem, and What Jackson Hole Could Fix
On the man himself, Turek says "he does come across as a very slick politician," to which Hafeez adds "Little too slick"
Turek's follow-up is that "after Fed chair, he's going to run for president or something" — it really does feel like that when he watches him present
The market's tolerance is the open question. June read as refreshing; July did not. "there's a difference between, you know, having a framework, having a reaction function, and like spoon feeding us"
Most people had written Jackson Hole off, in Turek's account, as another performance where "he'll say nothing task forces" — but he thinks it is a good outlet to put more meat on the bone, because Warsh is talking to colleagues and "He's speaking to a bunch of his colleagues" without a Nick Timiraos follow-up
What Turek does not expect is Waller-style explicit rules. He says it is not going to be anything "Odyssean or Delphic like that"
He thinks the curve steepened after the July press conference in a way that was negative feedback both narratively and in markets, and that the market will not keep tolerating "we're going to get inflation back down. How? Can't tell you what's going on."
A Different Nominal Growth Backdrop Than Last Year
Turek contrasts today with the themes of his last appearance: tariff inflation that never showed up, very low energy prices, an unemployment rate nudging up, and three Fed cuts toward the end of the year
Now it looks inverted — "it's certainly not a $50 to $60 energy, you know, backdrop," and "The unemployment rate seems biased lower, not biased higher"
Housing looked headed for a self-fulfilling negative spiral late last year; it has stabilized, though sentiment is poor and it is not strong
On inflation, "the Fed had in their December dots had '26 core PCE finishing the year at 2.2%" — and "it seems like it's going to be another year at three or plus"
The growth picture is uneven but still producing a lot of nominal growth. He points to the quarter's private domestic final purchases over the headline: "PDFP printed I think it was you know close to 4% in the second quarter", which he says spoke a lot louder than 1% GDP or negative retail sales
The composition has narrowed because so much of it is AI capex, in an economy people are used to reading through consumption
Turek finds it hard to worry much about a US downside scenario from inside the economy — unemployment is low, consumption is roughly trend, housing is weak but not getting weaker — and "we keep rerating go forward AI capex"
Why a Falling Unemployment Rate Isn't Lifting Wages Yet
Hafeez puts the puzzle directly, asking "with the unemployment rate falling, I mean, why is wage growth not stronger", and offers his own frame: "is this like a Japan scenario where policy makers are in reality really trying to target wage growth rather than inflation"
Turek thinks it is too soon to conclude anything, since the fall is only a couple of months old
He grants the evidence on the other side: "average hourly earnings in the last NFP report came down from 3 and a half% to 3.2%", and aggregate weekly payrolls, an indicator he likes for where nominal growth is going, have been slowing
The mechanism is what makes him cautious. For the first half of the year "the unemployment rate was very stuck at around 4.3, 4.4" while payroll growth accelerated; the rate only started falling as monthly payrolls went negative
His conclusion: "the unemployment rate of the last couple months has been falling more from a supply side than the demand side" — a labor market tightening that is "The taking the unemployment rate, unemployment rate down 30 basis points over the last few months" without the usual wage transmission
He also flags the Fed's preferred wage tracker moving the other way: "Q2 was a 0.9" on the employment cost index, after trending lower
The contrast he draws is with the clean version of the mechanism — "we jammed 150,000 jobs", pulled hiring forward and raised wages to do it
Both Tails of the AI Capex Distribution Have Gotten Fatter
Hafeez checks whether he read optimism into the subtext correctly. Turek confirms it and immediately qualifies it, saying he would define his optimism as "the tails in both parts of the distribution have gone up"
The spend is powerful because of the accounting. One company's capital expenditure is another's revenue — "it's immediately booked as you know, profit and revenue for someone"
What drives the hyperscalers, in Turek's reading, is the left tail rather than the right. The spend comes from a fear of being out of business, which is why credit market indigestion does not slow it: "why would 50 basis points in Google's 10-year or 30-year bond is not changing the calculus for them"
The calculus he ascribes to them: "if they don't do this the business could be gone and if they do this it could be a 10x on return"
He rejects the 2022 Meta comparison clients keep pitching him. Meta was overinvested in VR, cash flow negative and the stock fell 60 or 70% before Zuckerberg's stopout and the year of efficiency — but Meta did not view VR as existential, and Turek says today's spenders do
On the guides, most hyperscalers said in Q2 earnings that "'27 is going to be significantly higher than '26", with "Meta was the only one that didn't really give a guide for '27"
"over the next 12 months, it seems reasonable there's going to be over a trillion dollars spent between the hyperscalers. I mean, these are fiscal programs." His benchmark for that: "Germany announced a trillion dollar fiscal program in March of last year", and it was a very big deal for markets
He frames the whole thing as "these like two actors in the economy who are commanding a massive amount of capital" who both believe they have no choice — governments and AI
The warning attached to the optimism is about the shape of any slowdown. "you don't usually slow from you know a trillion dollars of spend to 800 billion it usually goes from a trillion to 200" — and "capex busts are painful and this is turning out to be the biggest capex cycle that we've seen"
Why 30-Year Yields Are at Multi-Year Highs
Hafeez sets it up with the levels: "if we look at the 30 year it's moving to like multi-year highs I think the highest in almost 10 years or so", with the 10-year moving up as well
Turek's first reason is global and structural — "the market wants to price upward sloping curves everywhere", in the US, Japan, Europe, the UK and even Switzerland, because inflation volatility has risen and is asymmetric above central bank targets
The credibility question is real but secondary for him: "You're about to have your third year in a row of 3% core. It is fair to ask you know what is actually the inflation target."
The big thing is the reversal of the 2010s. That decade had lots of savings and little investment, governments contracting rather than expanding, and companies geared toward buybacks and free cash flow — reflected in low nominal growth
Now fiscal is expanding everywhere five years after Covid, on top of a private capital spending cycle he compares in relative terms to shale in the early 2010s and to US housing in the 2000s
The change that finally moved yields was how the spend is funded. Google was making money and investing that money; now "the free cash flow is now gone because they're spending so much and that means they have to come to market to get the money" — and the amount keeps rising
His arithmetic for a client who disagreed: in 2024 and 2025 "the capex bill was 200 billion that was 400 billion" in round numbers, and that was the period that took free cash flow to zero
Using Richard Koo's balance sheet framing, he says the public and private sectors are now on the same side. The savings left to meet the call are East Asian households — "It's not American households anymore. The savings rate has gone down." — and European savings, which German fiscal policy is already claiming
The incremental point he thinks is notable: "budget deficits globally are still, you know, at best they're staying and at worst, you know, they're going materially wider"
The Treasury Buyback Was Small, and That Wasn't the Point
Turek calls it a big deal while conceding the numbers are not — "taking buyback operations from two billion to four billion", though "if you annualize those numbers they're actually not nothing"
The context is what made it notable. "it really happened when there was no Treasury dislocation at all going on" — "swap spreads on the month were tightened", not widening, and there was no off-the-run versus on-the-run problem
"the Treasury seemingly decided that they didn't really like the number on the screen in terms of the level"
The first consequence he draws: "you really invite the market to fight as we kind of seen with dollar yen"
The second is the reaction across other markets the following day, which he thought was the interesting part — "dollar Swiss goes down one and a quarter% euro dollar is almost up 1% gold's up almost four. Those are big moves."
The question that follows is whether anything has actually been solved: "are you just kind of moving one market imbalance from one asset class to another"
He compares it to Japanese policy going back and forth between long-dated JGBs and the yen — "you kind of just doing a whack-a-mole because you don't want to deal with the underlying problem"
Joining the Dots: Yen Intervention, Swap Lines and the 30-Year
Hafeez lines up the pieces: joint US-Japan intervention in the yen, the buyback, and a widened FX swap facility
Turek says they connect, and dates the start earlier than the buyback: "this really started in January of this year when Bessent rate checked through the New York Fed dollar yen"
That episode made more local sense to him. It was more specifically about JGBs, a rising dollar-yen was pressuring US yields, and it bought time for Takaichi and her finance minister, Satsuki Katayama, to formulate a plan alongside the Bank of Japan after Takaichi won the LDP race
This time is trickier in his view. Yes, "we broke 163 dollar yen was at the highs" — but it did not feel like a JGB-driven global selloff
His conclusion: "the solve is 30-year Treasuries and every action is sort of like a derivative of that", with everything else a question of what can plug the gap in the moment
He thinks the linkage has expanded policy risk rather than removed it
How to Trade a Treasury That Will Fight You
Hafeez asks the position question directly — steepener, outright short, or is this an overread of something that goes sideways again
Turek's first rates takeaway is that the level is now a target. "you end up making the market a little bit more of a floor than a ceiling", the same way the Japanese did at 160 in dollar-yen, and it was done in benign macro conditions with financial conditions on balance easy
His second is to respect the firepower. The mechanism available in the US is "it's basically a form of Operation Twist right where it's you basically have to sell the front end to buy the back end"
"I don't think this changes the structural trend by any means. But yet you now would know that one of the main actors in the game is going to fight you."
The third takeaway is about what the intervention protects. The biggest economic risk in recent weeks had become a cost of capital that cuts off AI capex; now the most meaningful actors are trying to limit that damage
His two exhibits are "Nvidia announced that they have this $500 billion fund backed by every major Wall Street firm", which he reads as smoothing financing for the most capital-intensive part of the supply chain, followed by Bessent effectively intervening
If the spend is more durable, the downside case for the parts of fixed income that track growth shrinks — "the growthy parts of fixed income, which I would argue is sort of like the reds and SOFR and five-year notes and Treasuries"
Which leaves him with the counterintuitive conclusion: "you could actually price terminal rates higher in the US because the chances of something, you know, cutting off this positive growth impulse has actually gone down"
The likely path, he says, is that the market does this whack-a-mole between all of it
The Dollar's Downside Is Now the Bigger Tail
Hafeez asks whether a weak dollar is the cleaner expression, then answers himself — if the Fed ends up hiking because rates further out are up, it goes in circles. Turek's reply: "And you answered your own question."
He thinks the dollar was right to sell off on the day, and lays out the game theory he had been turning over on his walk in that morning
If Bessent upsizes the buybacks and sells bills to fund them, that pressure on funding rates should return, especially now the Fed is not doing many reserve management purchases — the operation it started last year to smooth bill issuance into funding rates that were dragging above where it wanted them
If the Fed then has to restart those purchases to defend its policy rate, Turek's verdict is unsparing: "I don't know if that's technically QE but that sounds like QE to me"
He can envision a world where core PCE, the last two CPI prints and the payroll print give the Fed rope not to hike, while reserve management purchases effectively fund Bessent's intervention — which he calls a terrible outcome for the dollar
The complication is that the modal path is genuinely unclear. A more durable cycle means "maybe the Fed could actually have like a mini hiking cycle here", and he thinks the chances of that went up
Where he lands: "the dollar now has its left tail is much bigger than its right tail"
Europe: An Asymmetric ECB and a Deficit That Grows on Its Own
Asked whether Europe has anything to say, Turek says there is a lot. The consensus view is meager growth, supply shocks, an ECB forecasting inflation too high, one more hike and then nothing — and yet "the market prices a full hike into the next year"
What he thinks the Europe narrative misses is that "the ECB has communicated a very asymmetric reaction function" — "slightly below two doesn't really mean much and slightly above two is still going to get dealt with", a stance he dates to the 2025 strategy review
His proof is behavioral: "it's why they didn't really cut when inflation was printing 1.8 annualized"
That asymmetry sets a floor under European yields, and the structural backdrop pushes the same way — "Germany now is a 4% budget deficit" that gets mechanically bigger as growth fails to arrive
On politics, "you have a French election coming up next year", and in German local polling he cites one survey where the AfD is leading: "in this one poll I saw they're polling at 42%. They're in the lead."
He flags a turn in Germany's approach to China over the summer after a hands-off stretch in which he says China completely ate their lunch on trade, crystallizing at a French-German summit at the end of June or beginning of July, and pointing toward "either German tariffs or broad European tariffs on China"
Why that matters for rates: "the big disinflationary impulse is coming from Chinese goods and that might be about to change"
Hafeez asks if the bias is therefore for short-dated European yields to rise more than expected. Turek says mostly, but corrects the maturity — "I'm more looking in the belly and long end in Europe"
The bigger question he wants asked is whether the ECB moves past neutral: "are you hiking to a new policy stance i.e. are moving from neutral to restrictive", which he frames as "Hiking to 275 or three" rather than the low-stakes hikes within neutral so far
His conclusion on the bund: "the direction of travel and things like Bunds is still higher in yield even from you know 325 where we are today"
Oil Is Stuck, and That Is Worse for Inflation Than a Spike
Hafeez frames the disconnect: "when you speak to oil specialists or commodity specialists they keep saying like this is real this is something really bad but then macro markets don't seem to care", and notes many people were burnt in the first phase of the war
Turek describes himself as a taker of information here rather than a forecaster, and puts the fundamental case against the tape: analysts tell him "this is the most blatant supply and demand imbalance of all time" and then "I turn on the screen, I'm like, well, oil's at 78"
His explanation is elasticity. "we saw how China was able to shut off imports", sellers still moved refined product in the heat of it, and "trade has been rerouted through the US", which he says became a big export beneficiary — so "the elasticity of these markets was just greater than people assume"
That is not an argument to fade it, he says: "I'm more convinced in the floor than a ceiling at the moment"
Two things support the floor. First, "the actual product side is quite tight", and a Macro Hive analyst has been highlighting that refined product is what actually feeds into inflation
Second is the standoff itself, which he thinks is durable. A low oil price costs Iran leverage and gives the US operational room; a high one risks both a bombing campaign and an energy crisis — "Iran doesn't really want to get bombed. And the US doesn't want an energy crisis."
Neither side looks close to a deal, on the strait or on the nuclear file, and both are buying time — Iran to hand its deterrent off from the strait to something more durable, the US because the strait as a leverage point carries negative carry while "the Emiratis the Qataris the Saudis are building pipelines around it"
He thinks that could show up soon: "we could be looking in like the middle of next year even that there's like there's pipelines online that totally circumvent the strait"
The irony he draws out is the inflation consequence of a war that neither ends nor escalates: "it actually more durably raises inflation risk because it's you're letting it linger longer" instead of a spike that cuts off growth and turns disinflationary
Central banks have been positively surprised that "there's been no second-round effects", with the Bank of England "the BOE sort of been like the most notable of celebrating this fact" — but Turek notes it has only been about six months, and demand is uneven
His bottom line is not a return to an obvious choke point with oil at 130. It is that "both parties are incentivized to act below 70 in their own ways", keeping the market above there and keeping energy a cyclical upward pressure on inflation rather than the downward one it was last year
Advice for Graduates: The Friction Has Never Been Lower
Hafeez says the young listeners he interacts with are mostly pessimistic about the job market, and that Macro Hive has been taking on a large number of interns to get people on the ladder
Turek is sympathetic on timing — "I don't envy it from a timing perspective" — but thinks the offsetting change is enormous for anyone genuinely passionate about something, because "the friction has never been lower" to build technical depth and then share it
His comparison is his own start ten years ago, trying to get a hedge fund job without a well-known school or good resources, when getting up to speed in macro was genuinely hard: "I go on the CME website, it'd be prices from like a day ago"
What he thinks would have changed it: "what I could have done with, you know, a ChatGPT or a Claude", combined with the low barrier to publishing
He offers his own hiring as evidence. "I'm taking on an intern in the fall" — he put the call out on Twitter and was, in his word, blown away by who applied
His actual advice is one line: "write and share it and keep sharing it", because "it really is to some degree an idea of meritocracy" and "everyone's desperate for information and everyone's desperate for a new way of thinking about things"
His own path proves it — "I shared a blog with someone who worked at Moore and he was like why don't you come here for a summer" — and he does not think that is a one-of-one story in the industry
"share your thoughts and you'll be surprised who's interested in them"
Asked where to find his work, he points to "I have a website JST Advisors" and to his X account, @JTurek18, for the research and advisory side
Turek's bottom line is that the two biggest actors in the economy — governments and the AI spenders — both believe they have no choice but to keep committing capital, and that every official attempt to hold yields down while that is true simply moves the imbalance into another market.
Products, Companies & Tools Mentioned
Federal Reserve (The core subject: Turek's argument that Warsh is an owl without a strong pull over the committee, with Waller and Powell the board's intellectual leadership and Logan, Hammack and Kashkari the regional voices to watch)
US Treasury and Scott Bessent (Raised buyback operations from two billion to four billion with no Treasury dislocation going on, which Turek reads as an unwillingness to accept the level rather than a market-functioning fix)
European Central Bank (Turek's template for Warsh, via Lagarde's owl line, Lane and Schnabel as leading indicators — and separately the source of the asymmetric, hawkish reaction function he thinks floors European yields)
Google (His example of a hyperscaler whose operating cash flow used to fund the capex, and whose 10-year or 30-year bond moving 50 basis points does not change the spending calculus)
Meta (The 2022 VR overinvestment comparison clients pitch him, which he rejects because Meta never saw VR as existential — and the only hyperscaler that did not guide 2027 capex in Q2)
Nvidia (The $500 billion fund backed by every major Wall Street firm, which Turek reads as smoothing financial conditions for the most capital-intensive part of the AI supply chain)
Bank of Japan, Takaichi and Katayama (The January dollar-yen rate check bought time for the new Japanese leadership to formulate a plan, in a case Turek thought made more local sense than today's)
Germany (A 4% budget deficit that grows mechanically as growth fails to appear, a trillion-dollar fiscal program in March of last year as Turek's benchmark for the hyperscaler spend, and a turn on China trade policy after the French-German summit)
AfD (Leading one German local poll Turek saw at 42%, part of the political backdrop he thinks is underweighted in Europe)
Bank of England (The central bank he singles out as the most notable in celebrating the absence of second-round effects from energy)
ChatGPT and Claude (What Turek says he could have used to get up to speed in macro ten years ago, when the alternative was day-old prices on the CME website)
JST Advisors (His own firm, and where he points listeners along with his X account)
Brevan Howard and Moore Capital (His previous roles, and the source of the story about a shared blog turning into a summer job)
Books & Resources Mentioned
Richard Koo's balance sheet economics (Turek's framework for arguing that the public and private sectors are now on the same side, leaving East Asian households as the marginal savers)
The July FOMC minutes (Released the night before this conversation; the "several participants favored" language is Turek's evidence that Warsh had committee support to hike and did not)
The ECB's 2025 strategy review (Where he dates the central bank's asymmetric treatment of inflation slightly below versus slightly above target)
JST Advisors' weekly research note (Described in Hafeez's introduction as global macro trade ideas for hedge fund clients)
Get the latest market chatter and takes as they happen:
X | Threads | Instagram | YouTube | TikTok | Facebook