Never Go Full Macro
Waller changed the tone
This is a side note for my thoughts, mostly because the last few weeks have forced me to rethink the way I was trading rates.
Since Waller's recent speeches, I think there has been a real shift in tone. He has usually been one of the more dovish people on the FOMC, probably somewhere near Bowman, though I do not know exactly how to rank them. He was appointed by Trump and has generally been more open to cuts.
The important point in his April 17 speech was not simply "inflation is a supply shock, so look through it." It was more subtle. A single temporary supply shock can be looked through. But if you keep getting one temporary shock after another, the sequence itself stops being temporary. That is the turning point. You can call each shock transitory, but after enough of them, inflation is just staying high.
His May 22 speech moved even further in that direction. He still does not sound eager to hike immediately, but he also no longer sounds comfortable with a one-way easing bias. In my view, that is probably the right way to think about the current macro environment.
The ZQ mistake
When I was long ZQ, I was betting on short-term rate cuts. ZQ is the CME 30-day Fed Funds futures contract, so it is basically a way to trade the expected path of the effective fed funds rate.
My reasoning was this: I thought 2026 would be the year when AI capex started turning into real revenue, which would also mean more visible labor disruption. AI would start replacing jobs more clearly. We had already seen weak labor data across developed countries, including the U.S. and Canada. I also had personal bias here, because as a recent graduate, the job market felt horrible. Not just for me, but for basically everyone I talked to, including people with strong degrees and double degrees.
The labor market felt like it had shifted into "no hire, no fire." I thought the next step would be a real crack, from no hire, no fire into actual layoffs.
That was one of the biggest mistakes I have made.
After reading the Dallas Fed research, I realized the labor market can stay stable even with very low or negative job growth, because labor supply is also shrinking. People are leaving the labor force through retirement, health reasons, deportation, and demographic change. Healthcare and services have also carried a lot of labor growth. I still did not exit. I kept hoping the weakness would show up more clearly. That turned into one of the worst losses I have taken.
So the labor market is more resilient than I expected. The Fed has two mandates: employment and inflation. Employment is not breaking clearly enough, while inflation is obviously the problem. The old argument was that this inflation was supply shock rather than demand shock. If that is true, raising rates may put more pressure on the supply side while not doing much to reduce demand for essential goods like energy.
When I looked through CPI components, I had the same reaction. Insurance is going up. Healthcare is going up. Services inflation exists, but not as much as feared. Manufacturing inflation is being driven partly by AI capex, but higher rates will not really stop data centers, memory, electronics, or the AI buildout. These companies are not going to stop building. Unfortunately, we are in a supercycle.
Why Warsh made sense to me
That is why the Warsh-style approach made sense to me: shrink the balance sheet to reduce liquidity and money-supply excess, while cutting rates to reduce pressure on the real economy. I was probably too ideological and too optimistic.
QE mainly benefits banks, asset owners, and companies with too much debt. Normal people do not benefit the same way. Gallup's latest stock-ownership number is closer to the low 60s than my old 55% estimate, but the exact number is not really the point. The point is that ownership is uneven. A large part of the population gets the inflation without getting much of the asset appreciation. That is basically how you get a K-shaped economy.
I overlooked the liquidity risk. I thought Warsh and Bessent, both connected to Druckenmiller's framework, plus Treasury support, could make the approach work. But I now think Powell also tried some version of this, and QE is just too convenient. It is basically a free lunch until the asset bubble fully pops. Because of how the U.S. financial system works, liquidity support becomes the default answer.
This caused memorable suffering. It hurt my mental health, my physical health, and I am exhausted.
The pricing also looked ridiculous to me. At one point, the market was pricing roughly two net hikes by 2028. I thought that could go either way. If the war lasted long enough, demand destruction would eventually matter. If it stopped quickly, then two years of QT should eventually achieve something and create room for cuts. But it was a bad call.
I held the spread almost exactly to the bottom and sold around negative 25 bps. Right after I sold, it moved back toward negative 10 bps, and now it is around negative 6 bps. That was brutal.
To be clear, I was trading a spread. I shorted this year's rate path, meaning I was betting on hikes this year, and went long later cuts two years out. I thought that made sense: if inflation stayed out of control, the Fed could remain hawkish near term, but eventually the economy would crack and cuts would come later. Instead, the market priced no change this year and hikes next year. I understand that logic, but I think it leaned too hard into a world where the war lasts forever and growth concerns never matter.
The long end
This changed part of my view. Originally, I was considering going long the front end and hedging with a short in long-duration bonds, because of the deficit problem, interest expense, and QT starving the long end of the curve. That should push long-end rates higher, especially when term premium still looked low. Instead, I did the ZQ spread. It did not work.
Now I have pivoted more toward the long-duration bond trade, though very late. As of May 22, the 30-year Treasury yield was around 5.07%, the 20-year was around 5.06%, the 10-year was around 4.56%, and the 2-year was around 4.13%. Effective fed funds was around 3.62%. So the 30-year was only about 145 bps above fed funds, and the 30s/2s spread was only around 94 bps.
Given the deficit, interest expense, military spending, election pressure, possible stimulus, inflation volatility, and political uncertainty, that does not look like enough compensation.
Japan is also a warning sign. Japan's 30-year yield is around 4% now. Japan used yield curve control in a deflationary world. It is much harder to do that when inflation is back. The same logic applies to the U.S. If the U.S. tried yield curve control now, the market could read it as QE, fiscal dominance, and inflationary monetization.
If we end up with 4% inflation and only 1% to 2% real GDP growth, then a 6% 30-year yield does not seem unreasonable. If inflation has a 5 handle, then 7% or 8% long yields are feasible in an extreme case.
I also think shorting long-duration bonds is a cleaner expression than my ZQ trade. If the war lasts, or even if there is a ceasefire, clearing mines, reopening routes, getting ships through, and getting insurance back in place will take time. Any ceasefire is likely fragile. I do not think the supply shock disappears overnight.
Recent inflation data also shows pass-through across different layers: producer prices, intermediate inputs, and final consumer prices. We are already seeing the first phase of input inflation around 4% to 5%, but it has not fully passed through yet, partly because inventories have cushioned the shock. Recent Fed research also suggests energy inflation is spilling over into other regions.
So my current framework is simple. If the Fed hikes, long-duration yields should go up. If the Fed cuts, inflation concerns and term premium may rise, so long-end yields could stay high or move higher. If the Fed stays on hold, that is possible, but I think it is the least likely path if inflation keeps moving the wrong way.
Oil underneath it
Oil is the trigger underneath all of this. The latest EIA data showed U.S. commercial crude inventories fell by about 7.9 million barrels for the week ending May 15, while refinery inputs were around 16.3 million barrels per day and utilization was 91.6%. That is still strong.
Refinery cracks are also high. A crack spread is basically the margin from turning crude into refined products like gasoline, diesel, or heating oil. If cracks are strong, refiners still have a reason to buy crude. Demand is still there.
There are risks. If the war ends quickly, oil could fall, yields could fall, and risk assets could rally. Before the war, the U.S. was probably already moving toward disinflation or even recession. Excluding healthcare and AI capex, underlying growth looked much weaker. The labor market is also still horrible, even if demographics make the headline data look stable. This is not a healthy economy.
But I am not convinced oil just goes back to normal and stays there. Ceasefires are fragile. We could easily get one, two, or three ceasefires that later break. If Israel and Iran remain in conflict, or if U.S. involvement creates more instability, this becomes very hard to trade. I also think some central banks may be forced to sell Treasury holdings to raise dollar liquidity during the energy crisis.
The instrument problem
There are two main ways to short long-duration bonds: buy puts on TLT, or use futures like ZB and UB. ZB is the standard 30-year Treasury bond future. UB is the Ultra Bond future, which gives more exposure to the very long end.
The problem is that the risk is much larger than the tick size makes it look. ZQ is about $41.67 per basis point per contract. For Treasury futures, tick value is not the whole story. DV01, the dollar value of a one-basis-point move in yield, depends on the cheapest-to-deliver bond and its conversion factor. CME examples put Ultra Bond DV01 around $277 per full-size contract, though it changes with the cheapest-to-deliver bond.
I used UB, unfortunately. I do not know why. That was bad position sizing. It forced me to monitor headlines every day, and I am completely exhausted. My portfolio is also too exposed to the war.
I also hold Brent calls because of SPR releases, inventory draws, refinery cracks, and the possibility that governments are treating the war as temporary when it may not be. We are not yet at the theoretical oil-price breaking point because governments are actively releasing supply. But if the war does not end as fast as the market expects, and inventories keep drawing, oil could spike toward $120 or even $150.
So both positions are very leveraged, and that is bad. But I already messed up the portfolio. I am not selling immediately because futures reopen Sunday night, and by Monday or Tuesday we should get a better sense of whether this temporary ceasefire or de-escalation actually holds. I personally doubt it, because the uranium-enrichment conflict is still unresolved. But I am not a war expert. Nobody really knows.
Even if a ceasefire happens, I think it can be breached later, and we could face another round of energy disruption. On top of that, midterms and the new Fed chair setup could be bad for bonds, equities, and broader risk assets.
At this point, I only have limited remaining capital. If it blows up, it blows up. If it does not, I will clean up my positions and seriously limit my risk. I still do not want to be long equities here. I am very bearish on equities and still think the setup looks similar to 1999, as I wrote before.
Maybe the better version is broad ETF diversification plus one short on the Ultra Bond, or just TLT puts. HYG puts could also make sense because high-yield spreads are still tight. If rates stay high and financial conditions tighten, option-adjusted spreads should widen.
Macro is fascinating because of the language and flexibility it gives you. You can express views across futures, bonds, commodities, credit, rates, and options in ways that equities do not really allow. But I learned the hard lesson: never go full macro.
So yes, this is just a personal reflection. I think we are at a turning point for central banks. The market is no longer cleanly biased toward cuts, and the next policy surprise could plausibly go either way. We will see how it plays out.
With this administration, it feels like every time you bet on the worst-case scenario, you have an unusually high batting average. It is genuinely hard to process how poor the execution has been since the start of this year.