I increasingly view the current market environment as a contest between two fundamentally different forces: the deteriorating conditions in the bond market and the extraordinary growth expectations embedded in equities, particularly around artificial intelligence. The unusual feature is that neither side has yet decisively won. Bond investors are pricing a substantially more restrictive monetary environment, while equity investors continue to place considerable faith in economic growth, AI earnings, Treasury intervention, and the Federal Reserve's eventual willingness to support financial conditions.
The result is a market that has become extraordinarily difficult to interpret using conventional relationships. The most important signal is not necessarily the day-to-day movement of the S&P 500 or Nasdaq. It is the behavior of the Treasury market, because changes in long-term yields affect virtually every other asset valuation in the financial system.
The Bond Market Is the Foundation
The 10-year Treasury yield approaching 5% represents a major change in financial conditions. A move of this magnitude matters because bond prices and yields move inversely, and longer-duration securities are particularly sensitive to changes in interest rates.
A useful approximation is duration. If a 10-year bond has a duration of roughly eight years, a 100-basis-point increase in its yield can produce approximately an 8% decline in its price, all else equal. That loss becomes dramatically more consequential when the underlying position is leveraged.
This is where leverage transforms an ordinary bond-market decline into a potential balance-sheet problem. An institution that owns $100 million of bonds might use those securities as collateral to borrow additional money and purchase more bonds. The process can be repeated, creating a much larger gross position relative to the institution's actual equity capital.
Leverage therefore creates an asymmetric vulnerability. A relatively modest decline in the underlying asset can produce a much larger percentage loss on the investor's equity. The important lesson is not that every bond investor is literally facing an 80% loss, but that duration and leverage can interact in ways that make a seemingly manageable increase in yields extraordinarily destructive to highly leveraged portfolios.
This is one reason I pay much more attention to the bond market than to a relatively small decline in equities. Equity prices can remain surprisingly resilient even while the financial system underneath them is becoming more restrictive.
Why Are Long-Term Yields Rising?
I see several interacting forces.
The first is geopolitical risk and its effect on energy prices. Disruptions involving major oil-producing regions can raise the probability of a persistent energy shock. Oil is not merely another commodity. It is embedded throughout the global economy, affecting transportation, manufacturing, chemicals, logistics and consumer prices.
If crude oil rises substantially and remains elevated, the inflation consequences can extend well beyond the gasoline pump. Higher energy costs eventually propagate through production costs and supply chains.
That creates an uncomfortable problem for central banks. A geopolitical shock can simultaneously weaken growth and increase inflation. Monetary policy is poorly suited to solving that combination because raising interest rates cannot produce additional oil supply. Nevertheless, persistent inflation expectations can force central banks to maintain or increase restrictive policy.
The second factor is competition for capital in the bond market.
The U.S. Treasury is not issuing debt in isolation. Large technology companies and other highly rated corporations are also accessing the debt markets to finance enormous investment programs. The expansion of borrowing by AI infrastructure companies is particularly important because building data centers, computing capacity, energy infrastructure and associated networks requires extraordinary amounts of capital.
The result is a broader competition for fixed-income investors. Government bonds and corporate bonds compete for the same pool of capital, with relative yields helping determine where investors allocate that capital.
This does not mean that every dollar of corporate issuance mechanically causes Treasury yields to rise. The relationship is more complicated. But the broader structural point is important: the capital requirements of the AI investment cycle are becoming large enough to matter for financial markets themselves.
AI Has Become a Macroeconomic Variable
One of the most interesting developments is that AI is no longer simply a technology-sector story.
The AI boom has created an enormous investment cycle involving semiconductors, data centers, cloud computing, electricity generation, networking equipment, real estate, debt issuance and corporate capital expenditures.
That means AI increasingly operates as a macroeconomic variable.
On one side, AI creates enormous demand for capital. Companies must spend heavily to build the infrastructure necessary to support increasingly sophisticated models. This can increase corporate borrowing and therefore add to the supply of bonds competing with government securities.
On the other side, AI can produce extraordinary productivity and earnings growth. If the earnings and cash-flow potential of the leading technology companies increases rapidly enough, equity valuations can remain attractive even as bond yields rise.
This gives me a useful framework for thinking about equity valuations:
Equity valuation is ultimately a competition between expected growth and the required return demanded by investors.
Higher Treasury yields raise the risk-free rate and therefore increase the discount rate applied to future corporate cash flows. That normally puts downward pressure on equity valuations, particularly for companies whose expected cash flows are far in the future.
But exceptionally strong earnings growth can offset some or all of that valuation pressure.
This is why AI has become so important to the market. Investors are effectively asking whether the earnings growth generated by AI can outrun the increase in the discount rate caused by higher bond yields.
If it can, equities can continue rising despite a difficult bond market.
If it cannot, the bond market eventually wins.
The AI Investment Cycle Contains Its Own Paradox
The AI boom therefore contains an interesting feedback loop.
AI requires enormous investment.
That investment requires capital.
Capital can come from corporate cash flows, equity markets or debt markets.
Debt financing increases the supply of corporate bonds.
More corporate bond issuance can increase competition for fixed-income capital.
At the same time, the investment itself can generate stronger future earnings and productivity.
So AI can simultaneously be inflationary for capital demand today and potentially disinflationary through productivity tomorrow.
That distinction is crucial.
The market is effectively pricing the future productivity benefits of AI while simultaneously having to finance the infrastructure required to obtain those benefits.
This creates a difficult transition period in which the financial markets must absorb enormous investment spending before the full productivity gains become visible in economic data.
Inflation Has Become the Critical Variable
The inflation picture is particularly important because monetary policy is ultimately constrained by the inflation trajectory.
A single CPI reading is less important than the direction and persistence of inflation.
It is easy to dismiss an individual inflation component as a temporary anomaly. But if every elevated reading can be explained away as a "one-off," the cumulative result can still be persistent inflation.
That is why I care more about the underlying trajectory than about whether one particular category happened to distort one month's report.
The energy situation makes this even more important. If oil prices remain elevated for an extended period, inflation could become more persistent rather than simply experiencing a temporary headline spike.
That changes the expected path of monetary policy.
The market can tolerate a temporary inflation shock if investors believe central banks will look through it. The situation becomes much more dangerous when investors begin to believe that the shock will persist long enough to alter the central bank's reaction function.
The Central Bank's Reaction Function Matters More Than Its Words
One of the most useful concepts for interpreting monetary policy is the reaction function.
Central bankers often communicate conditionality: if inflation reaches a certain threshold, if employment weakens sufficiently, or if financial conditions change, policymakers respond accordingly.
Once the relevant data arrive, investors should focus on the reaction function rather than simply listening to individual sentences in speeches.
If policymakers have previously established a threshold for what would justify tightening and subsequent economic data cross that threshold, the probability of a policy response should rise materially.
This is more informative than attempting to interpret every adjective used in a speech.
The larger question is whether inflation caused by energy prices will prove temporary or persistent. If markets increasingly believe that energy inflation could remain elevated for an extended period, they will naturally price a longer hiking cycle.
That is precisely why the expected number of future rate hikes can change so rapidly. The market is not merely reacting to today's interest rate. It is continuously repricing the entire future path of monetary policy.
The Yield Curve Contains a More Nuanced Message
I also find it important to distinguish between the front end, the belly and the long end of the Treasury curve.
Shorter-term yields are closely connected to expectations for Federal Reserve policy. If the market expects more rate hikes, short-term yields should generally rise.
Long-term yields incorporate additional considerations, including long-run inflation expectations, economic growth, fiscal policy, term premia and the supply and demand for long-duration securities.
Therefore, a situation in which the front end sells off while the long end behaves differently can contain useful information.
A rising short end combined with relative stability or declining long-term yields can indicate that markets expect the central bank to tighten enough to control inflation rather than allowing inflation to become permanently unanchored.
Conversely, if both the short and long ends rise sharply, the market may be expressing a much more serious concern about the inflation and fiscal outlook.
This is why simply saying "rates are going up" is insufficient. The shape and composition of the yield move matter.
Treasury Buybacks Are a Powerful Concept—but Size and Execution Matter
Another important development is the use of Treasury buybacks as a tool for influencing the bond market.
The basic mechanism is straightforward.
If the government buys existing Treasury securities from the market, it removes some supply from private investors. Increased demand for a particular maturity can raise its price and therefore lower its yield.
But there is an important distinction between announcing a program and actually exerting enough force to move the market.
If investors expect a very large intervention and the eventual program is relatively modest, the announcement can actually become counterproductive. Expectations themselves become part of the market.
Suppose investors believe policymakers are preparing an enormous intervention. They may position themselves ahead of it. Bond prices can rise in anticipation. But if the actual intervention is much smaller than expected, those positions can unwind, pushing yields higher.
This is a classic lesson in financial markets:
Policy effectiveness depends not only on the absolute size of an intervention but also on its size relative to expectations.
A $6 billion intervention can be meaningful in isolation while simultaneously being disappointing if the market expected $10 billion, $20 billion or more.
Implementation Can Matter as Much as the Amount
There is also an important distinction between the quantity of securities purchased and the way they are purchased.
Official-sector purchases conducted through auctions and dealers can be structured around valuation and liquidity considerations. If the objective is simply to acquire securities at attractive prices, the official buyer behaves differently from an institution whose objective is to maximize the market impact of the intervention.
If policymakers genuinely want to push prices higher and yields lower, purchasing securities aggressively—even at prices above what private investors consider attractive—would have a much larger market impact.
This illustrates a broader principle of market intervention:
The mechanism of execution can be as important as the headline size of the program.
A theoretically large intervention can have limited impact if it is executed in a way that minimizes its price effect.
Conversely, a smaller intervention can have an outsized effect if it credibly changes expectations about future policy.
Fiscal Policy and Monetary Policy Are Becoming Increasingly Intertwined
The Treasury's behavior also highlights a broader structural issue: fiscal policy and monetary policy do not operate in completely separate worlds.
The Treasury controls the government's financing strategy and debt issuance.
The Federal Reserve controls monetary policy and short-term interest rates.
Financial markets, however, evaluate both simultaneously.
If fiscal deficits remain large, Treasury issuance remains substantial. If inflation is elevated, the Federal Reserve may maintain restrictive monetary policy. If long-term yields rise sharply, Treasury financing costs increase.
This can create a feedback loop:
larger deficits → greater debt issuance → greater interest expense → increased fiscal pressure → greater market sensitivity to yields.
That does not automatically imply a fiscal crisis. The United States possesses an enormous and highly liquid Treasury market, and Treasury securities remain central to the global financial system.
But the marginal sensitivity of markets to additional supply and inflation can increase when investors are already concerned about the trajectory of government debt.
The Bond Market Can Discipline Fiscal Policy
This creates what I think of as a market-based fiscal constraint.
Governments can authorize spending and issue debt, but they cannot completely dictate the price at which private investors are willing to absorb that debt.
The bond market therefore becomes a form of discipline.
When investors demand higher yields, the cost of financing fiscal policy increases. Higher yields then affect mortgages, corporate borrowing, equity valuations and the broader economy.
This is why the bond market can become more important than the stock market in determining financial conditions.
Stocks can sometimes ignore deteriorating fundamentals for extended periods.
Bond yields ultimately influence the discount rate applied across the entire financial system.
Why Aren't Equities Already Crashing?
This is the central puzzle.
If long-term yields are approaching 5%, oil prices are elevated and markets are pricing substantially more restrictive monetary policy, one might expect equities to fall much more aggressively.
Yet equities can remain surprisingly resilient.
I see three major explanations.
The first is confidence that policymakers will ultimately respond if financial conditions become sufficiently disorderly.
Investors may believe that the Treasury and Federal Reserve possess tools capable of stabilizing markets.
The second is AI-driven earnings growth.
Investors may believe that the leading technology companies can produce enough earnings growth to compensate for higher discount rates.
The third is confidence that monetary policy will eventually become more accommodative.
Markets frequently look beyond the current policy environment. Investors may tolerate restrictive conditions today if they believe the Federal Reserve will eventually reverse course once inflation is controlled.
This creates a peculiar equilibrium: investors simultaneously recognize that conditions are deteriorating while believing that policymakers and corporate earnings will eventually rescue valuations.
The Most Dangerous Assumption Is That Policy Will Always Rescue Markets
This is where I become cautious.
The belief that policymakers will intervene can itself become a source of financial stability—but it can also encourage excessive risk-taking.
If investors believe that Treasury or Federal Reserve intervention will prevent significant asset-price declines, they have an incentive to remain invested and potentially increase leverage.
That can suppress volatility temporarily.
But if policymakers ultimately lack the ability or willingness to deliver the intervention investors expect, the unwinding can be abrupt.
This is the essence of what is often called a policy put: the assumption that authorities will respond whenever asset prices fall far enough.
The danger is not necessarily that the policy put is nonexistent. The danger is that investors may price in more intervention than policymakers are actually prepared to provide.
Expectations Have Become a Market Force of Their Own
The Treasury buyback episode illustrates something broader about modern financial markets: expectations can become more powerful than the underlying policy action.
Markets do not simply react to what policymakers do.
They react to what policymakers are expected to do relative to what investors have already priced in.
That means the same policy action can produce opposite market reactions depending on expectations.
A large intervention can disappoint if something larger was expected.
A small intervention can surprise positively if almost nothing was expected.
The information content of policy therefore depends on the gap between expectation and realization.
This is one of the most important principles I use when interpreting market reactions.
Why Market Prediction Is Becoming Less Reliable
At some point, an environment can become so dominated by competing narratives that traditional forecasting loses its usefulness.
If the Federal Reserve might hike, hold, or eventually reverse course; if geopolitical developments can suddenly alter oil prices; if Treasury intervention can change bond-market dynamics; and if AI earnings can simultaneously overwhelm traditional valuation concerns, then there is no single clean macroeconomic path.
Multiple scenarios become internally coherent.
That is fundamentally different from simply saying that the market is "uncertain."
Uncertainty means I do not know which outcome will occur.
A highly reflexive market means several opposing outcomes can each be rationally justified using the information currently available.
That is a much more difficult environment for forecasting.
The Market Can Become a Battle Between Narratives
When different asset classes appear to be pricing different realities, I need to be especially careful.
Bond investors may be saying:
Inflation is too high, monetary policy will be restrictive, fiscal supply is enormous and long-term yields must rise.
Equity investors may effectively be saying:
AI earnings will be extraordinary, policymakers will stabilize financial conditions and the Federal Reserve will eventually become more accommodative.
Neither side necessarily has to be irrational.
They are simply assigning different probabilities to different future states of the world.
This is why market prices can look contradictory without actually being inconsistent.
The contradiction is not necessarily in the prices.
It is in the assumptions behind the prices.
Oil Is the Wild Card
Among all of these variables, I consider energy particularly important because it can affect both inflation and monetary policy simultaneously.
If geopolitical developments disrupt multiple transportation routes and reduce the effective availability of Middle Eastern oil, the resulting supply shock can propagate throughout the global economy.
Strategic petroleum reserves can temporarily cushion the shock, but inventories are finite.
Once inventories decline sufficiently, the market becomes increasingly dependent on actual production and transportation capacity.
At the same time, stronger demand from major economies can intensify the squeeze.
This creates a dangerous combination:
less supply + disrupted transportation + depleted inventories + persistent demand = greater price pressure.
And because refined products such as gasoline and diesel are themselves becoming more expensive, the inflationary effect can spread beyond crude oil.
Central Banks Face a Difficult Trade-Off
The resulting environment puts central banks in an uncomfortable position.
If they respond aggressively to energy-driven inflation, they risk weakening economic growth.
If they ignore persistent inflation, they risk allowing inflation expectations to become entrenched.
The policy challenge therefore becomes a question of credibility.
A central bank does not need to eliminate every temporary price shock. But it must convince markets that temporary shocks will not become permanent inflation.
This is why communication matters so much.
When investors lose confidence in the central bank's commitment to price stability, long-term yields can rise even before actual policy rates increase substantially.
The Long-Term Investment Lesson
For me, the broader investment lesson is that I should not treat equities, bonds, commodities and monetary policy as independent variables.
They are interconnected.
Oil affects inflation.
Inflation affects monetary policy.
Monetary policy affects Treasury yields.
Treasury yields affect corporate borrowing costs and equity discount rates.
AI investment affects corporate borrowing.
Corporate borrowing affects bond supply.
AI productivity affects corporate earnings.
Corporate earnings affect equity valuations.
Fiscal deficits affect Treasury issuance.
Treasury issuance affects the supply-demand balance of government bonds.
And all of these relationships feed back into one another.
The market is therefore not a collection of isolated asset classes. It is a network of financial relationships.
What I Would Watch
Rather than attempting to predict every market move, I would concentrate on a relatively small number of variables.
First, I would watch the 10-year and 30-year Treasury yields. The long end provides important information about inflation expectations, fiscal concerns, term premium and bond-market demand.
Second, I would watch the shape of the yield curve. The distinction between front-end and long-end movements can tell me whether the market is primarily repricing Federal Reserve policy or becoming more concerned about longer-term inflation and fiscal conditions.
Third, I would watch oil prices and energy-market disruptions. If the energy shock becomes persistent, it changes the entire inflation and monetary-policy equation.
Fourth, I would watch core inflation and its trajectory, rather than dismissing individual components as isolated anomalies.
Fifth, I would watch the Federal Reserve's reaction function rather than trying to interpret individual words in isolation.
Sixth, I would watch Treasury intervention, but I would pay particular attention to actual execution rather than announcements.
Finally, I would watch AI earnings and capital expenditure. The crucial question is whether AI's economic payoff is becoming large enough to justify the enormous capital requirements being imposed on the financial system.
The Ultimate Question: Growth or Discount Rate?
Ultimately, I believe the entire market can be reduced to a contest between two forces.
One is growth.
The other is the required rate of return.
If corporate earnings and productivity accelerate sufficiently, equities can withstand higher interest rates.
If interest rates rise faster than earnings expectations, equity valuations eventually come under pressure.
AI is currently the market's strongest argument for the growth side of that equation.
Inflation, oil, fiscal deficits and monetary tightening are the strongest arguments for the discount-rate side.
That is the battle I am watching.
Conclusion: I Would Rather Understand the Mechanism Than Predict the Outcome
I increasingly believe that trying to predict the exact direction of the market in an environment like this is less useful than understanding the mechanisms that could produce different outcomes.
If inflation persists, monetary policy can remain restrictive.
If oil remains elevated, inflation can become more persistent.
If Treasury issuance remains large, the bond market must absorb substantial supply.
If corporate AI investment accelerates, the demand for capital can increase further.
If AI productivity and earnings explode, equities may justify valuations that would otherwise appear excessive.
If Treasury intervention becomes sufficiently aggressive, yields can potentially be restrained.
If intervention disappoints, yields can rise further.
And if the Federal Reserve loses credibility or appears politically constrained, the bond market can react violently.
None of these possibilities exists independently.
The most important lesson I take from this environment is therefore not that I should become more confident in predicting the next market move. It is almost the opposite.
I should become more rigorous about identifying which variables matter, how they interact, what the market has already priced in, and what would cause those assumptions to change.
Markets become most dangerous when the apparent certainty of a narrative exceeds the certainty of the underlying evidence.
When the bond market, monetary policy, fiscal policy, energy markets and AI investment cycle are all moving simultaneously, I do not need to know exactly what happens next.
I need to understand the transmission mechanisms.
Because ultimately, prices are not the story.
Expectations, capital flows, incentives, leverage, policy and economic fundamentals are the forces underneath the prices.
And when those forces begin pulling in opposite directions, the most valuable investment skill is not prediction.
It is knowing what to watch when the prediction stops working.