Good morning, traders. Let’s get started.
Last week was another highly volatile and difficult trading week. We did not make many trades. Our main tasks were to manage existing positions, realize profits, and proactively reduce risk before the market structure deteriorated.
First, we exited the SOXL shares that had been assigned to us through short puts two weeks earlier. The average cost of the position was approximately $142, and we sold the entire position on Tuesday at an average price of around $160. Although we did not try to capture every subsequent move, we successfully converted an involuntarily assigned position into a profitable exit within a relatively short period. Overall, the execution was quite satisfactory.
On the individual-stock side, we went long UNH and generated a substantial profit.
At the same time, we established a short position in AAPL. Half of the position has already been closed at a profit, locking in gains. The remaining half is still being held and is currently showing a small unrealized loss. Whether we continue holding it will depend on the upcoming Big Tech earnings, the broader index structure, and AAPL’s relative performance versus the NDX. We will not change the trade plan simply because of one day’s price action.
More importantly, on Thursday we observed a rapid deterioration in the market’s internal structure and a noticeable increase in risk signals. We proactively liquidated the vast majority of our positions and successfully avoided Friday’s sharp selloff in technology and semiconductor stocks.
At present, more than 90% of the portfolio is in cash, with only a small position in MXL common shares remaining. The market currently lacks a clear trend, while sector rotation and single-stock volatility remain extremely intense. We are therefore in no hurry to increase exposure again.
The priority is to patiently wait for opportunities with clearly defined entry levels, stop-loss conditions, and catalysts, rather than forcing trades merely for the sake of staying active.
In our Discord, I post updated thoughts every day, while many experienced traders in the group continue to share their insights. If you have not yet joined the subscriber chat, please click the link below:
https://sidestack.io/franktrading
Without further ado, let’s move on to the next section.
The SPX fell approximately 0.61% this week. The NDX significantly underperformed the broader market, declining 1.62%. The DJI fell around 0.50%, while the RUT declined approximately 0.97%.
As in recent weeks, the headline index declines were relatively modest. Beneath the surface, however, market volatility was far more severe than the index performance suggested.
GOOG, TSLA, and several semiconductor stocks were subjected to concentrated selling following their earnings reports. The SOX fell approximately 4.5% on Friday alone.
Meanwhile, energy, utilities, industrials, real estate, and materials continued to rise. This indicates that the market is still not experiencing a systemic risk-off move in which all stocks decline together. Instead, it is undergoing a structural repricing centered on AI returns on capital, real interest rates, and crowded positioning.
The market’s focus has shifted toward one central question:
Can AI-related revenue cover capital expenditures, depreciation, interest expense, and the steadily rising cost of capital?
AI demand remains strong. Cloud revenue, data-center orders, servers, memory, and optical communications have not experienced any obvious collapse.
However, with real interest rates near multiyear highs, the market is no longer willing to reward capital expenditures based solely on revenue growth. Investors increasingly want companies to demonstrate free cash flow and satisfactory returns on invested capital.
The energy sector rose approximately 4.07% this week, ranking first among the 11 major sectors. Utilities rose 2.33%, ranking second. Industrials gained 1.60%, real estate rose 1.39%, materials advanced 1.11%, and healthcare increased 0.92%.
Technology edged up around 0.17%, financials declined approximately 0.11%, and consumer staples fell 1.39%.
The two worst-performing sectors were communication services and consumer discretionary, which declined approximately 5.90% and 6.38%, respectively.
Energy was the only sector with a clear trend advantage this week.
WTI and Brent rose approximately 8% and 10%, respectively. Conflict in the Middle East, reduced vessel traffic through the Strait of Hormuz, attacks on Red Sea shipping, and damage to Kazakhstan’s export infrastructure collectively increased the supply-risk premium.
The relative strength of energy stocks also reflects two additional factors.
First, energy companies generate strong current cash flow, and their valuation duration is significantly shorter than that of AI and software companies.
Second, rising energy prices can be transmitted relatively quickly into corporate revenue, making energy companies less sensitive to rising real interest rates.
Energy is therefore not merely passively following oil prices. In an environment of high inflation and high real rates, it has become a comparatively clean cash-flow asset.
The strength in utilities, industrials, and materials follows a similar logic to what we saw earlier this year. It shows that the AI trade is broadening away from highly valued software, GPU, and model companies toward the physical construction chain supporting AI capital expenditures.
AI data centers require more than just chips. They also require:
Electricity
Natural gas
Power grids
Transformers
Cooling equipment
Industrial automation
Copper and aluminum
Construction and engineering services
Data-center real estate
The market is gradually shifting from “buying the AI concept” to “buying what AI needs to build.”
The rise in real estate was primarily driven by data-center REITs, defensive rotation, and investor expectations that short-term rates may be approaching a cyclical peak.
However, as long as long-end real rates remain elevated, traditional commercial real estate will still struggle to achieve a broad valuation recovery.
Communication services and consumer discretionary ranked second-worst and worst, respectively. This did not reflect a simultaneous collapse in the fundamentals of the entire sectors. Instead, the declines were concentrated in heavyweight constituents.
Tesla fell approximately 18% for the week, dragging the consumer discretionary ETF down around 5.1%.
Alphabet declined approximately 7.8%, pulling the communication services ETF down about 3.8%.
The two sectors share several common vulnerabilities:
High valuations and long-duration cash flows
Sensitivity to real financing costs
Exposure to the wealth effect among high-income households
Pressure from rising oil prices and consumer financing costs
Heavy index concentration in a small number of large technology companies
In addition, if AI stocks continue to decline, the negative equity wealth effect could weigh on US consumption during the second half of the year.
The US economy grew at an approximately 2.25% trend pace in the first half, but consumer spending may slow during the second half.
For the week ending July 22, global equity funds received approximately $30.4 billion of net inflows, down from $56 billion the previous week.
Global bond funds received approximately $15.1 billion of net inflows, while money-market funds experienced approximately $33.9 billion of net outflows.
Over the past four weeks:
Global equity funds received approximately $128.7 billion
Bond funds received approximately $95.6 billion
Money-market funds experienced approximately $58.9 billion of outflows
Overall, this combination remains risk-on. Capital is leaving cash and moving into both stocks and bonds.
However, the internal structure of equity flows has changed materially.
Equity funds in developed markets generally experienced outflows, with the exception of Japan. The United States was the main source of weakness.
Emerging-market flows, meanwhile, were concentrated in mainland China, South Korea, and Taiwan.
At the sector level, technology funds continued to receive the largest inflows, while industrial funds recorded the largest weekly outflows.
A notable divergence has once again emerged between technology fund flows and technology stock prices.
Technology funds are still receiving inflows, but the Nasdaq, high-beta AI stocks, and semiconductor stocks continue to decline.
This suggests that at least three different types of capital are currently operating in the market.
First, passive ETFs and long-term allocation capital continue to buy technology.
Second, hedge funds, momentum strategies, and leveraged investors are rapidly reducing gross exposure.
Third, short sellers are increasingly targeting companies whose capital expenditures, free cash flow, and credit quality are most vulnerable to scrutiny.
Strong fund inflows therefore do not mean that technology stocks have already bottomed.
On the contrary, when substantial inflows fail to push prices higher, it indicates that marginal selling pressure remains significant.
Global markets continued to display a pattern of weak US growth stocks, strong European value stocks, and substantial dispersion within Asia.
Germany’s DAX rose approximately 1.36% for the week. The Euro Stoxx 50 gained around 1.14%, the UK’s FTSE 100 increased approximately 0.91%, and France’s CAC 40 advanced around 0.88%.
Europe outperformed the United States for four main reasons.
First, technology and large AI companies have lower index weights in Europe.
Second, energy, industrials, and financials have higher index weights.
Third, euro-area economic data improved in July.
Fourth, the European earnings season began on a relatively stable note.
The euro-area composite PMI rose to 51.9. Services increased to 51.6, while manufacturing output rose to 53.0, the highest level in more than four years.
Germany, France, and the peripheral economies all improved sequentially. The UK composite PMI also rose to 52.1.
However, Europe faces risks from both energy prices and interest rates.
The ECB left rates unchanged this week, but Lagarde acknowledged that some members supported a rate increase.
The market has materially increased the probability of a September hike.
If energy prices continue to rise, the economy remains resilient, and core inflation reaccelerates, Europe may tighten policy further.
Europe’s relative strength is therefore not being driven by low interest rates. Instead, its lower technology exposure, cheaper valuations, and improving manufacturing activity have temporarily made it more resilient to the real-rate shock.
Dispersion within Asia-Pacific was even more pronounced than in Europe or the United States.
Chinese markets were relatively strong, as capital began rotating away from the most crowded North Asian AI hardware trades and into lower-valued, more lightly positioned Chinese assets.
Taiwan declined on Friday but remained relatively resilient for the week.
Japan was pressured by falling chip stocks, rising oil prices, yen depreciation, and the risk of US tariffs.
South Korea was the most important market to analyze this week.
The KOSPI fell approximately 1.9%, the KOSDAQ declined around 5.5%, and MSCI Korea dropped approximately 2.3%.
At the same time, foreign investors purchased approximately KRW 2.33 trillion of KOSPI equities, with most of the buying directed toward technology. The Korean won also appreciated approximately 1.6% against the US dollar.
This indicates that the decline in South Korean equities was not simply the result of foreign capital leaving the country. It was more consistent with domestic deleveraging and high-beta position reduction.
South Korean margin balances have fallen from a peak of approximately $25 billion to $22 billion.
Assets under management in leveraged ETFs have dropped from approximately $53 billion to $26 billion, nearly halving.
Leveraged exposure still represents approximately 2.1% of South Korea’s free-float market capitalization.
However, margin calls as a percentage of receivables have fallen to 0.7%, suggesting that deleveraging is still underway but has not yet become a disorderly forced-liquidation event.
Foreign ownership of South Korean semiconductor companies remains approximately 1.8 standard deviations below its long-term average.
In other words, foreign investors have returned for two consecutive weeks, but positioning remains well below historically neutral levels.
South Korea currently presents two opposing sets of conditions.
Bullish conditions include:
AI capital expenditures remain strong
Exports and the current account remain robust
Foreign investors have started returning
Foreign positioning in semiconductors is extremely low
Margin financing and leveraged ETFs have already been substantially reduced
Valuations are approaching cyclical lows
Bearish conditions include:
Memory-sector earnings may be approaching a cyclical peak
The market has begun questioning returns on AI capital expenditures
Korean interest rates remain under upward pressure
Deleveraging has not yet fully ended
The KOSPI responded poorly to positive news such as Alphabet’s higher capital spending
South Korea is therefore gradually entering a medium-term value zone, but short-term price confirmation has not yet occurred.
WTI rose approximately 8.3% this week and closed around $89.31.
Brent gained nearly 10% and closed around $96.78.
Brent briefly traded above $102 on Thursday, but fell sharply over the weekend following a temporary ceasefire.
The rise in oil prices reflected four components:
Actual supply disruptions
Shipping and insurance costs
Geopolitical risk premiums
Short covering and systematic trend-following purchases
Daily vessel traffic through the Strait of Hormuz has fallen to extremely low levels, while alternative Red Sea routes have also faced attacks by the Houthis.
The oil rally was therefore not purely a headline-driven move. Genuine logistical constraints were present.
However, oil’s extreme sensitivity to negotiation headlines on Friday also indicates that a large portion of the price above approximately $100 reflected a substantial war-risk premium.
Gold rose approximately 0.8% this week, while silver gained around 4%.
The most important development was not the size of the gains themselves, but the fact that precious metals demonstrated clear resilience in an extremely unfavorable macro environment.
The dollar continued to strengthen this week. The 10-year real yield rose to approximately 2.43%, while the 30-year real yield approached 2.95%.
Based on historical pricing relationships, an increase of this magnitude in real rates should have placed substantial pressure on gold.
Yet gold did not continue falling. Instead, it held its range and posted a small gain, while silver performed even more strongly.
This suggests that although real rates are still rising, their marginal negative impact on precious-metal prices is fading rapidly.
What matters most for asset prices is often not the absolute direction of the underlying driver, but how the asset responds to that driver.
When negative forces continue to intensify but the price no longer makes new lows, it often means that selling pressure has been largely absorbed and that persistent structural demand exists beneath the surface.
Gold may currently be supported by a combination of:
Central-bank and official reserve allocation
Concerns about fiscal sustainability
Geopolitical risk
Global reserve diversification
The market may also be beginning to price in the possibility that real yields are approaching a cyclical peak.
Even if real yields have not yet begun to decline, a slowdown in their rate of increase could materially reduce valuation pressure on gold.
If the FOMC, core PCE, employment data, or falling oil prices cause the front-end policy path and the dollar to ease even modestly, gold could quickly break above its previous range.
Silver could deliver even greater upside because of its higher beta.
The risk-reward profile of precious metals is therefore beginning to tilt clearly in favor of the bulls.
Copper rose modestly this week and remained at elevated levels.
Industrial metals are currently being influenced by two opposing forces.
Supportive factors include:
AI data centers
Power-grid construction
Improving European manufacturing
Global infrastructure investment
Limited incremental copper-mine supply
Negative factors include:
A stronger dollar
Weak Chinese domestic demand
Rising global real interest rates
Higher financing costs
Copper is therefore no longer a pure China-property trade.
Even if Chinese retail sales, property, and import data remain weak, copper prices may continue to demonstrate resilience as long as AI data centers, power grids, capital goods, and electrification investment keep expanding.
It is also important to differentiate within industrial metals.
Compared with traditional real-estate-related demand, copper, grid equipment, transformers, optical-communications materials, and certain scarce upstream materials have more reliable demand visibility.
US Treasuries experienced another significant selloff this week.
From July 17 to July 23:
The 2-year yield rose from 4.18% to 4.37%, an increase of 19 basis points
The 5-year yield rose from 4.28% to 4.46%, an increase of 18 basis points
The 10-year yield rose from 4.55% to 4.71%, an increase of 16 basis points
The 30-year yield rose from 5.06% to 5.17%, an increase of 11 basis points
At the same time:
The 5-year real yield rose from 2.01% to 2.17%
The 10-year real yield rose from 2.31% to 2.43%
The 30-year real yield rose from 2.87% to 2.95%
The key fact behind this increase in yields is that inflation compensation did not rise materially.
Nominal yields can be decomposed as:
Nominal yield = real yield + inflation compensation
Real yields can then be further decomposed as:
Real yield = expected real policy-rate path + real term premium
The difference between nominal yields and corresponding TIPS yields for the 5-, 10-, and 30-year maturities remains broadly stable at approximately 2.2%–2.3%.
In other words, medium- and long-term breakeven inflation remains anchored near the Fed’s target.
The market is not pricing in a complete loss of control over long-term inflation expectations.
Almost all of this week’s increase in nominal yields therefore came through real yields.
However, the drivers of higher real yields differ substantially across the various parts of the curve.
The 2-year is the purest policy-path trade.
Based on the nominal yield less short-term inflation compensation, the 2-year real yield is now close to 2.1%.
The main driver has been the market’s decision to reprice the probability of rate increases later this year, including the possibility of a September hike, into the expected real short-rate path.
The month-over-month decline in June CPI did not increase short-term inflation compensation. Inflation compensation remained relatively stable.
The approximately 15–20 basis-point increase in the 2-year nominal yield this week therefore primarily reflected repricing of the real policy-rate path.
The Federal Reserve’s internal debate has clearly shifted from “there is no need to raise rates” toward “we must be prepared to raise rates if inflation fails to improve.”
Logan has explicitly supported moderately higher rates.
Kashkari included one rate hike in his June SEP forecast.
Cook, Jefferson, Waller, and Hammack have all articulated conditional thresholds for tightening.
At the same time, Jefferson, Williams, and Waller still believe that if inflation improves over the next several months, current policy can be maintained.
The 2-year is therefore straightforward: it is driven by the Fed’s reaction function and by how the market interprets Warsh.
It does not carry the same long-term fiscal and physical-supply burdens as the long end.
If the FOMC does not validate the market’s current hawkish pricing, the 2-year should be the first part of the curve to reverse.
The 5-year reflects both the expected policy path over the coming years and the influence of medium-term real rates and capital flows.
CFTC data show that short positions in 5-year Treasury futures are near a two-year extreme, with net shorts representing approximately one-fifth of total open interest.
The 5-year is one of the most commonly used instruments among macro funds, relative-value funds, and basis traders.
As a result, this part of the real-yield curve contains a substantial amount of flow-driven noise.
The current 5-year real yield of approximately 2.17% reflects a combination of:
A higher expected real policy path
Reassessment of the neutral rate
Directional hedge-fund shorts
Corporate-bond and MBS duration hedging
Futures shorts associated with Treasury basis trades
This helps explain why flow pressure can make the 5-year appear relatively cheap even when fundamentals have not changed by the same magnitude.
However, the same positioning structure means that if the FOMC or PCE causes rate-hike expectations to unwind, the 5-year could experience short covering that moves faster than the underlying fundamentals.
The 10-year real yield has risen to approximately 2.43%, driven by three forces.
The first is the Warsh premium.
Even if the market does not believe that the Fed will continue hiking indefinitely, a higher-for-longer reaction function still raises the expected average real short rate over the next ten years.
The second is a reassessment of r∗r^*r∗.
Hyperscalers are spending hundreds of billions of dollars annually on AI capital expenditures, competing for savings.
Fiscal deficits and Treasury issuance are competing for the same pool of savings.
The price of savings is the real interest rate.
When both the private and public sectors simultaneously increase their demand for capital, the market requires a higher neutral real rate.
AI capital spending is therefore no longer merely an earnings narrative for the equity market. It is also becoming a neutral-rate narrative for the bond market.
The role of war is also nuanced.
The oil shock has not materially entered long-term inflation expectations.
Instead, it has entered real yields through:
A higher expected policy path
Tighter financial conditions
The absence of safe-haven buying in long-duration Treasuries
A higher required real risk premium
This explains why long-term breakevens remain stable even as the 10-year nominal yield rises toward approximately 4.7%.
The 30-year real yield is close to 2.95%.
This portion of the curve has become almost entirely a real-term-premium trade.
Its primary drivers include:
Treasury supply
Budget deficits
The market’s ability to absorb long-duration debt
The price sensitivity of pension and insurance demand
The composition of foreign official reserves
Concerns about fiscal sustainability
The spread between the 30-year and 2-year yields has widened to approximately 80 basis points.
This steepening is not primarily an inflation-expectations steepener. It is a steepening in real yields and term premium.
This also explains why a dovish FOMC cannot automatically repair the entire curve.
The 2-year can change dramatically because of one FOMC meeting.
The drivers of the 10- and 30-year yields—such as the Treasury, AI capital spending, government debt supply, and global long-duration buyers—do not disappear because of a single meeting.
The front end may reverse quickly in response to Fed communication.
For the long end to decline sustainably, however, oil prices, fiscal supply, or growth expectations must also change.
Comments
Nothing yet. Say the first thing.
Sign in to join the conversation.