After a volatile period in July for AI-related stocks, global markets have been steadily recovering since August. The South Korean Kospi index and the Philadelphia Semiconductor Index, key indicators of the AI sector, have rebounded 33% and 19% from their recent lows, respectively. The S&P 500 has even surpassed 7,800 points to set a new all-time high, while the Nasdaq is just 1.4% away from its own record. This recovery aligns with our mid-year outlook where we raised our S&P 500 target to 7,800-8,000 points.
Our analysis suggests that the recent correction, as well as the seven prior pullbacks during the AI rally, were all caused by a combination of three factors: high valuations and market crowding, external macroeconomic headwinds, and industry development bottlenecks. Now, as the market rebounds, we see that high leverage and crowded positioning have largely stabilized. For instance, the combined margin debt and leveraged ETF size in South Korea has shrunk by nearly 60% from its peak before stabilizing. Other global markets have also seen various degrees of leverage reduction. Our proprietary AI Bubble Pressure Index has also improved. However, the market is still waiting for a technological breakthrough to break the current ceiling. In this context, liquidity has become a key focus, particularly concerns over a potential reversal of the yen carry trade and increased US Treasury issuance.
From a liquidity perspective, despite recent positive data on non-farm payrolls, CPI, and PPI easing pressure for rate hikes, investors remain worried about potential unknown risks for the US dollar. Some are concerned that events like the US-Japan joint currency intervention, the Bank of Japan's rate hike, and increased US Treasury issuance could trigger a global storm similar to the yen carry trade reversal seen in August 2024. While this concern lacks concrete evidence, the suspicion that "if there were no pressure, why would the US help Japan intervene in the exchange rate" persists among some investors. This article will analyze the three key factors influencing US dollar liquidity in the coming months: Federal Reserve policy (cost), US Treasury issuance (supply), and the Bank of Japan and yen carry trade (major overseas demand).
The Federal Reserve's Policy: Less Pressure for Hikes, Inflation is Not the Main Issue; a Hike Could Be Beneficial if Data Deteriorates Again
Since Bessent took office in late May, market expectations for Fed rate hikes have fluctuated, especially after the ambiguous stance at the July FOMC meeting. This confusion pushed long-term bond yields above 4.7%. Recently, however, the market has received good news. Following the weak July non-farm payrolls report, both CPI and PPI have fallen, pushing the CME利率期货 implied rate hike expectations to December. Our mid-year outlook highlighted that the US economy is experiencing a K-shaped recovery, where traditional sectors like real estate remain suppressed by high interest rates. Therefore, the Fed does not need to be aggressive with rate hikes. Inflation is not the primary concern; the fear of hikes is mainly driven by AI-led growth and Bessent's communication style.
Inflation: It is likely to continue falling and is not a key issue. The July CPI fell from 3.5% to 3.4% year-on-year. Although core goods and services components saw a slight rebound, we believe the decline is highly probable due to the high base from last year's tariffs and the increased likelihood of Trump adopting a TACO approach on the Iran situation ahead of the mid-term elections. The question is only about the pace of the decline. Our calculations suggest that if oil prices gradually fall to $70 by year-end, the CPI could drop to around 2.7%.
Growth: While the overall picture is K-shaped, traditional US demand (excluding real estate) is not weak, and there are signs of it spreading due to AI. AI-related growth remains strong. The Q2 real GDP annualized quarter-over-quarter growth was 1.5%. While the contribution from AI-related investment fell from 1.11 percentage points in Q1 to 0.23 percentage points, it remained positive. The decline in inventories and increase in exports also reflect the pull from AI demand. The ISM制造业PMI rose from 47.9 at the end of last year to 55.6 in July. Private consumption, while not booming, has been stable, contributing 2.12 percentage points to Q2 GDP. However, July retail sales unexpectedly fell 0.6% month-over-month. Although July's non-farm payrolls were negative and missed expectations, seasonal factors and potential data adjustments mean it's too early to conclude that the labor market is cooling sharply. The rise in the US real interest rate from 1.9% in early May to 2.4% currently, which accounts for most of the increase in nominal rates, also reflects this economic strength.
Bessent's Style: This is a key reason for the recent swings in rate hike expectations. His ambiguous and wavering stance on the issue, shifting from focusing on trimmed mean inflation to AI inflation, and then recently suggesting that rising market rates are having a similar effect to a rate hike, has confused the market. The lack of clear signals on rate hikes at the June and July FOMC meetings, particularly the "hide-and-seek" communication in July, has fueled concerns that the Fed cannot effectively control long-term inflation expectations, pushing long-term bond yields higher. At this point, we see two possible paths for the market. If growth and inflation continue to weaken, the Fed will naturally have no need to raise rates, which would be a positive outcome. Recent economic data is trending in this direction. However, if August non-farm payrolls and inflation data worsen again, reigniting rate hike expectations, a "preventive rate hike" in September could be a beneficial move. It would help stabilize long-term inflation expectations and bolster the Fed's credibility. Furthermore, the actual implementation of a hike could mean the "bad news is out of the way," similar to the 1997 "preventive rate hike," after which bond yields peaked and the stock market quickly rebounded. Conversely, if market expectations for a hike rise but the Fed remains inactive in September, which would be politically difficult in October near the mid-term elections, it could amplify market anxiety.
US Treasury Issuance: Supply Increases in Q3, Size Nearing H1's Total, but July Issued Half, Mostly in Short-Term Bills
On the supply side, net US Treasury issuance in Q3 is projected to rise to $739 billion, a factor pushing up bond yields and affecting liquidity. According to the August borrowing plan from the Treasury Borrowing Advisory Committee (TBAC), Q3 net issuance will be $739 billion, and Q4 will be $628 billion. This is a significant increase compared to the $577.3 billion and $189.8 billion net issued in Q1 and Q2 of 2026. Since July, supply pressure and fears of long-term inflation失控 have pushed the 10-year term premium from 50bp to 80bp. This has caused the 10-year yield to rise from 4.47% to 4.7%, even as interest rate expectations have fallen.
However, a closer look reveals the impact may not be as severe as feared. First, the lion's share of Q3 supply pressure was concentrated in July. Of the planned $739 billion in net issuance, $355 billion was already issued in July, meaning the remaining $384 billion needs to be issued in August and September. This is a smaller disturbance than July's. Second, the issuance structure is heavily tilted towards short-term bills, limiting the supply pressure on long-term bonds. Short-term bills have consistently accounted for over 80% of total Treasury issuance in the past three years. In July 2026, they made up 86% of total issuance and 84% of net issuance, having a smaller impact on long-term yields. Looking back at the two issuance peaks in Q3 2023 and Q3 2025, supply pressure primarily affects bond yields by raising the term premium. Without the help of rate cut expectations, the impact can be more significant, as seen in 2023. In Q3 2023, the surge in issuance following the "debt ceiling" resolution pushed 10-year term premiums up nearly 120bp, driving long-term yields from 3.8% at the end of June to a peak of 5% in October. The interest rate expectations remained broadly flat after the last rate hike in July. In contrast, Q3 2025 saw a larger issuance of long-term bonds, but rate cuts drove yields down. That quarter also faced a peak in issuance after the debt ceiling was resolved, with net issuance of $1.058 trillion. However, two consecutive months of weak non-farm payrolls and a slower inflation increase spurred rate cut expectations, and the Fed resumed cutting rates in September 2025. During this period, the 10-year term premium oscillated in the 60-80bp range, while rate cut expectations pushed long-term yields from 4.5% in mid-July down to 4%. Overall, the Q3 2026 issuance size is smaller than those two peaks. Without the benefit of rate cut expectations, the market impact will likely fall somewhere between the Q3 2023 and Q3 2025 episodes.
The Bank of Japan and the Yen Carry Trade: A Repeat of the 2024 Unwind is Unlikely; Focus on the Yen Exchange Rate
Given the easing of Fed rate hike expectations, but still high costs, and minor supply-side disturbances, the stability of demand is crucial for the Treasury market to avoid increased volatility. US Treasury TIC data shows that as of May, foreign holdings of US Treasuries increased from $9.27 trillion at the end of last year to $9.37 trillion, representing 24% of outstanding Treasuries. Japan remains the largest foreign holder with $1.14 trillion, accounting for 12% of foreign holdings. As Japan's official sector is the largest overseas holder and the yen carry trade is one of the largest macro trades for buying Treasuries, it naturally becomes a focal point. Japan has a need to intervene in the currency market, but doing so requires selling US Treasuries, which is why the US stepped in to "help" prevent a sell-off. When the yen fell to nearly 164 at the end of July, the Japanese Ministry of Finance intervened in the currency market on the night of July 30, causing the yen to appreciate by up to 3.4% in one hour. The yen then experienced more volatility, appreciating by up to 5.5%. According to CICC International's estimates, the scale of this intervention exceeded 12 trillion yen (approximately $750 billion), making it the largest single round of intervention since the 1990s. On August 2, US Treasury Secretary Bessent confirmed a joint US-Japan currency intervention. Japan's currency intervention primarily involves selling dollar-denominated assets (like dollar deposits and Treasuries) from its FX reserves in the spot market to buy yen. To prevent a massive sell-off of Treasuries, the US Treasury cooperated through verbal intervention, by selling euros and buying yen, though the scale was small. The FIMA repo facility has not yet been used. If the intervention proves ineffective, it could force the BOJ to raise rates. This would narrow the US-Japan interest rate differential, potentially increasing pressure for a carry trade unwind, which could be further exacerbated by yen appreciation and risk aversion. Following the intervention, expectations for a BOJ rate hike have surged. Historical experience shows that currency intervention can only stabilize the exchange rate in the short term, not reverse a long-term depreciation trend. To stabilize expectations long-term, the BOJ may need to raise rates. The market's probability of a September BOJ rate hike jumped from 21% on July 30 to 81% currently. In this scenario, the yen exchange rate becomes a key variable. If the intervention leads to the expected yen appreciation, pressure on the BOJ to hike would ease. However, if the BOJ hikes alongside the intervention, it could trigger a yen carry trade unwind, leading to private sector selling of Treasuries and creating greater demand-side pressure on the bond market. The yen recently depreciated again from 155.2 on August 3 to 159.32, making its future trajectory a key point to watch.
So, will the August 2024 yen carry trade unwind storm repeat? We believe it is less likely under static conditions, but the risk of multiple factors coinciding unexpectedly should be monitored. The 2024 unwind was triggered by a simultaneous satisfaction of several conditions, which are currently not fully met. Yen short positions are significantly lower than they were. They were extreme at the end of July but have been drastically reduced since the intervention. Current net short positions are around 40,000 contracts, down from a peak of 170,000. BOJ rate hike expectations are already fully priced in. The market is pricing in over an 80% probability of a hike in September, making it a well-anticipated event. The US-Japan interest rate differential has not narrowed. A key driver of the 2024 unwind was the narrowing of the US-Japan rate differential, triggered by a rapid decline in US bond yields following weak US data. This has not occurred yet. The biggest difference between now and 2024 is the absence of a US recession narrative. The US economy still appears strong. Therefore, under static conditions, a repeat of the 2024 shock is difficult. However, we caution that the amplification effect of various negative factors coinciding could still occur, as liquidity shocks are often unexpected and have a "black swan" nature. The yen exchange rate is particularly crucial here. It is important to emphasize that even if a liquidity shock occurs, it does not change the underlying trend and often presents a better buying opportunity. Historical liquidity crises show that while short-term shocks are severe, central bank intervention to provide liquidity is effective, creating better entry points, as seen with the UK pension crisis, Silicon Valley Bank, and Credit Suisse.
Market Implications: Liquidity is Unlikely to Tighten Sharply; September is a Key Milestone; Any Volatility Could Offer a Better Entry Point
Current US dollar liquidity is not yet showing signs of significant tightening. In terms of price, the SOFR-OIS spread has narrowed after a brief spike, indicating manageable repo market pressure. Corporate bond spreads and credit spreads have also narrowed, suggesting easing corporate financing pressure. In terms of quantity, the Fed has stopped expanding its balance sheet. However, the use of the FIMA repo facility could create a similar effect. The reserves-to-bank-assets ratio is around 11.7%, which is not ample but not tight. Our financial liquidity indicator suggests that if the Fed does not start quantitative tightening this year, US financial liquidity should remain stable and not contract sharply. Overall, US dollar liquidity is expected to remain broadly stable, without a sharp contraction. The key focus is the yen exchange rate, with September being a critical period. For the market, greater upside potential lies in an improvement in earnings (molecular side). If investors are concerned about liquidity disruptions (denominator side), they could reduce exposure in the short term. However, any liquidity-driven pullback could offer a good entry opportunity.
For specific assets, we maintain a positive outlook on US stocks. Our mid-year target of 7,800-8,000 for the S&P 500 is being realized. Future upside will come from catalysts in the industrial trend, such as AI breakthroughs in Q3. If investors are worried about liquidity volatility, they can reduce positions or hedge tech exposure with consumer staples or gold. But any significant market decline due to liquidity issues should be seen as a buying opportunity. For US Treasuries, if there is no rate hike this year, the 10-year yield should trade in a 4.1-4.5% range. Supply pressure is mostly on short-term bills, with limited impact on long-term bonds. The US dollar is likely to continue its range-bound trend and will not weaken significantly. Gold has a well-defined floor, currently around $4,500 based on the current dollar and real interest rates. The upside potential depends on the narrative, such as de-dollarization. If the downside is limited and the upside requires waiting, gold can be held as a portfolio hedge. Key events to watch include the US August non-farm payrolls and CPI data, the September FOMC meeting (September 16), the BOJ's September rate decision (September 18), and any further US-Japan currency intervention.
Comments