JPMorgan's Perspective on the Great Capital Migration: Tech Debt-Issuance Anxiety Is Overdone, September Stock-Buying "Sharp Brake" Brews a Good Entry Opportunity

**JPMorgan's Perspective on the Great Capital Migration: Tech Debt-Issuance Anxiety Is Overdone, September Stock-Buying "Sharp Brake" Brews a Good Entry Opportunity** JPMorgan's recently released *Flows & Liquidity* global capital movement research report shows that this Wall Street financial giant judges the bond supply shock brought about by the expansion of tech financing can still be actively absorbed, and that the September cooling in stock market buying is more likely a phased pause rather than an alarm signal that a trend of capital retreat has been established. JPMorgan's quantitative model still shows a very high probability that the S&P 500 will rise over the next six months. In addition, the institution's breakdown of fund flows and global investment dynamics shows that over the past 12 months, inflow intensity into US materials, industrials, and energy sector ETFs unexpectedly led the tech sector; the globally highly popular AI semiconductor bellwether indices in Asian marketsnamely South Korean and China Taiwan stock market ETFsrecently received relatively strong inflows, but remain significantly underweight relative to global equity index weights, while the US weight leans "neutral." Corporate net debt issuance remains positive, while net equity issuance remains negativea statistic that highlights that while the corporate sector is increasing debt financing, it is also reducing net equity supply through capital activities such as buybacks. It is understood that JPMorgan's model calculations show that net issuance of US and European tech corporate bonds in 2026 is expected to increase by about $260 billion compared with 2025, an increment of about 5.3% relative to its previous forecast of roughly $4.9 trillion in global net bond issuance; based on supply-demand model estimates, the upward impact on the Global Aggregate (Global Agg) yield is only about 1015 basis points, i.e., 0.100.15 percentage points. After adding policy rate expectation variables, the estimated impact falls further to 510 basis points. JPMorgan's estimate of the tech debt-issuance shock is far more optimistic than the market's previously expected negative estimate of about 5 percentage points. At the same time, equity fund inflows have approached zero after five consecutive months of strong growth, but JPMorgan's senior analyst team tends to believe this weakness may be as short-lived as in April 2024; some European government bond momentum signals have entered extreme territory, potentially triggering profit-taking; both gold and Bitcoin have received funding support, while Bitcoin's higher hedging positioning may provide greater subsequent room for covering. JPMorgan says the latest global fund flow data highlight that what truly deserves attention now is "changes among supply pressure, marginal fund flows, and existing positioning." **Breaking Down the "Global Tech Debt-Issuance Frenzy": How the $260 Billion Increment Transmits to the Global Bond Yield Curve** JPMorgan expects net issuance of US tech investment-grade and high-yield bonds to rise from about $140 billion in 2025 to about $335 billion in 2026, with the year-on-year increment rounded to about $200 billion in JPMorgan's research report; European tech bond issuance this year is annualizing at a pace above 90 billion, higher than 350 billion in 2025the bank assumes the vast majority of the new portion constitutes net supply, estimating an increment of about 55 billion, equivalent to $60 billion, yielding a combined US-Europe net issuance increment of about $260 billion. To measure interest rate/yield curve risk, JPMorgan's analyst team multiplies bond stock by "index duration corresponding 10-year Treasury duration" to convert it into 10-Year Equivalents: the share of US corporate bonds in the combined corporate bond and Treasury duration increment rises from less than 10% in 2025 to about 40% in 2026, with US Treasuries still accounting for about 60%; the corresponding share of euro-market corporate bonds falls from about 40% to 20%. Overall net supply of US high-grade corporate bonds across sectors is expected to rise from $615 billion to nearly $1 trillion, with long-dated bond issuance by hyperscale cloud providers being an important reason for the increase in corporate bond duration supply. On this basis, JPMorgan uses the historical relationship between annual "excess supply"the change in supply minus demandand changes in the Global Agg yield to estimate an impact of 1015 basis points; after adding the annual change in the 6-month overnight index swap (OIS) rate, the impact falls to 510 basis points. However, because policy rates also affect bond supply and demand, the two explanatory variables are not completely independent, and JPMorgan ultimately still uses 1015 basis points as its main conclusion. Therefore, JPMorgan's analysts unanimously emphasize that this research report supports "moderate, absorbable marginal pressure from new tech bond supply," not that "AI financing will not affect the US Treasury yield pricing curve," nor does it treat this figure as the upper limit for the rise in the US 10-year Treasury yield; its estimates mainly target bond issuance and do not combine the effects of all bank loans, private credit, or AI productivity expectations. **Equity Buying: September's "Brake-Tapping"Why It May Be One of the Year's Best Contrarian Signals** JPMorgan's interpretation of September equity fund flows is relatively positive, based on the fact that fund and retail trading heat has cooled rather than fundamentals or long-term allocation logic being proven invalid. The September global equity fund-related flows observed at the time of the reportincluding mutual funds, ETFs, and leveraged ETF capital inflows and rebalancing tradesfell to near zero after five consecutive months of strong growth. Since retail buying through equity funds strengthened markedly in Q4 2023, a similar situation had previously occurred only in April 2024, when policy rate expectations and bond yields also rose, but funds resumed strong inflows from May. This time, the net open interest buying indicator for small single-stock call options, as well as the performance of retail-favored stock baskets relative to the S&P 500, are also at low levels, so the bank tends to believe the fund weakness will not last too longa so-called fund flow "fake-out." JPMorgan's quantitative model also provides important support: a logistic regression model combining trading activity, valuation, positioning, fund flows, economic momentum, and price momentum still has its latest chart position above the 75% threshold for judging the probability of the S&P 500 rising over the next six months. However, this is a directional probability model, not an expected 75% gain. Within this model framework, "above 75%" can be understood as a relatively strong bullish signal; a more accurate statement is that "the model estimates a relatively high probability that the S&P 500 will rise over the next six months." A curve above this threshold means the model estimates the probability that the index level six months later will be higher than the forecast starting point exceeds 75%. In addition, JPMorgan-compiled data show that as of September 15, composite equity positioning remained at the 75th historical percentile, government bonds at the 55th percentile, and credit bonds at the 21st percentile; global non-bank investors' equity allocation was about 50%, cash about 28%, bonds about 18%19%, and commodities including private gold about 3%; therefore, a more complete signal is that "incremental chase-the-rally heat has retreated while existing equity allocation remains relatively highwhich also provides a crucial observational clue for a recovery in buying." **Regional Rotation, Corporate Buybacks, and Long-Term Institutional Absorption Coexist** JPMorgan's fund flow dynamics data show that capital is still being allocated across regions and credit ratings, and corporates and long-term institutions continue to participate in the market. In the four-week average weekly data through September 9 on the front page, global equity funds saw net inflows of $7.2 billion and bond funds net inflows of $8.9 billion; US equity funds saw net outflows of $2.6 billion, while non-US equity funds saw net inflows of $9.7 billion; US high-grade bond funds saw net inflows of $7.4 billion, high-yield bond funds saw net outflows of $600 million, and European money market funds saw net inflows of $8.5 billion. These categories overlap and reflect regional and asset-quality divergence; they cannot be added together item by item. ETF monitoring further shows that over the past 12 months, inflow intensity into US materials, industrials, and energy sector ETFs unexpectedly led tech; South Korean and China Taiwan stock market ETFs received relatively strong inflows, but remain underweight relative to global equity index weights by about 0.6 and 0.5 percentage points respectively, Japan by about 1.2 percentage points, Europe overweight by about 1.6 percentage points, and US positioning appears "neutral"enough to show that "fund inflows" and "relative underweight" can coexist. On the corporate side, in Q1 2026 the G4 economies combinedthe US, UK, euro area, and Japanare expected to see non-financial corporate overall cash flow as a share of GDP exceed capital expenditure, with net debt issuance remaining positive and net equity issuance negative; as of August, globally announced buybacks are estimated at about $1.3 trillion, of which the US accounts for about $1 trillionthis is announced amount, not executed amount. G4 pension funds and insurance companies are still net buyers of bonds, with US and UK pension samples in a funding surplus as of July; credit creation in the US, Japan, and the euro area remains positive; and as of September 4, global IPO, follow-on equity issuance, and M&A year-to-date announced volumes were up 167%, 45%, and 35% year-on-year respectively. JPMorgan says these data together show that corporate financing, buybacks, and long-term institutional bond purchases are still operating. **CTA-Dominated Trends and Hedging: Bond Market Momentum Turning Extreme, Equities Not Yet Fully in Panic Pricing** The most noteworthy change in signals related to quantitative strategy CTAsknown as "fast money"is that some bond downtrends are approaching the zone that would trigger mean reversion and profit-taking. The short- and long-term average momentum standard score for German 10-year Bund futures briefly fell below "-1.5," returning to about -1.4 at the time of the report; US and Japanese 10-year government bonds returned from about -1.4 to -1.2, the UK from about -1.2 to -0.8; the futures momentum signal for French 10-year government bonds relative to German Bunds reached -1.6 on September 15 and was still about -1.5 at the time of the report. JPMorgan's trend framework has a mean-reversion filter; when signals exceed the threshold of plus or minus 1.5, they may turn neutral. Therefore, these are technical clues for potential position reduction and profit-taking, not statistics on all actual CTA short positions. On the equity side, average momentum signals for the S&P 500, Euro Stoxx 50, Nikkei, and MSCI Emerging Markets remain between +0.4 and +0.8, not reaching the extreme levels seen earlier this year; in the rule model on page 17, the S&P 500 is long both short- and long-term, the Nasdaq 100 is short-term short and long-term long, WTI and Brent turned neutral due to overly strong momentum, and gold and silver also show short-term short and long-term long. Hedging and institutional indicators are also inconsistent: as of September 15, the S&P 500 three-month at-the-money implied volatility was 15%, close to the one-year low of 14%, while Brent reached 52%; SPY and QQQ short interest ratios remain below the highs of earlier years, tech sector short interest standard scores are near neutral, while consumer staples, utilities, and industrials stand out more relative to history. Active US Treasury funds' beta to the Global Aggregate is close to 1.0, US balanced funds' equity beta has declined, while risk-parity funds' equity beta remains above its long-term average; as of September 15, CTAs, the risk-parity fund sample, and a 60/40 US equity/US bond portfolio were up 9.4%, 7.3%, and 6.7% year-to-date respectively, while MSCI global equities rose 12.4% and the Global Aggregate fell 0.8% over the same period. These data basically indicate that high interest rate pressure is prompting strategy divergence rather than a synchronized withdrawal of all capital; similarly, active trading does not equal net inflowsin the report's trading monitoring, emerging and developed market equity year-to-date notional trading volumes rose 129% and 43% year-on-year respectively, which can occur alongside the slowdown in new fund buying in September. It is worth noting that the above compiled signal-nature data are not statistics on CTAs' actual positions, but rather "directional signals that CTA/momentum funds might adopt" simulated by JPMorgan using a trend-following rule model. The model records signals as +1=long, -1=short, 0=neutral, and calculates short-term and long-term lookback windows separately; therefore, as of the latest model state in this September 16, 2026 report, the S&P 500 is short-term long + long-term long, with lookback periods of 84 days and 315 days respectively; the Nasdaq 100 is short-term short + long-term long, with lookback periods of 84 days and 462 days respectively. More specifically, the S&P 500 short-term long signal has been maintained for about 34 days and the long-term long for about 66 days; the Nasdaq 100 short-term short signal just turned about 2 days ago, while the long-term long has been maintained for about 55 days. So it more accurately reflects that JPMorgan's model judges CTA-type trend funds are currently "short-term long + long-term long" on the S&P but "short-term short + long-term long" on the Nasdaq. **Gold and Bitcoin: Gold Sees Stronger Fund-Return Trend, but Bitcoin"Digital Gold"Has the Edge in Potential Position Covering** Since the end of July, gold ETFs have fully recovered their previous fund outflows, while ETFs linked to Bitcoindubbed "digital gold"have recovered only about half; however, IBIT's short interest ratio remains near its year-to-date high and significantly above GLD's, so if hedging demand declines in the future, Bitcoin may receive stronger marginal covering support. Both are attracting capital, but JPMorgan places more emphasis on Bitcoin's yet-to-be-fully-released position-repair space. Since the end of July, both gold and Bitcoin ETFs have seen significant inflows; gold ETFs have fully recovered their earlier year-to-date outflows, while Bitcoin ETFs have recovered only about half, with some slight giveback recently; the bank believes that under improved news conditions, Bitcoin ETF demand still has room for further normalization. Futures positioning proxy indicators for both are also at relatively high levels, reflecting that institutional capital has also participated in the "debasement trade," and the recent rise in real bond yields has caused this trade to pull back somewhat. The key difference is not that gold lacks funding support, but that Bitcoin still retains more cautious or protective positioning: IBIT's short interest ratio is near its year-to-date high, while GLD's is below its historical average; IBIT's put/call open interest ratio is also higher than GLD's. Therefore, if hedging demand declines, related position adjustments may provide stronger marginal support for Bitcoin than for gold. After the short-term cooling in flows, opportunities are more likely to come from a recovery in demand, adjustment of excessive trend trades, and covering of hedging positions.