1. The broader question remains where equilibrium rates ultimately settle. In June and again in July I argued that AI could matter as much for the neutral rate as it does for productivity, because the investment arrives first, competing for capital, power, infrastructure and labor, while the productivity benefits arrive over a much less certain timeframe. That still leaves me with the view that equilibrium rates sit structurally higher than markets became accustomed to after the GFC, and that this cycle is ultimately constrained through the cost of capital rather than a collapse in demand. The past two weeks bring that back into focus. Higher energy, persistent inflation, resilient activity and changing central bank reaction functions are forcing markets to revisit how restrictive policy actually is. An energy shock does not itself mean r* has moved higher, but the two can reinforce one another, and after the repricing we have just seen I would be careful assuming that inflation eventually normalizes and rates simply return to familiar levels.

2. Energy is now a question of how long the physical buffers can last. The precautionary shutdown of Saudi Arabia’s East-West pipeline and renewed threats to Red Sea shipping matter because they compromise routes that had helped offset disruption through Hormuz, and the IEA estimates global inventories have fallen by roughly 500m barrels since the war began, which is a meaningful amount of cushioning to have removed. I am uncomfortable with the assumption that political incentives produce a timely resolution. Military superiority does not guarantee normal commercial flows, and continued economic pressure can just as easily broaden the set of infrastructure at risk. Equally, any credible reopening would create a very sharp reversal in oil and rates, so both tails matter. The macro point is persistence. Each additional month of disruption works through household purchasing power, corporate margins, inflation expectations and policy, and by the time weaker demand balances the oil market, the adjustment may also be showing up in growth and earnings.

3. The other thing the past week reinforced is how difficult the path becomes once volatility itself is part of the problem. The proliferation of multi-strat platforms, leveraged RV businesses and systematic strategies means positions that look diversified in normal markets can become highly correlated under stress: some share the same macro assumption, others are connected through financing, volatility targets and similar liquidation thresholds. Once the move starts, losses reduce drawdown capacity, estimated volatility rises and positions have to be cut simply to keep risk constant, so a trade can become more attractive fundamentally at exactly the point it becomes harder to hold. Calling every episode a six or eight standard-deviation event offers limited comfort if the volatility estimate was calibrated to the wrong regime. That is why sizing and portfolio construction matter as much as the eventual destination. Dealer balance sheets are finite, electronic liquidity can disappear remarkably quickly and markets seem capable of moving from low volatility to high volatility with very little in between. If a book is already fully expressing the view when that happens, there is no capacity to lean into the dislocation and sometimes very little capacity even to hold the original risk. Portfolios that are implicitly short volatility through carry, leverage, crowding or liquidity assumptions need some form of offsetting convexity, whether explicit or simply through lower starting size and genuinely diversifying positions. The same applies across asset classes: a rates position, an energy-importer carry trade, a consumer equity exposure and a credit book can all look independent while sharing exactly the same dependence on stable inflation, affordable energy and accessible financing. The objective is not to avoid drawdowns. It is to preserve enough risk capacity to hold or add when the opportunity improves. Being right about the eventual clearing level is only valuable if the portfolio can survive the path.

4. On the Fed, I still lean towards a September hike, but my probability remains below the market’s. I moved towards a hike after Jackson Hole and still think 25bp on Wednesday is the most likely outcome, but I am closer to 70% than the roughly 90% now embedded in pricing. Kevin Warsh made a hold harder for himself at Jackson Hole by saying the better summer inflation prints had not convinced him that the underlying trend was improving sufficiently, and Friday’s 0.3% core CPI does little to resolve that concern. At the same time, the economic case is not quite as clean as the pricing suggests. The read-through to core PCE looks closer to 0.25-0.30%, payrolls have softened and the Committee remains divided. My suspicion is Warsh can assemble the votes for 25bp, particularly with energy moving higher, but I would not extrapolate that into the beginning of a sustained hiking cycle. One hike could just as easily buy him time.

5. What matters more is what either decision does to the curve. A credible hike could actually help contain long-end inflation and term premium by reinforcing the Fed’s willingness to act. The more uncomfortable outcome is a hold that rallies the front end but sells off 10s and 30s because the market starts questioning whether the Fed is willing to confront inflation. We have seen that movie already. If front-end yields fall while long-end yields rise, particularly alongside a weaker dollar and higher gold, the curve is steepening for the wrong reason. That is a much more important signal to me than whether the Fed moves 25bp on Wednesday. Equally, if they hike and the long end behaves well, that probably buys broader risk assets some time. The policy decision matters, but the market’s reaction to it will tell us more about where credibility and term premium actually sit.

6. Europe looks more vulnerable to the growth side of the adjustment. The ECB’s 25bp hike came with higher inflation projections and a somewhat better near-term growth assessment, so its reaction function needs to be taken seriously while energy remains elevated. But Europe faces a much larger imported-energy burden than the US and captures far less of the direct investment impulse from the AI build-out, while still sharing in the rise in global financing costs, which is a tough combination. My medium-term bias remains towards lower European belly forwards relative to the US, but timing matters because the ECB may need to bring tightening forward before the growth damage becomes visible.

7. The UK is where the pricing looks most stretched to me. My base case remains a hold on Thursday, and the market broadly agrees, with only around 6bp priced for the meeting and Andrew Bailey’s appearance on Tuesday the last real chance to shift that. It is everything beyond September that has moved. November is now more than 80% priced, the strip carries roughly 45bp of tightening by December, nearly two full hikes by year end, and cumulative pricing runs to around 110bp by next summer, with the December SONIA contract implying close to 4.3% and March 2027 above 4.6% against a base rate of 3.75%. The question is therefore no longer whether the Bank sounds hawkish this week. It is whether they can out-hawk what the curve has already done without actually hiking, which I doubt. I struggle to see the economy absorbing everything priced into the end of next year, so the receiving side is more attractive on a medium-term basis than at any point this summer, and the two-plus-hikes entry condition I set out in July has now been met on pricing. What I want before leaning in properly is Bailey’s tone on Tuesday and some evidence the repricing has stalled, because the whites and reds are thin enough that terminal pricing can travel a long way on very little flow in both directions, and this remains a market that can overshoot well before it corrects.

8. Japan is also becoming more interesting, although I would be careful getting too far ahead of the evidence on repatriation. The yen’s strength despite higher global yields and energy prices is worth watching, but the latest flow data still show Japanese trust accounts buying foreign bonds in August, so I would not overstate the extent to which flows have actually turned. The more important point is that incremental allocation can change well before the existing stock of foreign assets is sold. If domestic yields continue to become more attractive and BoJ normalization looks credible, Japanese investors can simply direct more new money at home and hedge more of what they own abroad. That is enough over time to matter for global duration. My bias remains that yen exposure becomes more useful as a medium-term portfolio hedge, though I prefer being selective on expression rather than simply shorting USD/JPY at any level.

9. Equities continue to look remarkably relaxed relative to the macro distribution. Friday was instructive because equities rallied and volatility fell even as Fed hike expectations firmed. Oil also came off, so it was not a clean experiment, but it reinforced the point that higher policy rates do not automatically mean lower equities. What matters is why yields are moving. Higher real yields alongside stronger earnings and investment can be absorbed. Higher yields driven by energy, inflation expectations or deteriorating policy credibility are much harder. I would also be careful assuming equity positioning is universally stretched after the de-risking we have already seen. The more durable issue is that volatility underneath the indices remains much higher than the headline market suggests. Into the central-bank meetings, expiry and quarter-end, I still think there is a case for owning convexity while it remains reasonably priced rather than waiting to buy it after volatility has already moved. A constructive medium-term earnings view can coexist quite comfortably with a much more cautious view on the path over the next few weeks.

10. AI remains the other side of this. The fundamental investment case continues to look very strong, but the debate is no longer simply whether capex is too high. Hyperscaler capex was roughly $150bn as recently as 2023 and consensus is now above $1.3tn for 2028. At that scale the market is obviously right to obsess about returns, but I am less convinced that ROIC can always be reduced to a simple calculation. Some AI spending will generate directly measurable revenue, some will allow businesses to operate more efficiently, and some may increasingly resemble cyber security, spending that produces a poor standalone return but becomes necessary simply to remain competitive. Competitive pressure alone could therefore sustain investment considerably longer than a conventional capex model would suggest. Oracle captures the tension on the other side: demand is enormous, with cloud infrastructure revenue more than doubling and contracted demand growing rapidly, but so is the capital required to satisfy it, with free cash flow remaining deeply negative while the infrastructure is built. The consistent message across Nvidia, Oracle and the rest of the ecosystem is that power, data-center capacity, memory and execution remain binding constraints. I remain constructive on the theme, particularly where genuine physical scarcity supports pricing and returns, but the thematic is changing. The early phase rewarded identifying AI and owning the beneficiaries. The next phase is about who captures the economics, who funds it and which spending ultimately proves productive. I would expect the path to become much more jagged as a result, with more DeepSeek-type air pockets forcing investors to reassess where scarcity and value actually sit in the stack.

11. Regulation is another risk around the AI distribution that I think markets need to spend more time on. The largest frontier labs are increasingly advocating capability-based safety standards, independent evaluation and greater oversight of advanced models. The safety concerns are real, but so are the economic consequences. Regulation creates fixed costs, and the largest labs and hyperscalers are far better positioned to absorb evaluation, security, reporting and liability requirements than smaller challengers. Depending on how the rules are written, that could entrench incumbents while making it harder for open-weight or smaller models to compete at the frontier. Poorly designed rules could equally slow deployment and reduce returns on capex already committed. For markets, the questions are who helps set the standards, where the thresholds sit and whether regulation changes the pace of investment or simply who captures the economics. I would not assume today’s AI winners remain tomorrow’s simply because the aggregate theme stays intact.

12. Stepping back, I still think the equilibrium cost of capital settles higher than markets became accustomed to after the GFC, but the scale of the recent repricing changes the opportunity set. I would not simply extrapolate this week’s moves. Some parts of the curve may already be pricing more tightening than the underlying economies can ultimately sustain, while other assets still look too comfortable with the idea that inflation and policy eventually revert to the old regime. AI fundamentals remain exceptionally strong, but the trade itself is becoming more complicated as financing, regulation and the question of who captures the returns become more important. That argues for more selective cross-market opportunities, better entry levels and a much greater focus on path and liquidity. The biggest mistake available in this environment is probably not being wrong about the destination. It is holding the right view in a form that has to be liquidated before it can pay.

The views expressed herein are those of Schonfeld Strategic Advisors LLC (“Schonfeld”) as of September 13^th, 2026 and are provided solely for informational and discussion purposes. These comments represent our current judgment based on information available at the time of writing and are considered “forward-looking statements” subject to change at any time without notice. Certain information contained herein has been obtained from publicly available sources believed to be reliable; however, no representation or warranty, express or implied, is made as to its accuracy, completeness, or timeliness. This commentary should not be construed as investment advice nor a recommendation to buy or sell any security, strategy or fund. Future results may differ materially from expectations due to evolving market conditions, new information or subsequent events.

This commentary is intended as a general macro or strategy-level discussion and does not replace fund-specific disclosure materials. The views expressed do not guarantee the future performance of any security, asset class or market and should not be relied upon as predictive of any Schonfeld account, portfolio or fund performance.

The content included herein is proprietary and may not be reproduced or distributed, in whole or in part, without the prior consent of Schonfeld.

__PRESENT

__PRESENT

__PRESENT__PRESENT__PRESENT__PRESENT__PRESENT