On Wednesday 16 September 2026 the Federal Reserve raised interest rates for the first time in 38 months. It was not a close vote: 12 to 0. And 16 of the 18 officials who submit projections pencilled in at least one more hike before the year is out.
The market fell. The Dow dropped about 620 points.
By Thursday lunchtime the S&P 500 was up 0.93%. By the close it was up about 1.2%.
That is the whole story. But it is worth understanding why, because the 18 hours in between are the most informative thing that has happened to anybody's trading account this year.
- Dow Jones
- −1.2% · 51,461
- S&P 500
- −0.5% · 7,551
- Nasdaq Composite
- flat · 25,978
- Russell 2000
- −0.7%
- S&P 500 · lunchtime
- +0.93%
- Russell 2000 · lunchtime
- −0.40%
- S&P 500 · close
- ≈ +1.2%
- Russell 2000 · close
- ≈ +0.65%
The index first. The rate-sensitive stuff second.
What actually happened at the September 2026 Fed meeting
The FOMC moved 25 basis points to a target range of 3.75% to 4.00%. The last increase before this one was in July 2023. Three years and two months.
Chair Kevin Warsh was not ambiguous about it. The committee must be confident that underlying inflation is moving to target "clearly and at sufficient speed", and that standard, in his words, "has not been satisfied". He named three things: a labour market that is not softening, price pressures still running above 2%, and the Middle East.
Then the projections. Sixteen of the eighteen participants who submit a dot (Warsh himself does not) see at least one more hike this year: twelve have the year-end rate at 4.125% and four at 4.375%, against a current midpoint of 3.875%. The median path now sits at 4.1% for the end of 2026 and 4.1% for 2027, up from 3.8% and 3.6% in June. And the committee does not expect PCE inflation back at 2% until 2029.
Read that last one again slowly. The people who set the rate are telling you the job runs for another three years.
| Fed median projection | 2026 | 2027 | 2028 | 2029 |
|---|---|---|---|---|
| PCE inflation | 3.7% | 2.3% | 2.1% | 2.0% |
| Core PCE inflation | 3.4% | 2.5% | 2.2% | 2.0% |
| Federal funds rate (September) | 4.1% | 4.1% | 3.9% | 3.6% |
| Federal funds rate (June projection) | 3.8% | 3.6% | 3.4% | 3.1% |
The Fed is not on its own, either. The European Central Bank has hiked twice this year. The Bank of England held at 3.75% on Thursday with three of nine members voting to raise, and said a hike was becoming "increasingly likely". The Bank of Japan is expected to move on Friday. The 10-year Treasury yield is sitting around 5%, a level it last saw in 2007. Whatever this is, it is not a one-meeting event.
Eighteen hours
On Wednesday the tape did the correct thing. Dow down 1.2% to 51,461. S&P 500 down 0.5% to 7,551. Nasdaq flat at 25,978.
On Thursday it bought it back. Stocks opened higher, with the S&P up more than 1% in the first hour, helped by oil easing and Treasury yields snapping an eight-day rising streak. By early afternoon the S&P was up 0.93% and the Dow 0.88%. By the close the S&P had finished up about 1.2%.
There is a detail in that rebound that almost nobody will mention, and it is the part that matters. At lunchtime, with the S&P up 0.93%, the Russell 2000 was still down 0.40%.
Small caps are the most rate-sensitive thing in the equity market. They carry floating-rate debt, they refinance sooner, and they die faster when money costs more. So on the day after a unanimous hike with more signalled, the money went into the index and came out of the thing the hike actually damages.
Then, in the afternoon, small caps got bought too. The Russell finished up around 0.65%. Take either half of that session and the conclusion is the same. A view about a hiking cycle would have sold small caps and kept them sold. What the tape did instead was buy the index first and the rate-sensitive stuff second, which is not a view. It is a reflex, and it is a reflex with a very specific training history.
It is also happening on thin foundations. BTIG's Jonathan Krinsky pointed out on Wednesday that only 49% of S&P 500 stocks closed above their 200-day moving average, the first time since April 2000 that breadth has been that weak while the index sat within 4% of an all-time high. "You just aren't getting rewarded for buying strength," he wrote. The headline index is being held up by fewer and fewer names, and the dip-buyers are buying the headline.
A reflex is not an edge
For 38 months the direction of policy was one way. Every dip, without exception worth arguing about, was a buy. Anyone who bought weakness got paid, repeatedly, on a schedule, for three years.
Here is the uncomfortable part.
Nothing about that experience tells you whether buying weakness is a good idea. It tells you that buying weakness worked in an environment where the largest single input was pushing in one direction. Those are not the same claim, and from the inside they are completely indistinguishable, because both of them produce the same thing on your statement: money.
This is the selection problem, and it is not a character flaw. Put ten thousand people in a three-year regime and the ones who survive to still be trading are disproportionately the ones whose instincts matched that regime. Not the ones who were right. The ones who were aligned. The regime did the selecting, it selected for a reflex, and it handed everyone with that reflex three years of evidence that they are good at this.
Nobody in that process did anything wrong. Nobody was reckless. The machine simply cannot tell you the difference between an edge and a tailwind while the tailwind is blowing.
Three years of profitable dip-buying is evidence that dip-buying worked in a one-way regime. It is not evidence that you have an edge. From inside a P&L, the two are indistinguishable.
The only instrument that measures this
And here is why Wednesday matters far more than the 25 basis points.
A regime change is the only event that can separate the two. It is a test you cannot study for, cannot opt out of, and did not know you had scheduled. For three years there was no way to find out whether you had a process or a posture. Now there is one, it is running, and everybody is being graded at once.
Most people reading this have never traded a hiking cycle with real size. Not because they are inexperienced, but because there has not been one to trade. The pattern library was built in a single climate.
The thing about a first exam is not that you fail it. It is that you do not know your own score until it arrives, and by then the paper is already marked.
The 2029 problem
The second half of this is a timing story, and it is the same one we told after the Korean crash in August: the trader who gets carried out is rarely the one who ran out of conviction. It is the one who ran out of time.
The committee told you inflation does not get back to 2% until 2029. The market spent Thursday morning pricing something shorter than that.
One of those two is wrong, and the interesting thing is that it does not matter which. What matters is that the distance between them is measured in years, and almost nobody is positioned for years. They are positioned for the next print.
Being right about the destination and wrong about the duration is the single most expensive combination in this business, and it is the one that a three-year one-way regime trains you directly into, because in a one-way regime duration never costs you anything. You were never charged for being early. You are about to be.
For what it is worth, our own 2026 macro playbook, written in January, had disinflation in the subtitle. That is rather the point. Nobody's January view survives a regime change intact, including ours.
What this looks like from inside a prop firm
We should say the part that is against our interest, because it is true and you will work it out anyway.
Our pass rates were set in the old regime.
A challenge passed in 2024 or 2025 was passed in a market with a persistent bid underneath it, and that is an easier exam than the one being sat this month. We did not make it easier on purpose. The environment did. Anybody in this industry quoting a historical pass rate right now, including us, is quoting a number generated under conditions that have just changed.
The same applies to the rules themselves. A drawdown limit is calibrated against an assumption about how far things move in a day. When that assumption shifts, the rule gets tighter without anybody editing it, and the trader experiences that as the firm moving the goalposts when actually the floor moved.
We are not going to pretend that is comfortable. What we would say is narrow: a rule that ends a position at a level you chose in advance is survivable in a regime change, and a size chosen because the last three years were gentle is not. The failure we expect to see over the next two quarters is not people being wrong about direction. It is people carrying a size that was calibrated to a market that no longer exists.
If you want the mechanics of how gearing turns a wider daily range into a faster exit, we covered it in What Is Leverage? If you want to see how your own numbers hold up when ranges expand, the drawdown calculator is built for exactly that, and the current limits for every account size are on the challenge rules page.
The thing worth doing this week
Write down what you believe your edge actually is. One sentence, mechanical, no adjectives.
Then write down what has to be true about the world for it to work.
If the second sentence contains anything resembling "policy is easing", "dips get bought" or "volatility stays low", you do not have an edge yet. You have a position on the regime, which is a completely respectable thing to have, as long as you know that is what you own.
Almost nobody does this, for the same reason almost nobody writes down a forecast with a date on it. A written premise can be falsified. An unwritten one can quietly be retired and replaced with a new one that also feels like it was there all along.
If you have never written one, the trading plan and risk management lessons in the Pipster Academy are the place to start.
Bottom line
The Fed raised rates for the first time in 38 months, said it expects to do it again, and told you the job runs to 2029. The next morning the index rallied while the most rate-sensitive part of the market fell, and by the afternoon that had been bought too.
Both of those things happened because a reflex trained over three years fires faster than a conclusion.
The reflex might be right. It has been right for three years. The point is that until Wednesday there was no mechanism anywhere in the market capable of telling you whether it was skill, and now there is one, and it has started.
You are about to find out what you have. So is everyone else. So are we.
Frequently Asked Questions
Did the Fed raise interest rates in September 2026?
Yes. On 16 September 2026 the FOMC raised the federal funds target range by 25 basis points to 3.75% to 4.00%. The vote was unanimous, 12 to 0, and it was the first increase since July 2023.
How many more Fed rate hikes are expected in 2026?
The September dot plot shows 16 of 18 participants projecting a year-end rate above the current range: twelve at 4.125% and four at 4.375%. The median projection is 4.1% for the end of 2026 and 4.1% for 2027, up from 3.8% and 3.6% in June.
Why did stocks rally the day after the Fed hike?
Falling oil prices and a pullback in Treasury yields gave the index a reason, and three years of rewarded dip-buying gave it a reflex. The S&P 500 finished Thursday up about 1.2%. The Russell 2000, the most rate-sensitive part of the market, was still down at lunchtime before being bought back in the afternoon, which is the detail worth paying attention to.
What is a regime change in trading?
A regime change is a shift in the dominant force driving prices, such as the move from a three-year easing cycle into a hiking cycle. Strategies that were profitable because they aligned with the old driver can stop working without any change in the trader's skill or discipline. That is why a regime change is the only real test of whether a track record reflects edge or tailwind.
How does a Fed hiking cycle affect prop firm challenges?
Historical pass rates at every prop firm, including Pipster, were generated in a market with a persistent bid under it. Drawdown limits are calibrated to an assumption about daily ranges, so when ranges expand the same rule becomes effectively tighter. The practical response is to reduce position size in line with the new volatility rather than to argue with the rule.
What should funded traders do after the September 2026 rate hike?
Write down your edge in one mechanical sentence and the conditions that have to hold for it to work. If those conditions include "policy is easing" or "dips get bought", you hold a position on the regime rather than an edge. Then check that your position size was not calibrated to the last three years of low volatility.
Sources
Federal Reserve FOMC statement and Summary of Economic Projections (16 September 2026); CNBC, CNN, TheStreet, Kiplinger and Quartz coverage of the decision and of Thursday's session; Bank of England 17 September decision via CNBC; BTIG market note via CNBC; index levels from Kiplinger, TheStreet and Trading Economics. Thursday lunchtime figures are intraday. Thursday closing figures are approximate, taken from CNBC index quotes shortly after the close on 17 September 2026.