Update — September 1, 2026
This article was written on August 28, before the week it describes began. Warsh delivered his speech on August 28, and his tone turned out to be hawkish: he stated that inflation remains far too high, that the better inflation readings this summer do not indicate that the underlying trend has meaningfully improved, and that rate hikes may be necessary in the coming months. The %33.9 September pricing mentioned in the text shifted markedly upward following the speech; these lines have been left as they were to reflect the picture on that day. The first days of the week are now behind us; the employment report on September 4 and the FOMC on September 16 are still ahead. We have a separate article for the full September calendar.
Kevin Warsh's first opening speech as Fed Chair at Jackson Hole is being delivered today at 14:00 GMT. Symposium speeches set the direction for the market, but what they set direction for is mostly a framework: they explain which variables monetary policy will look at and which risks it takes more seriously. Figures are brought by data, not speeches. Following the symposium that ends this week, it is precisely time for that.
There are two full data weeks until the September 16 FOMC decision, and the first is by far the busiest. In the week of August 31 - September 4, two central bank decisions (RBNZ and Bank of Canada), Eurozone inflation closely concerning the ECB, two US ISM surveys, and the main event of the week, the US non-farm payrolls report, line up.
The current state of pricing is as follows: According to CME FedWatch, the probability of a 25 basis point rate hike at the September meeting is %33.9, while the probability of at least one rate hike by December is %74. Rate cut pricing is almost non-existent. The distance between these two numbers means the market is saying "a hike will come, but not this month." The data for the coming week could bring these two ratios closer together or push them further apart. Below, we will address the calendar day by day, not as a list, but by showing how each topic connects to September 16.
Monday and Tuesday: China PMIs and European Inflation Open the Week
The week begins on 31 August at 01:30 GMT with China's official manufacturing and services PMI data. Previous readings were 49.2 for manufacturing and 49.0 for services. In PMI surveys, the 50 level is the threshold between expansion and contraction; both series being below the threshold means that both main pillars of the Chinese economy reported a contraction compared to the previous month. The Caixin manufacturing PMI arriving on 1 September at 01:45 GMT (previous 50.9) uses a different sample than the official series and covers mostly private sector and exporting firms; the divergence between the two is important information on its own.
Why do Chinese data concern prices in the European and US sessions? The transmission channel is not direct, but operates through commodity demand. When Chinese industrial production slows, expectations for copper, iron ore, and energy demand decline; this translates into pressure on the currencies of commodity-exporting economies, namely the Australian dollar and New Zealand dollar. AUD and NZD are among the major currency pairs that react fastest to Chinese data during the Asian session. The second link in the same chain is that weak commodity demand acts to reduce global goods inflation - which serves as a disinflation argument, albeit a weak one, in the US inflation debate.
On 31 August at 12:00 GMT, Germany's preliminary harmonised CPI arrives (previous +2.8%). On 1 September at 06:00 GMT, German retail sales (previous -1.1%) will be released, followed at 09:00 GMT by Eurozone preliminary CPI and core CPI (previous annual +2.9% and +2.5%). The German data serves as a leading signal for the region as a whole; because national data arrives a day earlier, the Eurozone figure is often already partially priced in by the market.
The main issue here is the level. The ECB's medium-term target is 2%. A headline reading of +2.9% and a core reading of +2.5% point to a level above the target, but not far enough to cause panic. Core inflation remaining below headline suggests that a portion of the price pressure stems from volatile items such as energy and food; the reverse, meaning core remaining sticky, is the scenario central banks dislike most. For EUR/USD, this data will be the first real test of ground for a recovery attempt by the pair, which fell to around 1.1650 following the US PCE on 26 August.
Tuesday ISM manufacturing: The most concrete basis for the "economy is not collapsing" argument
On 1 September at 14:00 GMT, US ISM manufacturing PMI will be released; previous reading was 55.6. To understand why this figure is the week's most underappreciated headline, one needs to recall the debate at the July FOMC. At the meeting, interest rates were held steady in the 3.50-3.75% range, but Logan, Hammack, and Kashkari dissented in favor of a 25 basis point hike. In other words, there is a visible minority within the committee saying "inflation is still high, and the economy is in a position to handle it."
ISM manufacturing stands at 55.6, well above the 50 threshold. This is data that directly weakens the recession narrative. The manufacturing sector is a relatively small part of the US economy, but it is the most cyclically sensitive part: it reacts quickly to rate hikes, the inventory cycle, and global demand. In an environment where interest rate hikes break the economy, it is among the first to deteriorate. A manufacturing survey above 55 means this breakdown is not yet visible.
The logic leading from here to rate hike expectations is as follows: if the growth side holds up, one of the reasons holding the Fed back in the face of core PCE remaining sticky at an annual 3.3% weakens. In the 26 August data, headline PCE came in at an annual 3.7%, above expectations (3.6%), strengthening the dollar broadly and lifting Treasury yields. A strong ISM manufacturing reading will reinforce this reaction, while a weak reading will work to pull down the 33.9% probability of a hike in September. Second-quarter GDP at an annualized 1.5%, remaining below the first quarter's 2.1%, acts as a counterweight preventing the picture from being one-sided.
Wednesday: Two rate decisions, one GDP, and NFP's flawed harbinger
On 2 September, the calendar intensifies. At 01:30 GMT, Australian second-quarter GDP data arrives (previous quarterly +0.3%, annual +2.5%). When read alongside China's PMIs from the previous day, a consistent or contradictory story for AUD will emerge; how Australian growth behaves at a time when commodity demand is weakening carries information regarding RBA expectations.
At 02:00 GMT, the RBNZ announces its rate decision; the current policy rate is 2.50%. More important than the decision itself is the accompanying statement and rate path projection, if any - in small economies, central bank communication can generate larger price movements than the decision. At 13:45 GMT, the Bank of Canada decision arrives; current rate is 2.25%. The Canadian decision, along with the Canadian unemployment rate to be announced two days later at 12:30 GMT (previous 6.4%), could create pressure on USD/CAD extending into the second half of the week.
However, the main headline on the US side of the day is the ADP private sector employment report arriving at 12:15 GMT (previous 44,000). One needs to be honest about ADP: this data is not a reliable predictor of Friday's official non-farm payrolls figure. It is produced with a different methodology from payroll processing data; it does not cover public employment, and significant deviations from the official figure have been observed in the past.
So why does everyone watch it? Because it is the only concrete employment signal before Friday, and the market tends to react disproportionately to the single number it has. ADP's practical value lies less in its predictive power than in shifting expectations ahead of Friday, thereby altering the magnitude of surprise at the moment of NFP. A very strong ADP raises NFP expectations and can cause the same NFP figure to be priced as a "disappointment." It is more accurate to read this data not as a forecasting tool, but as an indicator of expectations.
Thursday: Why does ISM services carry more weight than manufacturing?
We start on 3 September at 01:45 GMT with China's Caixin services PMI (previous 50.1) - just above the threshold, representing an almost flat services sector. At 06:30 GMT, Swiss CPI arrives (previous monthly -0.1%, annual +0.4%). Swiss inflation running this close to zero is what structurally distinguishes the Swiss franc from other majors; this low-inflation regime is the backdrop for why CHF is favored when safe-haven demand arises.
The weight of the day, however, lies in the US ISM services PMI arriving at 14:00 GMT (previous 54.1). The services sector makes up a much larger portion of the US economy than manufacturing. This single fact makes the ISM services survey more decisive than ISM manufacturing - yet due to market habits, manufacturing data often gets more headlines.
ISM services has an additional significance regarding the inflation debate. Services inflation is much stickier than goods inflation; it is tied to wages and takes a long time to come down once established. A major part of the answer to why the Fed takes the 3.3% level in core PCE so seriously lies in the services item. The prices paid and employment sub-components of the survey therefore carry more information than the headline figure, and the employment sub-index of ISM services may serve as a more meaningful clue for the next day's NFP than ADP.
Friday: The employment report where three figures could contradict each other
4 September 12:30 GMT, the most decisive moment of the week and likely the period up to 16 September. US non-farm payrolls, average hourly earnings, and the unemployment rate are released simultaneously. Previous readings: 57,000 payrolls, monthly +0.1% earnings growth, 4.2% unemployment.
What gives extra weight to this report is recent history. July non-farm payrolls came in at -23,000, meaning the economy suffered net job losses that month; additionally, significant downward revisions were made to previous months. The revision issue is critical: the market reacts instantly to the headline figure, whereas in the same release, data for the preceding two months might be revised down substantially. Therefore, on 4 September, one needs to look not only at the new figure, but also at the July and June revisions.
The real difficulty is this: the three components in the report can contradict each other, and often do. The payroll figure may come in weak while average hourly earnings come in strong - in which case the labor market is cooling but wage inflation remains stubborn, representing the most uncomfortable combination for the Fed; it provides neither a clear justification for easing nor a comfortable one for tightening. Or payrolls could be strong while the unemployment rate rises; the reason for this is often people re-entering the labor force in hopes of finding work, raising the participation rate, which is actually a positive development but looks bad on the headline.
In terms of pricing, the link is direct: probability of a hike in September is 33.9%, and 74% by December. If strong payrolls and strong earnings come together, the market is expected to push the September probability toward the December probability - meaning the idea that "a hike is coming" is maintained while its timing is brought forward. Weak payrolls and soft earnings would pull both probabilities down; however, considering that rate cut pricing is near zero today, room for downward movement appears narrower than to the upside.
On the same day at 09:00 GMT, Eurozone retail sales (previous monthly -0.3%, annual +0.7%) and at 12:30 GMT, Canadian unemployment rate (previous 6.4%) will be released. The Canadian data coinciding with the exact same minute as the US report means that in USD/CAD, employment data from both economies is priced simultaneously at that moment; volatility in this pair at that minute can be higher compared to single-sided data.
When the report is published, the sequence to look at is as follows:
- –Payroll number - the headline figure, but the most misleading one on its own
- –July and June revisions - can completely change the meaning of the headline
- –Average hourly earnings - the primary indicator of the inflation side for the Fed
- –Unemployment rate and participation rate - participation explains why the rate changed
- –Not the price in the first five minutes after the release, but the close of the first hour - the initial reaction is frequently reversed
How to trade this week?
In high-data weeks, most of the money lost comes not from an incorrect directional prediction, but from carrying a correct directional prediction with the wrong position size. The items below are not a strategy recommendation, but a checklist of technical topics to verify during a calendar-heavy week:
- –Review the size of open positions prior to data release times. September 1 at 14:00, September 2 at 12:15, September 3 at 14:00, and especially September 4 at 12:30 GMT are the four key moments to mark this week.
- –Check your margin level before the data release, not during it. Trying to open the account summary and calculate during a sudden move means making a decision at the worst possible time. Calculating the required margin in advance using the Position Calculator simplifies this process.
- –Expecting spread widening during ISM and NFP minutes is realistic. This is not a broker issue, but a natural result of liquidity providers reducing risk; it is also seen in account types that offer tight spreads under normal conditions.
- –Widening the stop distance and lowering the lot size is a more consistent approach than carrying a large lot with a tight stop. While maintaining the same risk amount, the likelihood that market noise at the moment of the data release closes the position prematurely is reduced.
- –Account for the possibility of order slippage: at the time of the data release, a stop order may execute not at your specified price, but at the next valid price. Accounts with negative balance protection limit the extreme end of this risk, but do not eliminate the loss.
- –Convert the times for the week to your local timezone in advance. All times listed here are GMT; leaving a position open at the time of data release due to a time miscalculation is a common and entirely preventable error.
- –Monitor not the data itself, but the position of the data relative to expectations. What moves the market is not the absolute value of the figure, but its deviation from consensus.
Where to follow the calendar?
FXPARTNER's Economic Calendar page is a sufficient starting point to see this week's headlines and times in one place; the importance level of data and previous readings are listed here. The content we previously published on how to read the calendar can also be found under FXPARTNER Academy.
On the broker side, XM's interactive economic calendar is a practical tool in that it shows the expectation-actual comparison within the platform as soon as data is released. On the educational side, XM also has daily live webinars in 23 languages and the Tradepedia video library. These are tools that make tracking the calendar easier; they do not substitute for trading decisions and are not a source of advantage on their own. XM also has its own limitations: raw-spread account options are limited, leverage varies by region and legal entity, and cTrader is not supported.
Instead of choosing a broker based on this week alone, comparing license, cost, and withdrawal conditions via the Broker Search tool and Broker Rankings is a healthier approach. Heavy data weeks are a good opportunity to observe a firm's execution quality - but you do not have to test this with real money.
Conclusion: Will the gap between the two rates narrow?
It is possible to look at this week with a single question: will the distance between the 33.9% the market gives for September and the 74% it gives for December close or widen? Strong ISM readings and a resilient employment report narrow this gap, effectively making September 16 a live meeting. A weak data set, on the other hand, maintains or widens the distance; rate hike expectations do not disappear, they simply shift forward on the calendar.
Within the Fed, there are already three names that entered dissenting votes in favor of a hike, and core PCE stands at 3.3%. In contrast, July employment came in negative, retail sales were -0.6%, and second-quarter growth fell below the first quarter. Thus, there is tangible data in favor of both tightening and holding - the function of this week will be to show which side of the scale will gain weight.
Perhaps the most useful thing that can be said from a practical standpoint is this: in weeks like this, guessing correctly is less important than surviving. Managing position size and margin level in a week with four separate data events is a higher-yield endeavor than knowing which number will arrive.
This content is for general information purposes only and is not investment advice; leveraged transactions involve high risk and you may lose all of your capital.
Best 5 brokers
01XM Global
9.5IndexXM Global
- Min. deposit
- $5
- Leverage
- 1:1000*
02Lite Finance
9.2IndexLite Finance
- Min. deposit
- $10
- Leverage
- 1:1000*
03FxPro
9.1IndexFxPro
- Min. deposit
- $100
- Leverage
- 1:2000*
04MultiBank Group
9.0IndexMultiBank Group
- Min. deposit
- $50
- Leverage
- 1:1000
05AvaTrade
8.9IndexAvaTrade
- Min. deposit
- $100
- Leverage
- 1:400*
Sponsored links. Ranking is based on the FXPARTNER Index score, which consists of regulation, cost, platform, and withdrawal axes; affiliate revenue does not affect the ranking.