Markets are surprisingly good at handling one worry at a time. Give investors a single villain, an oil spike, a hawkish central banker or a wobble in technology stocks, and they will argue about it, price it and move on. What markets handle far less gracefully is a pile up.

That is what the past week delivered. Three separate market shocks arrived almost at once. Oil pushed back above US$100 a barrel, the US Federal Reserve raised interest rates for the first time since 2023, and the people building the most advanced artificial intelligence systems publicly asked their own industry to slow down. Any one of these would have unsettled a nervous market. Together they produced something closer to a stampede.

The temptation in weeks like this is to treat everything as one big risk and react to the whole bundle. I think that is a mistake. These three shocks arrived together, but they travel through the economy at very different speeds, and they will not leave together either.

Sorting out which is which is most of the work.

Why this week felt different

The local numbers tell the story. The S&P/ASX 200 fell 264 points, or 2.94%, last week. On Tuesday it closed at 8,672.5, its lowest finish since mid June and 6.7% below the record intraday high of 9,296.7 set on 6 August. By the halfway point of September, the index was down 4.5% for the month, well beyond its average September decline of 1.78%.

None of that is catastrophic. It is a pullback, not a crash. But speed matters, because speed is what turns reasonable concern into poor decisions.

Shock one: oil is back above US$100, and the workarounds are wobbling

The Strait of Hormuz has been effectively closed to normal commercial traffic since late February. Daily transits have been running at a fraction of the roughly 130 ships a day that passed through before the conflict.

For months, markets had learned to live with that. What changed this week was the workaround. Saudi Arabia’s East-West pipeline, a vital route for moving crude around Hormuz, was shut after an attack, with no clear timeline for its return. Brent crude closed above US$105 a barrel and West Texas Intermediate pushed towards US$102.

Here is the point worth holding onto. Oil is not just a headline. It is a transmission mechanism.

It shows up at the petrol pump within days, in freight costs and airfares within weeks, and in wage claims and inflation expectations within months. That is why oil matters more to the next shock than it does to the share market on any given afternoon.

Shock two: the Fed has moved, and the rates story has gone global

The August US inflation report, released on 11 September, set the scene. Consumer prices rose 0.4% for the month and 3.4% over the year, with petrol accounting for more than a third of the monthly increase. Core inflation, which strips out food and energy, rose 0.3%, a touch hotter than expected. By the eve of the meeting, futures markets were pricing better than a 90% chance of a Federal Reserve hike.

The Fed delivered. Overnight, the committee voted unanimously to lift its target range by 25 basis points to 3.75% to 4.00%, its first increase since July 2023. Chair Kevin Warsh described the move as removing “a dose of accommodation”, and the committee said it would support a “timelier return” to its 2% inflation goal.

The hike itself was fully priced. What mattered more was the message around it. The Fed’s updated projections show 16 of 18 participants expecting at least one more increase this year, four of them seeing room for two, with the median rate ending 2026 around 4.1%. Officials also nudged their inflation forecasts higher and now see core PCE inflation at 3.4% by year end.

In other words, this was not framed as a one off insurance policy. It looks more like the opening move of a tightening cycle.

Warsh was also candid about the limits of his tools. He acknowledged the Fed cannot switch off an energy price shock. Its job is to stop that shock spreading into broader inflation. That is precisely the transmission problem described above.

Markets did not enjoy the message. US shares were higher before the announcement but turned lower during Warsh’s press conference. The Dow fell 1.2%, the S&P 500 lost 0.45% and the US 10 year Treasury yield pushed back above 5%, around its highest level since 2007.

The Fed is not alone. The European Central Bank moved last week, the Bank of Japan announces its decision on Friday with another hike in play, and some strategists now expect at least eight G10 central banks to lift rates by year end.

Australia is not an exception. It is closer to the front of the queue.

Markets are pricing a high probability that the RBA lifts the cash rate to 4.60% on 29 September, which would be its fourth increase this year and take the cash rate to its highest level since 2011. Australian 10 year bond yields are also sitting at their highest levels in around 15 years.

That combination matters because the Australian economy is unusually rate sensitive. Household wealth is heavily concentrated in housing, most mortgages are variable, and the housing market has already turned. Cotality’s national home value index fell 0.9% in August, the fifth consecutive monthly decline, leaving values 3.6% below the March peak. Through winter, 93% of capital city suburbs recorded falls.

Consumers have noticed. The Westpac Melbourne Institute consumer sentiment index fell 5.2% in September to 84.4, deep in pessimistic territory. Morgan Stanley’s local strategy team has gone further, moving to what it describes as a near maximum underweight on Australian shares, on the view that the local monetary and fiscal backdrop is more adverse than elsewhere.

Shock three: the builders asked to slow the build

The third shock was the strangest, because it came from inside the house.

Over the weekend of 12 and 13 September, Anthropic chief executive Dario Amodei published an essay titled “We Must Pace the Frontier”, calling for an intentional global slowdown in frontier AI development. Within hours, Sam Altman of OpenAI, Demis Hassabis of Google DeepMind and Elon Musk had publicly agreed. Altman separately said OpenAI would not go public in 2026.

Markets did what markets do. On Monday the Philadelphia Semiconductor Index fell 5.9%, SoftBank dropped nearly 11% in Tokyo, South Korea’s Kospi lost 3.3% and ASML fell around 6% in Europe. Nvidia’s Jensen Huang pushed back publicly, with President Trump dialling in on speakerphone to call the slowdown push a hoax.

It is worth being clear about what actually changed. A group of chief executives said they want to pace the development of the most capable models. Nothing, at least so far, has changed in the order books.

Oracle, reporting on 10 September, lifted its contracted backlog to US$664 billion after signing more than US$30 billion of new AI cloud contracts in the quarter. Bank of America continues to point to roughly US$3 trillion of AI infrastructure spending by 2030.

Well known investors are split. Ray Dalio wrote on 9 September that “we are now in a bubble”. Cathie Wood called the AI doom narrative “ridiculous”. Michael Burry has suggested the slowdown push is tied to upcoming listings. When the experts disagree this loudly, it is usually a sign the market is short on facts and long on narrative.

Market shocks are not all created equal

This is the distinction I think investors should focus on.

Oil and rates are transmission shocks. They move slowly, but they reach almost everything: household budgets, company margins, borrowing costs and valuations. Their effects tend to arrive after the headlines have faded, which is precisely why they are easy to underestimate.

The AI slowdown call, at least for now, is mostly a sentiment shock. It changed how much investors are willing to pay for future growth, not how much companies have already committed to spend. That could change if words become regulation, or if customers start cancelling contracts. But this week, the prices moved much further than the facts did.

Australia offers a neat illustration of how the shocks interact. Westpac recently cited the scale of data centre investment as one reason it expects the RBA to keep tightening. In other words, the AI build out is now big enough to be considered inflationary. Yet NextDC, one of the most direct listed exposures to that build out, fell 14% in a month. Building data centres consumes enormous amounts of capital long before it earns a return, and higher rates make that capital dearer.

At the same time, Firmus is reportedly preparing an ASX float that could raise up to A$7 billion at a valuation near A$50 billion. In the same week the industry’s leaders talked about slowing down, one of the largest listings in Australian history is being pitched on the idea that demand will keep compounding.

Both stories can be true. What they tell you is that the AI trade has moved from a question of whether to a question of price.

The second order effects worth watching

The first order effects are obvious: dearer petrol, higher mortgage rates and weaker technology shares. The second order effects are where the real investment questions sit.

The first is margin pressure in businesses that cannot pass on costs. Energy, freight and wages all feed into the cost base with a lag. Companies with genuine pricing power will manage. Those competing mainly on price are likely to feel it in the next reporting season, not this one.

The second is refinancing risk. Companies and property owners who borrowed on the assumption that rates had peaked may now face harder conversations with their lenders. Balance sheet strength, which looked boring during the rally, becomes a genuine competitive advantage.

The third is the wealth effect. With housing making up the bulk of Australian household wealth, even modest price declines tend to show up in discretionary spending over the following quarters.

The fourth is the cash hurdle. When cash and term deposits pay more than 4%, investors quite reasonably ask more of their shares. That does not make shares unattractive. It makes the gap between good businesses at sensible prices and fashionable stories at stretched prices far more important.

What could go wrong, and what could go right

It would be dishonest to pretend this is simply noise. The risks are real.

Oil could go higher. Another attack on Gulf infrastructure, or a prolonged pipeline outage, would add to the inflation problem just as central banks are trying to contain it. Rising rates alongside slowing growth is the stagflation scenario markets fear most, and Australia, with high household debt and a softening housing market, is more exposed to it than most developed economies.

The charts are not encouraging either. The ASX 200 has slipped below its 200 day moving average, and market technicians at IG are watching the 8,500 level as the next line of support. If that gives way, they see the March low of 8,262 coming back into view. A tightening cycle that spreads across much of the developed world also tends to compress valuations for longer than investors expect.

The AI debate could harden too. If the slowdown call turns into coordinated regulation, the spending assumptions behind many AI linked valuations would need to be revisited.

But the counterpoint matters. Markets tend to overprice clustered bad news. A credible reopening of Hormuz would pull oil sharply lower and ease pressure on inflation. A Fed that has now shown it is serious about inflation could, in time, help anchor expectations and take some heat out of long bond yields. And AI demand, judged by contracted backlogs, has not yet shown any sign of cracking.

The risk is not only that things get worse. It is also that investors sell good assets at poor prices just before the fog lifts.

What should long term investors actually do?

This is not a call to make heroic macro bets. Nobody knows exactly when Hormuz reopens or how many more hikes the RBA has in it. The practical question is how to hold a portfolio through a period when the macro backdrop has become less forgiving.

A few principles seem sensible.

Separate the shocks. Ask of each holding whether it is exposed to a transmission shock, such as energy costs and interest rates, or a sentiment shock, such as a shift in narrative. The first deserves more attention than the second.

Look hard at balance sheets. In a rising rate world, low debt and strong cash conversion are not defensive luxuries. They are the difference between a company that can invest through the cycle and one forced to raise capital at the wrong moment.

Respect pricing power. Businesses that can lift prices without losing customers are the natural hedge against an oil led inflation problem.

Distinguish backlog from cash. In AI and infrastructure, contracted revenue is encouraging, but it only matters if it converts into cash flow and decent returns on capital. Oracle’s shares are down more than 20% this year despite a record backlog, which tells you the market is already asking that question.

Do not extrapolate one week into a decade. Selloffs driven by a cluster of headlines often create the best opportunities for patient investors, particularly among quality small and mid caps that get sold alongside everything else.

Rebalance rather than react. If a portfolio has drifted after a strong run, a volatile month can be a reasonable time to tidy it. That is very different from abandoning a plan because of a bad week.

TAMIM Takeaway

In a single week, oil pushed back above US$100, the Federal Reserve raised rates for the first time since 2023 and signalled more to come, global bond yields hit multi year highs, and AI’s own leaders called for a slowdown. The ASX 200 responded with its sharpest fall in months.

The market has treated these as one big risk. They are not. Oil and rates are slow moving transmission shocks that will shape inflation, household spending and company margins for months after the headlines fade. The AI slowdown call, for now, is mostly a sentiment shock that has moved prices far more than it has changed spending.

For long term investors, the right response is neither panic nor bravado. It is discipline: strong balance sheets, genuine pricing power, sensible valuations and a willingness to buy quality when the crowd is busy selling everything at once.

Headlines arrive together. Their consequences rarely do.