US bond market turmoil is hitting stocks: What ASX investors need to know

Back in May I argued that rising bond yields are a tax on every asset you own. Government bonds set the price of risk-free money, and everything else – shares, property, gold, and your super balance – is priced relative to that. The higher the risk-free rate, the more everything else usually has to fall to compensate.

Three months on, that tax has gone up again – and this week something more interesting happened than another leg higher in yields:

On Tuesday: the 30-year US Treasury yield briefly traded as high as 5.327%, the most since June 2007, Germany’s 30-year Bund reached a 15-year high, and Japan’s 30-year cleared its own spring peak.

On Wednesday: hours before it had to walk into the market and ask investors for another US$16 billion, Washington blinked. A plan to quell rising yields was announced. Stocks, gold, Bitcoin and other risk assets spiked.

On Thursday: Treasury Secretary Scott Bessent appeared on CNBC to explain what he was doing. By the end of the trading session, the stock market rally was in tatters.

Let’s make sense of it – how bonds work, why the long end has come unstuck and what Washington’s attempts so far to solve the problem do and don’t fix. Most importantly, we’ll detail the key numbers ASX investors should be watching from here.

Bonds 101: The most important market in the world

A bond is a loan. You hand over your money, the borrower pays a fixed rate of interest for an agreed period, and at the end you get your money back. A government bond is that arrangement with a country’s treasury on the other side.

What many don’t realise is that once a bond is issued, it trades. The interest payment, or coupon, is fixed in dollars – but the price of the bond is not. The two move like a seesaw. Buy a bond paying $5 a year for $100 and your yield is 5%. If the price falls to $80, that same $5 is a 6.25% yield to whoever buys it next. Prices down, yields up, and vice versa.

Now to the bit that matters for your shares. Because US government debt is regarded as the safest asset in the world, its yield is treated as the risk-free rate – the return you earn for taking no risk. Every other asset has to beat it.

A share is a claim on future profits, and the future is uncertain. So, to value a share’s uncertain future profits today, you must discount them back at a rate analysts typically peg to the risk-free rate. Consider:

• A company earning $1 a year, discounted at 5%, is worth $20.
• A company earning $1 a year, discounted at 2%, is worth $50.

The same earnings, different discount rates, different values. That’s why the bond market is integral to your portfolio.

Why long-dated bond yields came unstuck

Short-dated yields largely follow the central bank, the RBA in our case, or the Federal Reserve in the USA. But the yields on longer-duration bonds – those maturing in 10, 20, and 30 years – are usually driven by several other factors.

Over longer timeframes, inflation risk, political risk, and risk that the government in question will be unable to pay you back all come into consideration. The extra compensation investors demand for wearing these risks is called the term premium, and it has been climbing hard.

Three main factors have driven the recent rise in term premium for US government bonds:

The volume of paper coming: The US budget deficit is running near US$1.8 trillion so far this year; the Congressional Budget Office has lifted its 2026 estimate to US$2.1 trillion, US$200 billion more than it forecast in February; and the national debt passed US$40 trillion for the first time this month.

An unresolved argument about interest rates: The Iran conflict has kept energy prices elevated, and markets that began 2026 expecting rate cuts now expect hikes. But the Fed is split on whether it should act, as tightening into an energy shock risks real damage to the economy. When the path of short rates is unsettled, lenders want more to commit money for 30 years.

A new competitor for the same money: The hyperscalers – the giant cloud and AI operators, led by Microsoft, Alphabet, Amazon, Meta, and Oracle – are funding an enormous data centre buildout with borrowed money. Big Tech bond issuance has already topped US$220 billion this year, double last year’s full-year total. Much of it is long-dated, and these are companies with pristine balance sheets, some rated above the US government itself. To a pension fund needing a safe 30-year asset in the current political environment – a 30-year Microsoft bond and a 30-year Treasury start to look very similar indeed!

What treasury did – and what the Secretary said next

From 9 September to 4 November, the Treasury will at least double its long-end buyback operations, from US$2 billion to at least US$4 billion each, across the 10 to 20-year and 20 to 30-year maturity buckets. In plain terms, the government intends to buy back its own long-dated debt: fewer long bonds outstanding, less supply for the market to absorb, prices supported and yields pushed down.

Importantly, Treasury isn’t printing money to pay for the purchases – that would be quantitative easing – and it’s the Fed’s tool. The buybacks will be funded by issuing more short-dated bills, so Treasury is effectively swapping long-term debt for short-term, leaving the total unchanged. And it’s not large – at US$4 billion per operation, we’re talking roughly one hundredth of one percent of all US government debt.

The official framing was that the move “reflects Treasury’s desire to provide greater liquidity support in longer-dated nominal sectors.” In other words, they’re shuffling the US debt deck.

The next day, Bessent went on air, and the framing changed. His comments made clear that Treasury believed the market wasn’t pricing long-term US debt appropriately, telling CNBC that the aim was “to signal that we think that this is a thinly traded area of the market,” and that “we believe that the yields don’t reflect the underlying fundamentals.”

He also hinted that Treasury’s operations “could be more than the $4 billion per issue.” When asked how much more he was willing to do, he said “we have a big toolkit, so we will see.” That’s not deck shuffling, that’s a policymaker telling the market its price is wrong and warning he’ll keep buying until it agrees.

And the timing invites obvious scrutiny. The announcement landed hours before a US$16 billion 20-year auction, a week after a 30-year auction cleared at 5.216%, and after three straight sessions of falling equities with rising yields widely blamed as the cause. It has the look of an administration that didn’t want a failed auction on the front page while it’s prosecuting a war.

The market has seen this move before. In April 2025 a savage bond sell-off preceded the White House’s 90-day pause on its reciprocal tariffs, a reversal the president explained by saying people were getting “yippy” and “a little bit afraid” about the bond market. That episode gave us TACO, or Trump Always Chickens Out, which describes the administration’s habit of flinching at market pressures and changing its course. Treasury’s buyback is the latest flinch, and Bessent’s job the next day was to try and convince the market that it wasn’t.

When interventions work, and when they fail spectacularly

To understand what happens next, we must take a quick look at history. An official intervention is not really about the size of the stick (or the carrot). It is about belief. When a credible institution says it will defend a level, traders won’t put up a fight – they assume it has a bigger balance sheet, a longer time horizon, and no impatient shareholders watching profits. The threat does the work.

The modern textbook case is the European Central Bank in 2012: three words from Mario Draghi, “whatever it takes”, and the eurozone bond crisis eased. The bond-buying programme built to back those words up was never used at all – not one euro was ever spent under it.

Now a counter-example: in September 1992 the Bank of England was defending sterling’s place in Europe’s exchange rate mechanism at a level the underlying economy could not support. It raised its policy rate from 10% to 12%, announced a further rise to 15%, and spent billions of pounds of reserves buying its own currency. George Soros’s Quantum Fund concluded the fundamentals would win and sold sterling anyway – reportedly clearing more than £1 billion in profit when Britain crashed out of the mechanism.