Gold rallies to around USD 4,400, but the case against it hasn't moved

• On 5 August 2026, hopes of a Strait of Hormuz reopening eased oil supply risks and improved the inflation outlook, prompting markets to price greater odds of Fed rate cuts. Lower expected rates pushed real yields down, lifting gold 4% to around USD 4,250. On 10 August, those hopes collapsed after Trump added sweeping compensation demands, triggering renewed conflict concerns and a flight to safety that pushed gold to around USD 4,400 by 12 August.
• The July FOMC held rates at 3.50% to 3.75% on a 9-3 vote, the most divided decision since 2016, with three regional presidents dissenting for an immediate hike; markets were pricing one to two hikes by year end as of 30 July.
• A reheated Iran conflict (a collapsed ceasefire, tanker strikes in the Strait of Hormuz, an oil price spike) and a violent late-July chip and memory stock rout explained gold's safe haven bid through most of July.
• China's demand is pulling away from India's: the PBoC logged its largest monthly reserve addition in over two and a half years and June imports hit a two-year high, while India's net imports fell to their lowest since 2020 and the duty hike is fuelling a grey market.
• We only recommend holding between 0% and 10% of your portfolio in gold, purely as a diversifier. Gold tends to move independently of shares and bonds, but while it is often viewed as a safe-haven asset, history shows it has not consistently protected capital during every market crisis.

In late June, we argued the case against a large gold allocation had strengthened on every front: rates, ETF flows and Indian demand had moved against the metal. Gold was trading near USD 3,979 (24 June), down 29% from its January peak.

Seven weeks on, gold trades near USD 4,400 (12 August 2026), a rally, not a continued slide we had expected. At first glance, that looks like a challenge to our thesis. We don't believe it is, but it's worth being precise about what our thesis actually says: we aren't forecasting gold falls, our case is that the structural drivers, particularly the rate outlook, don't support raising a strategic allocation beyond a small diversifier position.

Tactical safe-haven demand can and does push the price higher in the near term, and it has, twice in six weeks: a reheated Iran conflict and a sharp semiconductor selloff put a safe haven bid under gold through most of July, and hopes for a Strait of Hormuz reopening deal then drove gold sharply higher again in early August as markets pared back Fed hike expectations, before those hopes collapsed on 10 August and sent oil, and gold, higher still, this time on renewed conflict risk rather than a friendlier rate outlook.

If anything, that back-and-forth argues against leaning harder into gold as a crisis hedge, the same event risk that pushed it up this time could reverse just as quickly if the conflict cools or the current de-risking fades. Layered on top is a demand story splitting in two: China's buying accelerates while India's erodes.

Related article: Markets are celebrating a Hormuz deal. History says not to.

Related article: We warned about gold since late 2025. It is now down 29% from its January 2026 peak.

Related article: Three strikes against gold: rates, central banks and India

The Fed: more hawkish, not less

On 29 July, the FOMC voted 9-3 to hold rates at 3.50% to 3.75% for a fifth straight meeting. The hold matched expectations, but the vote was the most divided since 2016: three regional presidents, Cleveland's Beth Hammack, Minneapolis's Neel Kashkari and Dallas's Lorie Logan, dissented for an immediate hike, each citing persistent inflation. Governor Christopher Waller had also flagged inflation concerns beforehand but voted to hold. Unlike June, there was no dot plot this meeting, so the vote split itself is now the clearest read on the committee's direction.

That direction points one way. As of 30 July, markets were pricing one to two hikes by year end, building on June’s dot plot, where the median 2026 projection had risen to 3.8% from 3.4%. Warsh pushed back on the idea that softer June inflation settles anything, describing the Fed’s stance as "a period of watchful thinking rather than watchful waiting". The bond market agreed: the 30-year Treasury yield rose to a 19-year high near 5.2%, a steepening that suggests investors worry the Fed is moving too slowly on inflation.

The inflation data is more ambiguous than the headline suggests. June CPI eased to 3.5% (core 2.6%), the first cooling in five months, but almost all of it came from lower energy prices, which have since reversed: Brent and gasoline are both up over 18% since the ceasefire collapsed in early July. One soft print built on an energy dip that has already unwound is unlikely to change the Fed's direction.

For gold, this reinforces rather than offsets the June headwind: the opportunity cost of holding a zero-yield asset stays elevated while the Fed's hawkish tilt holds.

Table 1: Fed funds rate expectations — March vs June 2026 dot plots

	March 2026	June 2026
Median 2026 rate projection	3.4% (one cut)	3.8% (one hike)
Officials projecting ≥1 hike	0	9 of 18
2026 PCE inflation forecast	2.7%	3.6%
Source: Federal Reserve Summary of Economic Projections, March and June 2026.

Related article: July FOMC recap: The Fed holds, but the long end isn’t convinced

Risk events: Iran and semiconductors revive the safe-haven bid

If rates alone drove gold, it should have kept falling after 29 July. It did not. The April US-Iran ceasefire and June memorandum broke down in early July: strikes resumed, Iran hit two tankers in the Strait of Hormuz on 13 July, and oil jumped over 9% as the US reinstated a naval blockade. Fighting paused again around 25-27 July, with Trump saying strikes had halted at Iran's request while warning attacks would resume without a new deal. The conflict has not been resolved, it has simply gone quiet again, a pattern arguably more durable for safe haven demand than a clean ceasefire.

The second event was a market-driven rather than geopolitical: a sharp correction in semiconductor and memory stocks in the second half of July. The Philadelphia Semiconductor Index fell into a bear market, more than 20% below its June peak; SK Hynix and Samsung Electronics each fell more than 13% in a session, and South Korea's Kospi suffered its eighth circuit breaker of the year. The triggers, doubts about AI infrastructure spending and profit-taking after an extraordinary rally, are sector-specific, but the risk-off impulse may have spilled into gold: SPDR Gold Shares recorded a USD 446.8 million weekly inflow in mid-July after a run of outflows.

This looks more like an isolated flight-to-safety flicker than a shift in trend. The gold inflow came against much larger outflows (-USD 9.02 billion) over the 1H26, while the semiconductor sell-off was driven by sector-specific concerns rather than broad risk aversion. With higher rates still creating an opportunity-cost drag, the episode does not yet point to a durable revival in gold ETF demand.

None of this stopped market forecasts turning more cautious on the medium-term path. A Reuters poll of 29 analysts (28 July) cut the median 2026 gold forecast to USD 4,509, down from USD 4,916 three months earlier, the first downward revision in eleven quarters. This is consistent with our view: as we expect Fed rate hikes this year, the rising opportunity cost of holding a zero-yielding asset like gold reinforces our cautious stance.

The Hormuz deal: repriced hike odds, not the inflation outlook

Gold extended its gains into early August as the dollar softened and oil fell on signs of progress in US-Iran talks. Then on 5 August, gold jumped more than 4% to around USD 4,250, its largest one-day gain since February, after Iran and Oman said they had reached an agreement on reopening the Strait of Hormuz and Trump said a deal could be struck within days. Markets responded by paring Fed hike expectations sharply: futures now fully price just one rate increase by year end, down from two a week earlier, and the priced-in probability of a September hike fell from 67% to 59% in a single session.

The picture since has been more mixed than a clean reversal. Trump added sweeping new compensation demands on 10 August that Teheran is almost certain to reject, tying any deal to conditions neither side currently accepts. Oil responded sharply: Brent rose about 5% on 10 August to near USD 88 a barrel, extending a rally of more than 5% over the prior three sessions, while European diesel futures jumped over 10% after refinery attacks in Saudi Arabia, Libya and Russia widened the disruption.

Iran also hardened its posture, naming a hardline former IRGC commander to head its top security body, and continues to tie any Hormuz reopening to the US lifting its naval blockade, releasing frozen assets, and ending regional strikes, none of which Washington has agreed to.

Priced-in September hike odds still have room to fall further before this settles: they've eased from 58% on 4 August to 50% as of 11 August, and the July CPI print, due 12 August, the first reading to capture the latest oil move, is the next catalyst that could push them lower still if inflation cools, or reverse that drift if it doesn't.

That said, our house view remains that the Fed hikes this year; what's still unsettled is the scale of the move, and that will stay data dependent. Energy prices, the actual channel through which the conflict feeds into inflation, are moving the wrong way: our July FOMC update showed June's disinflation was almost entirely an energy story, and this latest oil spike works against that channel, not with it.

Gold, meanwhile, has kept climbing through all of this, to around USD 4,400 by 12 August, evidence that its latest leg higher is now being driven by safe-haven demand rather than the rates story that started the rally on 5 August.

China vs India: two different demand stories

Beyond rates and geopolitics, the demand side is also splitting in two. Put on the same footing, net bullion imports for the April to June quarter, the gap is still wide, though the precise China total blends two data series.

China's customs data puts June imports at approximately 173 tonnes, a two-year high, on top of 157 tonnes in April and 151 tonnes in May, taking the quarter to about 480 tonnes, still around five times India's 98.1-tonne Q2 net imports. China's central bank widened the gap further: the PBoC bought 15 tonnes in June alone, its largest purchase since October 2023, extending a 20-month buying streak and taking reserves to 2,346 tonnes, 8% of total foreign exchange reserves; H1 purchases totalled 40 tonnes.

India, on the same Q2 basis, weakened on the imports side, complicating the original narrative. Net imports fell 23% year on year to 98.1 tonnes, the lowest since September 2020, with total demand down 6% year on year as jewellery (75.1 tonnes, -15% yoy) underperformed a steadier bar and coin market (50 tonnes, +9% yoy).

India’s duty hike (enforced in mid-May 2026) has pushed some buying underground rather than just cutting it. Legally imported gold now carries an 18% mark-up (15% import duty plus 3% sales tax), making smuggled gold meaningfully cheaper. That gap has made smuggling more worthwhile: customs seizures nearly doubled between mid-May and end-June, and officials expect illegal imports to top 100 tonnes in 2026. Official dealers, meanwhile, have actually been selling gold at a discount, USD 19 to USD 56 an ounce below the benchmark through July, as buyers waited for prices to settle. And most Indian consumers are trading in old jewellery for new rather than buying outright, so demand looks weak on paper even though gold is still changing hands.

The ETF picture is mixed. Chinese gold ETFs added 28.7 tonnes over H1 as a whole, but Q2 alone saw a net outflow of 21.8 tonnes, and June was the worst month on record for the funds, as a falling gold price and a rotation into equities pulled roughly USD 2.2 billion out in that month alone. India's gold ETFs, by contrast, stayed net buyers throughout: +4.2 tonnes in Q2, +2.5 tonnes in June, one of only a handful of major markets, alongside the UK, to record positive ETF demand in the quarter. On this specific metric, India's investors were more resilient than China's.

On a quarter-for-quarter basis, China's import volumes and central bank accumulation continue to dwarf India's, a genuine structural offset to rate and Western ETF headwinds. This is not, however, a broad-based Asian recovery: on the ETF measure specifically, it was Chinese investors who turned net sellers in Q2, not India's. One country is stockpiling gold. The other is trading last year’s jewellery for this year’s. That is not the same gold market.

Our position: gold continues to be a diversifier

Putting these threads together: the rally to around USD 4,400 reflects two distinct stories, a rate repricing that has only partially unwound and remains unsettled pending this week’s CPI print, then renewed safe-haven demand as the Iran conflict escalates again, neither of which points to a structural improvement in the case for gold. The Fed's hawkish tilt, reinforced by July's split vote, still raises the opportunity cost of holding a zero-yield asset. The demand picture offers a genuine but narrow offset, concentrated in China's imports and central bank buying rather than a broad-based recovery. None of that changes our conclusion from June.

We continue to recommend a small 0% to 10% allocation to gold for diversification. Gold’s long-run correlation with US equities is essentially zero, useful as a diversifier, but not a reliable crisis hedge.

Table 2: Correlation between gold and other asset classes

Asset	Longer-term (30 Jul 1971 to 31 Jul 2026)	Recent period (30 Jul 2021 to 31 Jul 2026)
US Equities	0.007	0.136
US Corporate Bonds	0.067	0.186
US Treasuries	0.070	0.206
World ex US Equities	0.144	0.301
Emerging Markets Equities	0.176	0.323
Commodities	0.389	0.351
Source: World Gold Council. Data as of 31 Jul 2026.

Our recommended products are the SPDR Gold MiniShares (NYSE: GLDM) and the LionGlobal Singapore Physical Gold A Acc SGD. Investors should assess suitability against their own investment objectives, financial situation and risk tolerance before investing.

Declaration

For specific disclosure, at the time of publication of this report, IFPL (via its connected and associated entities) and the analyst who produced this report hold a NIL position in the abovementioned securities.

This research report was prepared with the assistance of artificial intelligence (AI) tools. iFAST Financial Pte Ltd does not rely exclusively on AI for content generation; the content of this report, including all investment theses, ratings, price targets and conclusions, has been independently reviewed and verified by the research analyst(s) to ensure accuracy and professional int