What is the major shift in the gold market? On Bank of America’s latest flow read, it is this: gold is drawing its strongest wave of fund demand in nearly a year. Investor inflows have reached their highest level since October 2025. Gold-backed exchange-traded funds recorded a $6.4 billion weekly increase in holdings in August, the largest weekly gain in roughly ten months. The four-week moving average of those flows has also turned higher. That last detail is the one that matters. A single fat week can be a fluke. A rising four-week average is participation.
That is Bank of America gold as a flow story. It sits next to a metal that spent Tuesday, September 1, under $4,400 — a session low near $4,326, then a bounce toward $4,366 after ISM Manufacturing printed 54.6. Last week’s three-month high was $4,697. Kevin Warsh told Jackson Hole the Fed has “work to do” if inflation is not moving toward 2 percent at sufficient speed. September hike odds moved into the mid-60s. The 10-year yield was cited near 4.77 percent. Fund demand returning into that tape is the shift. It is not a repeal of the rates channel.
Could this change the gold bull market outlook? It can change the mix of who is buying. It does not, by itself, change the year-end math. This article is a gold price analysis of a flow signal against a hawkish week. It is not a recommendation to buy gold, gold mining stocks, or any other security.
Bank of America gold forecast work comes from more than one desk, and the numbers are not interchangeable.
Michael Widmer’s metals team is the price-path book. In January the bank talked $6,000 by spring. By June 22 Widmer wrote that hitting $6,000 looked unlikely for now, that the shift from “inflationary cuts” toward tighter policy reduced gold upside by about 50 percent all else equal, and that high U.S. deficits, a lack of fiscal consolidation and funding needs still left fuel in the tank for a later rally. On July 7 the desk cut its 2026 average gold price forecast by 14 percent, to $4,360 from $5,093, and said $5,000 remained in reach once the Fed’s tightening cycle ends. A Bank of America gold prediction that quotes $6,000 as if it were still the base case for this quarter is quoting a winter sentence in September.
Michael Hartnett’s allocation book is the flow-and-hedge book. Hartnett has kept gold as protection against dollar weakness and currency debasement. Later-summer strategy notes from the firm have treated gold as a hedge against U.S. dollar credit dilution, with commodities and resources preferred when long-end yields stay high and the dollar softens. The firm’s Bull & Bear Indicator has been cited in the sell zone, a backdrop in which gold is a hedge, not a momentum toy. Bank of America gold forecast 2026, in that register, is a portfolio sentence: own some metal because the fiscal arithmetic did not improve.
The September flow note sits between those desks. Central bank gold buying was the 2022–2025 load-bearing wall. Gold ETF demand going quiet in the first half of 2026 was the missing second engine. A $6.4 billion week and the best fund demand since October 2025 is the second engine coughing back to life. Gold demand trends that only count official tonnes will miss it. Gold demand trends that treat one August week as a new regime will overfit it. Weekly flows since the second half of 2025 have periodically exceeded $6 billion — and similarly large outflows have shown that positioning remains two-way. Persistence is the test BofA itself flagged. Episodic is the risk.
Central bank gold demand did not leave the building. The World Gold Council put official purchases at 289 tonnes in the second quarter, a record Q2, after a first quarter revised to 57 tonnes. Forty-five percent of reserve managers in the Council’s survey still intend to add gold over the next twelve months. Seventy-four percent expect the dollar’s share of reserves to fall over five years. Goldman Sachs, in an August 28 note, still uses 50 tonnes a month of official buying in 2026 against 17 tonnes before 2022, and a $4,900 year-end target. That is the floor under the gold market outlook. It is not Tuesday’s bid.
Mine supply has not sprung a leak that would cancel a flow shift. Seniors are still guiding flat-to-down ounces. Agnico Eagle is at 3.3 to 3.5 million ounces, now expected at the low end after the Barnat movement at Canadian Malartic, with AISC of $1,400 to $1,550. Newmont is pointing at about 5.3 million ounces. Barrick is guiding 2.90 to 3.25 million at AISC of $1,760 to $1,950. Gold supply and demand at the margin in 2026 is still an official-plus-investment story, not a new-mine story. When ETFs add $6.4 billion in a week, they are competing with the same above-ground stock the official sector has been absorbing. That is why the shift can matter even if the ounces from the pit do not change.
Gold investment demand is the swing variable Widmer has used for years. In an earlier 2026 note he sketched that a 14 percent rise in investment demand could support $5,000, and that much larger jumps would be required for the tail numbers that make good conference slides. August’s fund week is a data point on that curve. It is not the whole curve. Gold ETF demand that arrived while hike odds were still falling in mid-August is a different animal from gold ETF demand that holds after Warsh. The next four weeks of flow prints will decide which animal this is.
What does BofA’s gold outlook mean for investors? It means three numbers that must not be collapsed.
The 2026 average of $4,360 is a path-through-the-year figure set in July. Spot near $4,366 on Tuesday is almost exactly that average. A gold market forecast that treats $4,360 as a year-end target is misreading the line. Averages include the $4,000s of mid-year and the $4,600s of late August. They do not promise $4,360 in December.
The $5,000 mark is the “once tightening ends” waypoint. Tightening has not ended. Hike odds for September 16–17 are in the mid-60s. Payrolls are Friday. A Bank of America gold price prediction that brings $5,000 forward into this week is ignoring Widmer’s own sequencing.
The $6,000 twelve-month target from January is the long-horizon fiscal-and-debasement case: deficits around 6 percent of GDP, fading foreign Treasury holdings, reserve managers expecting a smaller dollar share. Widmer already said the spring timing is off. Hartnett’s hedge language still lives in that neighborhood. Gold outlook 2026 that keeps the fiscal case and drops the calendar is the only way those two sentences stay friends.
For a reader running a precious-metals sleeve, the flow shift is permission to stop treating the book as official-sector-only. It is not permission to treat $4,697 as support. Gold price target language from other houses — Goldman’s $4,900 year-end, Standard Chartered’s $4,600 twelve-month wealth-desk mark — still has to clear $4,400 first. Bank of America gold is a demand-mix update inside that range. It is not a new range.
They can, if the flow persists and the metal holds $4,300. They will not, if the $6.4 billion week is faded and gold loses $4,300 into payrolls.
Gold mining companies lever the metal. Best gold stocks 2026 and gold mining stocks to buy are phrases this piece will not complete. Gold mining stocks to watch, as a liquidity screen into a flow-and-Fed week, remain Agnico, Newmont, Barrick, Wheaton and Franco-Nevada. Margins at $4,366 are still wide against $1,400 to $1,950 AISC guides. Multiples that capitalized $5,000 as a 2026 average are not. Could gold mining stocks benefit from changing gold demand? If ETF demand stays and official buying does not roll over, the equities will try to re-rate with the metal. If the shift proves episodic, the equities will lead the metal lower, as they did after Friday. That is beta. It is not a shopping list.
Gold investment opportunities that use BofA’s flow note as a buy ticket on Tuesday afternoon are using a weekly statistic as a daily order. Gold investment opportunities that ignore the return of private demand after two years of official-led tape are missing the only new sentence in the September note.
The gold bull market of 2026 is already a year with a January record near $5,595 to $5,600, a mid-year washout toward $4,000, an August squeeze to $4,697 and a Warsh giveback. Adding a second demand engine does not rewrite that year. It can thicken the floor under the next dip if the four-week flow average keeps rising after payrolls. It can also mark a sucker week if outflows resume as soon as the 10-year holds 4.8 percent.
BofA’s own takeaway, in the flow coverage, was whether the surge becomes persistent rather than episodic. That is the only test that should change an outlook. Until the next month of ETF prints is in, the bull-market architecture is the same: official buying, fiscal deficits, a hawkish Fed, and a metal that has to defend $4,300 this week. The shift is real. The change in outlook is conditional. Conditional is not the same as cancelled, and it is not the same as confirmed.
On BofA’s September flow read, private fund and ETF demand is back at the strongest pace since October 2025, including a $6.4 billion weekly rise in gold-backed ETF holdings in August. The prior two years were led by central banks. The new question is whether both engines run at once.
It means a $4,360 2026 average that the tape is already near, a $5,000 mark after tightening ends, a $6,000 long-horizon fiscal case that missed its spring date, and a flow signal that is constructive only if it lasts. It does not mean buy gold this afternoon.
Yes, if flows persist and gold holds support. No, if the ETF week fades and $4,300 fails. Equities will move more than the metal in both directions.
This article is for informational and educational purposes only. It is not investment advice, a solicitation, or a recommendation to buy or sell gold, gold mining stocks, ETFs, or any other security or commodity. Bank of America price targets, average forecasts, flow figures and strategy comments are opinions published by that firm or reported by secondary sources and can change. Other bank forecasts cited are likewise opinions. Prices, yields and calendar dates are as reported around September 1, 2026, and change continuously. Forecasts can be wrong. Gold and mining equities are volatile and can result in the loss of principal. Past performance is not indicative of future results. Readers should consult a qualified adviser and conduct their own due diligence.
Author
Ben McGregor authors the Weekly Roundup at CanadianMiningReport.com, providing sharp analysis of the metals and mining sector. With a talent for spotting trends, Ben distills complex market shifts into clear, engaging insights on TSXV junior miners. His weekly updates cover gold, copper, uranium, and more, blending data-driven perspectives with a knack for identifying opportunities. A vital resource for investors, Ben’s work navigates the dynamic junior mining landscape with precision.