Did funds sell gold to buy the tech dip?

The week, briefly: gold slid from about $4,070 an ounce toward $3,890. Alphabet fell about 13% on Thursday after earnings; Intel fell about 29% on Friday; the Nasdaq-100 fund (QQQ) closed Friday down 3.47% on the day. The Federal Reserve held rates on Wednesday (10–2), strong GDP and hot inflation data landed Thursday, and by Friday December rate-hike odds sat near 72–75%. Five separate checks follow.

Check 01Did funds actually sell gold that week?

In fund-flow aggregates, modestly — Monday to Wednesday. In futures, no. And the large exchange-traded redemptions belonged to the week before.

• Fund flows (BofA "Flow Show," Oct 29; EPFR data, week to Wed Oct 28): gold funds saw a net outflow of $1.7 billion — described as gold's first net outflow since July — inside a week when equity funds overall took in $56.6 billion, dominated by China (+$25.7B) and Japan (+$6.8B).
• The big gold ETF (GLD), day by day (ETF.com, verified): GLD appeared in neither the top-10 creations nor the top-10 redemptions on Monday, Tuesday or Wednesday. There was no large single-day GLD liquidation in the crash week itself. The big GLD outflow — $2.30 billion — was the week ending Friday Oct 23. The World Gold Council's monitor agrees on timing: global gold ETFs lost $1,199 million (−7.7 tonnes) in the week to Oct 23, led by North America.
• Futures positioning (CFTC, released Fri Oct 30, positions as of Tue Oct 27): managed money's net long position in gold rose by 1,356 contracts, to 105,562 net long (longs 122,508; shorts 16,946). The week's futures liquidation came from the "Other Reportables" category (−14,770 contracts), not from managed money. One caveat: a Tuesday snapshot cannot capture Thursday–Friday selling.

Check 02Did funds buy the tech dip?

Aggregates say money entered tech funds Monday to Wednesday — before the big dips happened. The QQQ fund itself churned flat. Hedge funds were net sellers. The genuine dip-buying was retail, in chip funds, and mostly the week before.

• Sector flows (week to Wed Oct 28): technology funds took in +$5.3 billion (BofA/EPFR, a fifth straight weekly inflow) and +$5.5 billion globally / +$4.54 billion in the US (LSEG Lipper via Reuters). Reuters' stated driver was a rebound in AI and chip shares led by SK Hynix's record profit — buying the earlier chip rebound, ahead of the big earnings. US equity funds overall lost $6.65 billion.
• Hedge funds (Goldman Sachs prime brokerage): roughly $20 billion of net tech selling over the two weeks to Oct 26, US equities net-sold three straight weeks, tech being cut at the fastest pace since 2016. Goldman's desk recommended selective dip-buying on Oct 26 — a recommendation with a validation window of Oct 28–Nov 6, not observed buying.
• Retail: retail investors bought $2.6 billion of semiconductors in the mid-October week hedge funds sold (a 96th-percentile reading), with record inflows into leveraged equity ETFs (JPMorgan) — real dip-buying, but concentrated earlier in October and in chips.

Check 03The link: any direct evidence of a rotation?

None found. Across Reuters, Bloomberg (via syndication), Kitco, and strategist coverage from Oct 26 to Nov 1, there is not one claim — from a named analyst, a bank note, or a flow provider — that investors sold gold to fund technology purchases that week. The documented rotations were out of AI/tech and into Treasuries (Bloomberg, Oct 30: the rout "accelerated a rotation out of the artificial intelligence trade and into safe havens such as Treasuries") and out of risk assets into cash. Where gold and tech appear in the same sentence, the mechanism is uniform: gold sold for cash to cover tech losses and meet margin calls — Commerzbank's reading via Bloomberg, and a Kitco analyst's summary on Friday: "when the tech selloff hit, people sold gold to cover margin calls elsewhere. Gold is liquid, so it is the first thing that gets sold when people need cash." That is the opposite direction of causation from the story being checked.

One methodological caution, because both numbers get spliced together in commentary: BofA called the $1.7 billion gold outflow the "first since July," while the World Gold Council recorded a gold-ETF outflow the previous week too. Different fund universes, different methods — the two series should not be stitched into one.

Check 04Timing and scale

Scale fails too. Even in the friendly Monday–Wednesday window, the gold outflow ($1.7B) is about a third of the tech inflow ($5.3B), and both are rounding errors beside the week's $56.6 billion of total equity inflows. In the prior week the mismatch is starker: GLD lost $2.3 billion while chip ETFs took in roughly $11 billion plus $2.7 billion into QQQ — bought mainly through leveraged retail products while hedge funds were net sellers of about $20 billion of tech over two weeks. The money does not line up on either side of the ledger.

Check 05So why did gold fall? The stronger explanation

• A hawkish Fed repricing. The Fed held on Oct 28 with two dissents; Thursday brought stronger-than-expected Q3 GDP and accelerated September inflation data; oil rose on the Hormuz tanker attack. December hike odds reached ~72–75%. A non-yielding asset loses that contest against ~5% Treasury yields (Reuters/UBS, Oct 30).
• Margin calls and liquidity, Thursday–Friday. Gold fell as much as 3.2% on Friday to its lowest in a year "as traders sold the precious metal to cover losses elsewhere amid a global tech-stock rout" (Bloomberg). On Friday itself, yields fell while gold dropped 3.6% — the signature of forced, month-end liquidity selling, not a fresh rates shock.
• Momentum and profit-taking. Gold broke below $4,000 after a run to a January record; trend-followers were testing long-term support in both metals (TD Securities), and managed money had already been trimming for a month before the week began.
• Money went to Treasuries. Bonds rallied to their best month since December 2023 — the defensive bid is documented, and it did not go into tech.

Where the story probably came from: the week ending Friday Oct 23 genuinely does pair chip-ETF dip-buying with a large GLD outflow in the same table. Retell that week as this week, drop the "who" (the buyers were largely retail and leveraged funds; hedge funds were selling), and the narrative writes itself. It is still wrong for Oct 26–30.