print-icon
print-icon
Add ZeroHedge as a preferred source on Google

Worst In 20 Years: Hedge Fund VIPs Crushed As Goldman Warns The Cyclicals They Fled To Are Next

Tyler Durden's Photo
by Tyler Durden
Authored...

Hedge funds spent the third quarter doing exactly what they were supposed to do: trimming AI, adding everything else, and cutting risk.

It didn't help much.

According to Goldman's latest quarterly Hedge Fund Trend Monitor, published late Friday by the bank's chief US equity strategist Ben Snider (available to pro subs), which picks apart the 13F holdings of 1,001 hedge funds with $5.1 trillion in gross equity positions ($3.3 trillion long, $1.9 trillion short), the 50 stocks that matter most to hedge funds have just gone through one of the worst stretches in their two-decade history:

The most popular hedge fund long positions (GSTHHVIP) have been tightly correlated with the AI trade... The VIP basket's 14 pp underperformance vs. the S&P 500 during the past six months matches its most significant underperformance vs. the market in its 20-year history. Elevated portfolio concentration and crowding have increased the challenge facing fund managers.

In other words, the smart money's favorite stocks have now lagged the market about as badly as they did in the depths of the financial crisis.

And here's the kicker. As Snider documents, funds have been "broadening" out of AI and into cyclicals and defensives. In his Friday evening Global Reflections note (available to pro subs), Goldman's Louis Miller, whose GSX notes cover the bank's thematic baskets, argues that the cyclical pocket many of them rotated into is precisely what gets taken out back if the "4Q'18 analog" everyone is suddenly talking about keeps playing out.

One Of The Worst Six Months In 20 Years

Start with the scorecard. Based on GS Prime Services estimates, US equity long/short funds are up 5% YTD through Nov 19, slightly behind the S&P (+6%) and well behind the equal-weight S&P (+12%). Goldman's VIP basket of the most popular hedge fund longs, however, is down 3%, trailing even 7-10 year Treasuries and a basket of the most-shorted stocks (+1%). Diversification at its finest.

Over six months the damage is far worse. The VIP basket has lagged the S&P by 14pp, roughly level with Dec '08 (-13pp) and May '22 (-15pp), and quite a reversal from June, when it was beating the market by 10pp.

Why so bad? Because for all the talk of stock picking, the hedge fund long book is one big AI trade: the VIP list's excess returns have a 0.9 correlation with Goldman's broad AI, semis, optical networks, neoclouds and memory baskets.

Then there's crowding. The median hedge fund holds 70% of its long book in its top 10 positions, and Goldman's Hedge Fund Crowding Index rose to a new record high in Q3. When everyone owns the same names, everyone sells the same names at the same time... as the past six weeks have shown.

Bruised: Nets At "Liberation Day" Lows... Gross Not So Much

Faced with that, funds did the obvious thing and cut. Per Snider, fundamental long/short net leverage has plunged to 42%, the 0th percentile of the past five years and the lowest since Liberation Day. Gross leverage has fallen too, to 205%, but it is still in the 74th percentile over five years. 

Across the overall prime book, gross sits at a towering 316% (98th percentile).

Miller's desk data, run on a beta-adjusted basis, tells the same story with fresher numbers: fundamental L/S net leverage fell 9 points this week alone and is down more than 22 points since the start of October, to 48.2%, a five-year low. That is quite a round trip from June, when we wrote that hedge fund leverage had hit 4-year highs after one of the sharpest buying periods in history (chart source GS Prime Brokerage).

We flagged the slide earlier this week, along with the chart that worries us most:

Like gross leverage, short interest across the S&P has come down in recent months but remains elevated: the median S&P 500 stock ranks in the 92nd percentile of the past 20 years, with Utilities (99th percentile), Staples and Health Care (both 95th) the most shorted relative to history.

...And Broadening: Out Of AI, Into Everything Else

The more interesting part of Snider's report is where the money went. In Q3, funds cut TMT and added to cyclicals and defensives alike. Relative to the Russell 3000, net tilts rose most in Financials (+254bp), Consumer Discretionary (+237bp) and Industrials (+190bp), while funds also added to Staples (+154bp) and Utilities (+120bp). The Info Tech "underweight" grew by another 619bp to a record -2,267bp (much of that reflects tech's huge 35% weight in the index).

The result: hedge funds entered Q4 with 10-year-high tilts to Health Care, Financials, Consumer Discretionary, Utilities and Consumer Staples, all at the same time. Translation: funds spread the crowding around.

The factor picture confirms it. Funds rotated from Growth to Value, and exposure to Momentum fell to the lowest level since the AI trade began in early 2023.

None of this will surprise regular readers: earlier this month we reported that hedge funds had gone underweight AI for the first time since 2023, as Morgan Stanley's prime book logged the largest sustained de-grossing on record:

Within AI itself, funds rotated. Intel (INTC) and Meta (META) were among Q3's "Rising stars" with the biggest increases in hedge fund popularity (Intel picked up 55 new owners, with 89 more funds adding than cutting), and Nvidia (NVDA) saw small net adds. Alphabet (GOOGL), Qualcomm (QCOM) and AMD were "Falling stars." Most of the rising stars came from outside AI, though: Allstate (ALL), Home Depot (HD), IQVIA (IQV) and Merck (MRK). New non-TMT names joining the VIP list include AerCap (AER), Danaher (DHR), Somnigroup (SGI), Toast (TOST), UnitedHealth (UNH) and Union Pacific (UNP).

The 4Q'18 Problem: Where Hedge Funds Fled Is Where It Hurts Next

This is where Miller comes in. The 2018 comparison, which we have been tracking all week (from The Market Ear's "Great Tech Liquidation: Is 2018 About To Repeat?" to Brian Garrett's "Absolutely Zero Idea" rerun, which pencils in S&P 6,294 if the 2018 path plays out), has now officially gone mainstream. Miller's take on what that means for positioning:

4Q'2018 Déjà Vu. Beta adjusted nets are at 5 year lows and High-Beta Cyclicals (GSXUHBCY) are still up +12% YTD while broad US AI (GSTMTAIP) is now negative on the year. Part of the resilience of the Cyclicals trade can be explained by the fact that growth optimism has remained high and the market has traded rotations out of AI into ex-AI, rather than owning outright equity downside. Back in 4Q-2018, High Beta Cyclicals were amongst the weakest pockets of US equities, declining -23% and underperforming SPX by 10pp. If the same backdrop does manifest, and growth uncertainty takes over from inflation anxiety, we'd expect this pocket to underperform sharply.

Here is the divergence in one chart: over the past year, high beta cyclicals are up roughly 31% while Goldman's broad AI basket has given back nearly all of its gains, now up just 4% (chart source Goldman, Bloomberg).

And valuation offers no cushion. Miller notes the high beta cyclicals basket trades at 15x forward earnings, a 40% premium to its 2018 trough multiple. Two more 2018 laggards look exposed:

  • Lending Sensitive (GSXULEND) is flat QTD, against a 15% plunge in 4Q18, and "remains susceptible to increased Fed hawkishness and tightening in the lending market." With the Fed still hiking into hot CPI prints, that seems like a reasonable worry.
  • European consumers: UK Consumption (GSXEUKCO) and EU Domestic Consumption (GSXEDCON) are both up 12% QTD; in 4Q18 they fell 28% and 16%.

Put the two notes together and you get the uncomfortable punchline: hedge funds just pushed their Financials and Consumer Discretionary tilts to 10-year highs (and added 190bp to Industrials), piling into the same cyclical corners of the market Miller wants to short. Good timing.

The Market Ear framed the stakes nicely this morning:

"Fast Money Is Light And Slow Money Is Full"

So if hedge funds have already de-risked, who is left to sell? Miller's answer is everyone else:

Taking stock of our latest positioning stats – one thing stands out: fast money is light and slow money is full. CTAs, vol control and hedge fund net exposure are all near the low-end of their range... The full end is US equity ETF inflows, record-low mutual fund cash and PB gross leverage, which is heavy because books are running gross while keeping net low. Additionally, worth flagging growth mutual fund managers beta is somewhat high versus history (90th percentile).

Here is that last point: the beta of US large-cap growth mutual funds vs the S&P sits at 1.38 on a 3-month basis, not far from the April 2022 high of 1.58 (chart source GS).

Put differently, the hedge funds that usually get blamed for liquidations have already liquidated. The mutual funds sitting on record-low cash and high beta haven't. Miller's summary line says it all: "Beta adjusted nets are at 5 year lows, but real money positioning is not low." This matches what Goldman's traders told us earlier this month: plenty of exposure, zero conviction.

Where To Hide: FCF, Healthcare... And A Tail Hedge 2.5% Away

On the long side, Miller sticks with Goldman's High vs Low FCF pair (GSPUFCFP in the US, GSPEFCFP in Europe), which the desk flagged last week as its preferred long exposure. Among defensives, he is selective:

We are sympathetic with the long Defensives pitch, but we want to be selective in that exposure and have a preference for Healthcare versus Staples.

The desk likes structural growth themes with low correlation to oil, rates and AI (ACA beneficiaries, commercial biotech, bioprocessing, tools) and bundled them into a new 2027 Top US Healthcare Themes basket (GSCBHC27), where upward earnings revisions through 2028E take the multiple from 28x this year's earnings to as low as 19x 2028's. Just one caveat: per Snider, Health Care is already hedge funds' single biggest net overweight at +960bp vs the Russell 3000, so this particular hiding place is hardly empty.

Elsewhere, Miller likes China High Dividends (GSXACHHN), up 3.4% since last week, and would fade any relief rally in the China Competition basket (GSXECHNX): cheap at 11x forward earnings, but only if you trust consensus estimates. Classic value trap.

And for those who think this ends with a 2018-style capitulation, the desk is pitching its Conditional Downvar Strategy (GSISCVS1), which is designed to provide deep OTM protection with flat carry in normal markets. Notably, with <5% of basket flows on Friday being buy-to-open, Miller says "no offense [is] being played." His explanation of the trigger:

After today's 3.7% down move, we are inside the protection corridor. If the selloff accelerates next week, the strategy's long puts will kick in and we can expect a convex payoff. A reversal next week would take us out of the corridor (which is how 2018 played out), potentially locking in losses on the delta hedge.

Here is the S&P against the strategy's barrier, which stood about 2.5% below spot as of Friday morning (chart source GS FICC & Equities):

The strategy shines in "sudden, deep selloffs" (2008, 2020) and struggles in slow grinds like 2022, as its 20-year (mostly backtested) record shows: flat for years, then two giant steps (chart source GS FICC & Equities).

Bruised, Broadened... And Still Crowded

Snider's data says hedge funds have done the hard part: nets at five-year lows, AI cut, Momentum purged. Miller's data says that's exactly the problem: what risk the pros still hold sits in record-crowded cyclical and defensive tilts, and the real money that never sold is "full."

The bull case is that washed-out fast money means any spark sends the shorts scrambling (Goldman's Privorotsky smelled a squeeze on Anthropic day). Perhaps. But the 2018 playbook didn't end with a squeeze. It ended with the cyclicals finally cracking, credit waking up, and a Fed capitulation that, this time around, nobody is pricing. Until one of those happens, we'd bet the next leg of this selloff hits the "safe" rotation trades hedge funds just crowded into... which is probably why Goldman's desk is pitching a tail hedge with spot 2.5% from the barrier.

The pros dumped the AI trade for cyclicals. History says they may have just traded one crowded trade for another.

Much more in the full Goldman Hedge Fund Trend Monitor and Miller's "Global Reflections" note, both available to pro subs.

0