Fund Managers Are Nearly as Bullish as Before the 2022 Bear Market — and Their Cash Is Nearly Gone

Fund Managers Are Nearly as Bullish as Before the 2022 Bear Market — and Their Cash Is Nearly Gone

Bank of America's August Global Fund Manager Survey is the third most bullish reading of professional investor sentiment since 2022. The number that matters more than the mood is the cash: roughly 200 institutional managers with more than half a trillion dollars in assets now hold just 3.5% of their portfolios in cash. It is the sixth-lowest level in the survey's 28-year history.

This is one of the best public windows a retail investor has into what the professionals are actually doing, because the survey measures positioning, not opinion — where the money sits, not what people say about it. The August answer: managers are a net 56% overweight equities (more exposed than their benchmarks), the most since November 2021, and they have been overweight stocks for 14 straight months.

Why the cash number matters

Cash is dry powder — the reserve a manager draws on to buy weakness. Decades of this survey show it has been one of the market's most consistent inverse indicators: when professionals are flush with cash, forward returns tend to be better; when they have spent it, there is less natural buying underneath the market. At 3.5%, the buying power is deployed.

Bank of America's own rules recognize the extremity. Its Global FMS Cash Rule is flashing "sell," a signal that fires whenever average cash falls to 4% or below. Its Bull/Bear composite, which weighs positioning across asset classes, sits at 9.3, above the 8.0 level the bank reads as a contrarian sell. The house recommendation to clients is to retreat or rotate within risk assets rather than reload.

The recent record is vivid

This is not a hypothetical warning, because we have seen this exact setup twice in the last nine months. Bank of America ranks only two surveys since 2022 above August's. December 2025: cash hit its all-time low of 3.3%, and managers were the most bullish they had been in more than three years. February 2026: positioning was the most bullish in five years, with the Bull/Bear gauge at 9.5.

What followed? Stocks printed record highs in early January 2026, then the first quarter delivered the market's worst quarter since mid-2022 — down roughly 4.6% — with the Nasdaq falling into its first correction of the year. Fair caveat: the December signal was about a month early. It was not a clock. These extreme readings are better read as late-cycle comfort than as a calendar.

Last month provided the cleaner demonstration. The July survey set an all-time record — 82% of managers named long semiconductors the most crowded trade. Within about two weeks, chip stocks shed roughly a trillion dollars of market value and the Nasdaq-100 slid into its second correction of 2026. Twice in the last nine months, an extreme reading has been followed by a flush.

The crowd fears its own biggest bet

Now look at August's list of fears, and the odd part becomes hard to miss: the crowd's biggest worry is the thing it owns most. The leading tail risk is an AI bubble, named by 32% of managers. Hyperscaler AI spending is what 38% call the most likely source of a systemic credit event. Yet semiconductors are still the most crowded trade on the board, and investors rotated into tech in August rather than out of it.

Bank of America has compressed the whole consensus into a single line: "no macro landing, no Fed hike, no AI capex cut, no bears." That is a consensus with no bear case left in the positioning — a crowd quietly worried about its own biggest bet and expressing that worry nowhere at all.

One more honest wrinkle: this is not retail euphoria. Weekly AAII data put individual investors at about 35% bullish against a long-run average near 38%. The froth is concentrated among institutions and inside the AI megacaps, which narrows the risk: this is a crowded trade, not yet a broad mania.

The bulls deserve their full case

Contrarian thinking is a conclusion, not a starting position, so let the optimists make their argument. A record 56% of managers expect no economic landing at all; 43% expect a boom, the most since early 2022; and the share expecting double-digit earnings growth is the highest since August 2021. The capex machine that carries the whole trade is, so far, intact: Reuters estimates hyperscaler spending will rise by roughly $534 billion, alongside about $340 billion more in annual operating cash flow forecast for 2027 than 2025. If those numbers keep landing, the crowd is early, not wrong — and the burden of proof sits with anyone calling a top.

But the cracks are in the same document. A "disorderly rise in bond yields" jumped to the second-biggest tail risk, named by 27% of managers versus 14% a month ago, and rising yields hit the long-duration valuations of exactly the AI names the crowd owns. The 30-year Treasury already sits at its highest yield since 2007. There is also a live disagreement parked in plain view: most managers say the Fed will not hike before the November midterms, while money markets now price roughly 96% odds of a quarter-point rate hike this year. Someone is wrong, and the consensus has usually been the wrong side on rates. The same crowd, after all, calls corporate balance sheets the most over-leveraged since March 2023 — and still raised its equity bet.

The test lands this week

The next two sessions make the question concrete. Nvidia, the stock at the center of the semiconductor trade, reports after Wednesday's close, August 26; analysts expect roughly $91 billion in quarterly revenue, and options markets are pricing an 8–12% move in either direction. Nvidia is already down about 4.6% over the past week, while the S&P 500 sits within about 2% of a record high it reached just weeks ago. Fed Chair Kevin Warsh's Jackson Hole keynote follows Friday, August 28.

My invalidation condition is the fundamentals, not the survey. A clean Nvidia beat with raised guidance and calm yields confirms the bulls' numbers; the right response then is to hold quality and refuse to pay up at these levels, not to fade the market because a survey is bullish. The crack I am watching for runs the other way: any hint of hyperscaler capex conservatism, a wobble in guidance, or yields breaking decisively higher. In that world, the crowd has no plan B — and, at 3.5% cash, no buying power left to buy its own rescue.

What the survey is actually for

Not a crash call — treat anyone selling it to you as one with suspicion. It is a risk-regime reading. When the professionals are unanimous and their reserves are spent, the easy part of the advance is arguably behind and the market's reaction to bad news gets sharper. The useful habits are the unglamorous ones: keep some of your own cash as dry powder, plan to buy real pullbacks instead of chasing record highs the institutions already bought, and own businesses that survive a crowded trade unwinding — because the crowd demonstrated in July exactly what that unwinding looks like.

Arguing against unanimity only pays when the fundamentals support you. Reading unanimity as a reason to be careful with your entry points pays regardless.