When Investors Can't Agree: What Extreme Bull/Bear Polarisation Often Signals

Last reviewed: July 2026

Your X feed right now is a warzone. One side screaming "greatest bull market ever." The other insisting the crash is imminent. And almost nobody calmly occupying the middle ground.

None of this is new. That kind of sharp split in investor opinion is actually one of the more reliable features of late-cycle markets. Research going back seven decades shows that when disagreement gets this extreme, you're often closer to a turning point than most people realise. Not tomorrow necessarily. But the conditions are there.

This piece walks through the evidence, the key charts, and what it actually means for your portfolio. No crystal ball predictions. Just the pattern recognition that matters.

What does the classic sentiment cycle look like?

Sir John Templeton nailed it decades ago: bull markets are born on pessimism, grow on scepticism, mature on optimism, and die on euphoria. Simple enough. And broadly true.

At the start of a bull run, almost everyone is gloomy. Prices climb. Scepticism slowly gives way to optimism, and near the end many investors become convinced the good times will roll on forever. Bear markets follow the reverse arc.

But here's what's different about the current moment. We're not seeing one side dominate. We're seeing both extremes screaming at each other simultaneously. That's not consensus euphoria. It's polarisation. And historically, that's a different animal entirely.

Phase	Dominant Sentiment	Market Behaviour	Typical Duration
Birth	Deep pessimism, capitulation	Bottoming, forced selling exhausted	3-6 months
Growth	Scepticism, "dead cat bounce" fears	Steady climb, wall of worry	1-3 years
Maturity	Optimism, rising participation	Broadening rally, FOMO buying	1-2 years
Euphoria / Polarisation	Sharp disagreement, extreme views	Narrow leadership, high volatility	3-12 months

Why does extreme disagreement matter more than average sentiment?

Markets aren't driven by one "average" investor. They're driven by the tug-of-war between groups with fundamentally different perspectives. And when those groups disagree violently, interesting things happen.

Retail investors tend to extrapolate recent returns. When the S&P 500 (or the ASX 200) has been running hot, they pile in expecting more of the same. Professional and institutional investors are more likely to focus on valuations and mean reversion, which makes them cautious when prices stretch.

When these two camps pull in opposite directions for sustained periods, you get three things:

1. Higher trading volume and volatility as both sides express their convictions
2. Prices that stay elevated longer than fundamentals alone would justify, because the bearish "smart money" faces real limits to arbitrage (shorting is expensive, and being early is the same as being wrong)
3. Lower future returns on average. Academic research covering seven decades of US market data finds that stocks with high forecast dispersion tend to underperform those with low dispersion over subsequent 12-month periods

A market where 40% of participants are extremely bullish and 40% are extremely bearish is a very different beast from one where most people are mildly optimistic. The average sentiment reading might look similar, but the underlying dynamics are poles apart.

What does the historical evidence actually show?

The AAII Sentiment Survey

The American Association of Individual Investors has been surveying retail sentiment weekly since 1987. It's not perfect. But it's one of the longest-running, most-studied sentiment measures we have.

The bull-bear spread (bullish percentage minus bearish percentage) has repeatedly hit extremes near major turning points. And the pattern is consistent enough to pay attention to:

• Extreme bearish readings clustered around major bottoms: late 2008, March 2009, March 2020, October 2022
• Extreme bullish readings appeared near peaks or late-stage advances: early 2000, late 2007, early 2018, late 2024
• As of late July 2026, bullish sentiment sat at 29.6% while bearish hit 42.3%, producing a notably negative spread

AAII Sentiment Extreme	Date	S&P 500 Level	What Followed (12 months)
Peak bearishness	Mar 2009	676	+68% rally
Peak bearishness	Mar 2020	2,237	+75% rally
Peak bearishness	Oct 2022	3,577	+22% rally
Peak bullishness	Jan 2000	1,469	-12% decline (dot-com burst)
Peak bullishness	Oct 2007	1,565	-38% decline (GFC)
High polarisation	Jul 2026	5,600+	TBD

Here's the interesting part. Extreme consensus pessimism has been a far better historical buy signal than extreme consensus optimism has been a sell signal. Bearish extremes produced average 12-month forward returns north of 20%. Bullish extremes were messier, sometimes preceding immediate declines, sometimes just mediocre returns for a year or two.

Analyst forecast dispersion

When professional analysts disagree widely about future earnings for the same company, it has historically predicted higher risk of business-cycle turning points. During long expansions, some analysts cling to optimistic projections while others start adjusting downward. The gap widens before the consensus finally breaks one way.

Cross-sectional research from Diether, Malloy and Scherbina (published in the Journal of Finance, 2002) found that stocks with high analyst forecast dispersion underperformed those with low dispersion by around 9.5% annually. The market, it turns out, doesn't price disagreement correctly.

Market internals and breadth deterioration

This is where it gets practical for ASX investors. Even when the headline index looks strong, the breadth of the advance can quietly fall apart underneath. John Hussman and others have documented that widening internal divergences frequently appear near major peaks.

What does that look like? More stocks making new 52-week lows even as the index holds near highs. Sector leadership narrowing to a handful of names. In 2000 it was tech. In 2007 it was financials. In 2024-25, it was the AI-adjacent mega-caps.

On the ASX, we've seen something similar play out with the S&P/ASX 200 being dragged higher by a narrow group: Commonwealth Bank (ASX: CBA) trading at historically stretched multiples, the big iron ore names like BHP Group (ASX: BHP) and Rio Tinto (ASX: RIO), and a few tech darlings. Meanwhile plenty of small and mid-cap names have gone sideways or worse.

What does the current set-up look like for mid-2026?

We reckon the current environment has several hallmarks of late-cycle polarisation. Not a prediction. Just pattern recognition.

Here's what we're seeing:

• Institutional cash levels have been low, with some professional sentiment indicators running elevated through early 2026
• Retail surveys have swung cautious. The AAII bull-bear spread recently turned notably negative (29.6% bulls vs 42.3% bears as of late July)
• Consumer confidence remains weak relative to where stock prices are sitting, a disconnect that's widened over recent quarters
• AI stocks remain a flashpoint. Ask ten investors whether NVIDIA (NASDAQ: NVDA) or the ASX AI-adjacent plays are overvalued and you'll get ten passionate, contradictory answers

Think of it like a party that's gone on too long. Some guests are still dancing enthusiastically. Others are checking their watches and muttering about calling an Uber. Very few are calmly enjoying the music. And social media's algorithms are making sure you hear the loudest voices from both camps.

What should you actually do with this information?

We're not going to pretend we can call the exact top. Nobody can, and anyone who tells you they can is selling something. But there are some practical things worth doing when polarisation is this elevated.

First, treat polarisation as context, not a timing tool. High disagreement raises the odds you're in a late-cycle phase, but cycles can stretch longer than anyone expects. It's one input alongside valuations, breadth, and your own time horizon.

Second, watch the spread, not the average. A market split 40/40 with 20% neutral is qualitatively different from one that's 60% mildly bullish. The distribution matters more than the mean.

Third — and this one's easy to forget — social media is a distorted mirror. Algorithms reward emotion. The calm, evidence-based middle view gets almost no visibility. Don't let your X feed set your risk tolerance.

It also helps to understand the asymmetry here. Extreme consensus pessimism has been a much better buy signal historically than extreme consensus optimism has been a sell signal. Polarisation sits somewhere between those poles, which makes it useful but not decisive.

And finally, focus on process over prediction. Rather than trying to call the exact inflection point, ask yourself: "Given current valuations and this level of disagreement, am I being adequately compensated for the risks I'm carrying?"

Major market cycles rarely end with everyone holding the same view. They tend to end amid growing disagreement, some investors still extrapolating the recent trend while others are already preparing for mean reversion. The split you see on X right now is consistent with that pattern.

That's not a reason to panic. And it's not a reason to abandon a long-term plan. But it is a reason to double-check position sizes, review your diversification, and make sure your portfolio can handle a stretch of higher volatility without forcing bad decisions. We've been doing exactly that with our own positioning over the past few weeks.

Frequently Asked Questions