Can Bubbles Be a Good Thing?

When people think of bubbles, we think of high and rising asset prices that far exceed the fundamental value of the underlying assets. 

The expectation, or at least the risk, is that at some indeterminate point in the future prices will reverse and cause large losses. 

But can bubbles actually be a good thing? 

Key Takeaways – Can Bubbles Be a Good Thing?

• Not all bubbles are equal: 
  • “Mean-reversion” bubbles mostly destroy wealth and then vanish. They enrich early participants who time their exits well, but are mostly pointless on net.
  • “Inflection” bubbles build new technology and infrastructure and can lead to long-term benefits, but still crush many investors.
• Inflection bubbles around railroads, electricity, internet, or AI may lead to an initial surge and subsequent losses. 
  • But they can massively accelerate progress by funding experiments and building infrastructure with long-term value that would never pass normal ROI screens or value-driven assessment criteria.
• After the bust, cheap leftover assets (e.g., fiber, data centers, rail lines) become the foundation for long-run growth and productivity.
• Socially, and via other knock-on economic effects, these bubbles can be net positive even if huge amounts of capital are misallocated or written off in the investments themselves.
  • Cars and airplanes have been poor investments directly despite their influence due to their high capital intensity. Their benefits have been clear.
• For traders/investors, bubbles are usually harmful.
  • Newness removes valuation anchors, lottery-ticket thinking dominates, memories of past crashes fade, and leverage amplifies mistakes.
• Equity can justify moonshot risk because of the upside of hitting on one or a few big winners. Debt usually can’t in high failure environments because the coupons of those with lasting success don’t compensate for all the failures.
• Practical approach: don’t go all-in or all-out. Take moderate, diversified exposure, avoid heavy leverage, and focus on behaving prudently if it is a bubble.

What is a bubble?

Bubble is similar to other financial terms like recession and bear market, where if you don’t have a formal definition it gets tossed around to the point where it loses its meaning.

There’s not an agreed-upon “bubble criteria” but here is our definition:

1) Prices are high relative to traditional measures.

For example, if stocks have a low forward yield, a rise in inflation or normalization in real rates could compress their multiples and hurt the most leveraged companies.

2) Prices are discounting high levels of earnings that aren’t likely to be justified.

This one is not necessarily obvious in the moment. Bulls often argue that the asset is growing a lot and the valuation is merited.

If a stock has a forward P/E ratio of 100x, that means if it would take an investor 100 years to get their money back if the company achieves that one-year-forward earnings and then doesn’t grow further.

But companies with those types of earnings multiples have a lot of expected growth. If it’s expected to grow its earnings to 10x that amount in 5-10 years, it could certainly justify the valuation.

3) Purchases of assets are commonly done with high leverage.

Purchases on high leverage are common to lever up to be in on the trend.

4) Bullish sentiment is broad

For example, there are few bearish investors and survey data is widely positive.

5) A wave of new buyers is entering the market.

There’s a wave of new buyers in the market who are inexperienced, speculating on assets they don’t understand well, don’t know how to price, are driven by narratives and groupthink rather than a careful analysis of future earnings potential, and riding the wave on the most popular assets.

6) Monetary policy is stimulative and pushing asset prices higher.

Monetary policy is often stimulative by keeping rates sufficiently low (i.e., low real rates) and/or directly monetizing debts.

When cash and bonds have poor or mediocre real returns, there’s more incentive to take risk.

7) Extended forward purchases are being made.

Buyers are making extended purchases of goods such as supplies and inventory to either speculate or help insulate themselves from future price gains.

Two kinds of bubbles

You can think of bubbles in roughly two categories:

1. “Mean-reversion” bubbles

These are driven by some financial fad or supposed “new thing” that promises high returns without much risk. 

Examples include things like exotic mortgage securities before 2008 or various “can’t lose” schemes in earlier eras.

They usually:

• Don’t fundamentally change how the world works
• Attract capital purely because they look like a money machine
• Inflate on the belief that profits are easy and almost guaranteed
• Attract a lot of new and unsophisticated buyers
• Collapse back to earth once reality shows the cash flows were never there

These bubbles mostly destroy wealth and then fade. 

When they burst, the world looks a lot like it did before, just with more bankruptcies and regret. As well as plenty of misallocated capital and lost opportunities elsewhere.

2. “Inflection point” bubbles

These form around genuine technological revolutions: railroads, electricity, radio, the internet, and now artificial intelligence. 

In these episodes, investors correctly sense that “the future will be meaningfully different from the past,” but then wildly overshoot in how fast and how profitable that future will be. 

They usually:

• Fund huge amounts of new infrastructure that wouldn’t have been built under sober, slow, spreadsheet-driven capital allocation
• Pull forward decades of experimentation and iteration into a few frantic years
• Dramatically overbuild capacity, then slam into a painful correction
• Leave behind a new permanent foundation that society and future businesses can use at a far lower cost

These bubbles also destroy enormous amounts of market wealth. 

There are typically also plenty of “innovation like no other” schemes run by promotional management teams and con artists that become rich without actually producing much in profitable output. 

But the good parts of these bubbles often accelerate progress and permanently raise what the world is capable of.

From society’s perspective, that second kind can be extremely beneficial.

Why “good” bubbles help the world

In a normal, cautious environment, capital for risky new technologies trickles in. Managers insist on clear business cases, near-term cash flows, and demonstrable demand. 

That keeps investors safer, but it also means:

• Fewer experiments
• Less redundancy
• Slower build-out of infrastructure
• More gradual learning

In a boom, optimism becomes a coordinating force. People start to believe “this changes everything,” and two key things happen:

1) Massive parallel experimentation

Instead of ten careful projects, you get hundreds or thousands.

Different teams try different architectures, business models, and use cases.

Most of those attempts fail, but they map out the landscape far faster than cautious, incremental investment can.

2) Front-loaded infrastructure

Expensive platforms that no single prudent firm would fund alone get built anyway: national rail networks, fiber backbones, data centers, power lines, satellite constellations. 

At the time, many of these assets are mispriced, overbuilt, or uneconomic. 

After the crash, they often get bought cheaply and reused by the next wave of operators.

So from a social and technological point of view:

• The bubble compresses the “installation phase” of a technology.
• The crash redistributes ownership and writes down costs.
• The “deployment phase” then unfolds on top of cheap, already-built infrastructure.

That is why you can say:

• The railroad boom was a bubble and it transformed America.
• The dot-com boom was a bubble and it gave us the backbone of the modern internet.

The overinvestment was irrational for many shareholders, but highly productive in the long run.

Bubbles and parallels to government infrastructure investing

Let’s say a local government wants to build a subway system with an expected 50-year useful life. 

The original idea is they’ll spend $10 billion and this will eventually be paid back through train fares.

But let’s say the system, with cost overruns and higher fixed costs than anticipated, ends up being only half as economic.

So they have $5 billion in bad debt that can’t be paid, but if they write it down and spread out the cost over that 50-year period – i.e., $100 million per year, or 1% of its cost per year – it’s not as bad.

Taking a 1% write-down per year, they might very well believe they were better off having the subway system than being overly cautious and not having it. 

So even though it was uneconomic, other social, political, or economic goals may have provided enough benefits to justify it, even if the project itself couldn’t cover it.

Why bubbles are usually bad for investors

The fact that some bubbles are socially useful doesn’t make them friendly to capital.

A few recurring patterns hurt investors:

1) Newness removes anchors

With no long history to constrain imagination, people start projecting limitless futures, extrapolating overly optimistic growth rates, and paying prices that assume almost everything goes right. 

2) Lottery-ticket thinking

When the potential payoff looks like “the biggest company of all time,” even tiny probabilities are used to justify large bets, often with very optimistic assumptions. 

A $1 trillion company used to be a big milestone (Apple in August 2018, in terms of reaching and sustaining the valuation). Then in 2025, we saw the first $5 trillion valuation (Nvidia). Now $10 trillion valuations won’t seem farfetched.

So, even if a company has a small sliver of a chance of reaching such valuations, large upfront investments may seem worthwhile doing the expected value calculations.

For example, if a venture capitalist invests $10 million into a company, if the prospective ceiling of the company is a $1 trillion valuation, you only need a 0.001% probability of that happening to justify the investment. 

If the bets of the VC are diversified and the portfolio companies can find ways of doing business and genuinely improving the value of each other to boost their odds of success, this kind of investment becomes easy to rationalize based on expected value.

3) Short financial memory

Lessons from the last crash fade quickly, especially when a new technology feels different from anything seen before.

4) Leverage

Once debt enters the picture, losses are magnified, projects become more fragile because of the obligatory nature, and failures spread more easily through the financial system.

In a winner-takes-most technological race, equity is usually the right way to play the upside. One huge winner can offset many zeros. 

For debt, the math is the opposite. You earn a modest coupon on the winner and face permanent impairment on the losers, and that mix can be very unattractive if the technology’s economics are still speculative.

It’s why a lot of VCs have trouble wrapping their heads around the bond market.

So for the average investor, a bubble is usually:

• Great in hindsight
• Brutal in real time, especially if they arrive late, concentrate their bets, or add leverage
  

So, can bubbles be a good thing?

It depends on whose “good” you care about.

For society and long-run technological progress

Yes. Bubbles around genuinely transformative technologies can be great accelerants.

They mobilize capital and talent, build out infrastructure, and push experimentation and learning into overdrive.

The resulting crash is painful. It can be bad for a large number of projects. Even for sophisticated VCs, most of their investments are commonly zeroed (no return of capital).

But the world often emerges with new capabilities, cheaper platforms, and a higher baseline for productivity.

For individual investors caught up in them

Usually not. Most participants pay too much, misjudge timing, back the wrong players, or use too much leverage.

Many see large permanent losses.

A small minority who are early, selective, and disciplined may do very well, but that’s not the typical outcome.

For someone deciding how to act today

The practical takeaway is nuanced:

• You probably do not want to be all-in on the latest story, since you can’t know how far enthusiasm will overshoot or which companies will survive.
• You also probably do not want to be all-out of every major technological shift, since some of the greatest long-term returns come from riding those inflection points in a measured way.

At the same time, getting in on a technological shift may mean not even having to allocate toward it directly.

To get in on the internet, a sector tilt in the late 90s would have potentially wiped out 95%+.

Being in the NASDAQ – a broad tech index – lost 80% peak to trough. 

The S&P 500, still heavily diversified, lost 40%+ with no specific tilts.

The airplane and car were some of the most influential inventions of the early 20th century, but provided very little value to shareholders of airlines and automakers.

A sensible stance is to treat a potential bubble as both opportunity and minefield:

• Accept that some overinvestment is almost inevitable in big technological waves
• Participate moderately and diversify across players and timing
• Avoid heavy leverage tied to highly uncertain cash flows
• Focus less on “is this a bubble?” and more on “is my behavior prudent if it is?”

Of the thousands of car companies that were founded in the 20th century, only a few lasted and even those didn’t turn into great companies as far as returns.

In that sense, bubbles around real technological change can be “good” for the world, even as they are “bad” for many portfolios. 

The trick is to benefit from the progress without volunteering as one of the casualties that pays for it.

George Soros on the positive aspects of bubbles

As his popular quote on bubbles goes:

“When I see a bubble forming, I rush in to buy, adding fuel to the fire. That is not irrational.” 

George Soros integrated bubbles as part of his investment philosophy of reflexivity, his idea that market sentiment and economic fundamentals have a two-way, self-reinforcing connection. 

Soros believed that bubbles start with a solid basis in reality, or the idea is at least conceptually solid.

As we discussed earlier, bubbles can easily form when expected value calculations become easy to justify large valuations.

But at the same time, this reality becomes distorted by misconceptions or unrealistic extrapolations.

This, in turn, can lead to a boom-bust type of dynamic. 

Key aspects of his approach to bubbles:

Acknowledging imperfection

Needless to say, Soros is far from an indexer. He believes there’s lots of money to be made from the markets misunderstanding things. 

Bet big when right

His strategy isn’t about always being right, but rather maximizing gains when he’s correct and minimizing losses quickly when wrong. Slugging percentage, not batting average. Essentially options-based type of payoffs.

As his associate Stanley Druckenmiller noted, “when you’re right on something, you can’t own enough.”

Of course, this is terrible advice for non-professional and most professional traders, as trades without access to specific information or analysis are very much two-way. 

Profit from the rise and burst

Rather than avoiding bubbles, Soros tries to profit from their formation and subsequent burst by correctly timing his entry and exit. 

He watches shifts in expectations rather than valuations, identifying self-reinforcing narratives and feedback loops beget more buying.

His preference is to enter when the crowd becomes increasingly confident and exit when there’s evidence that marginal buyers are slowing and that confidence turns fragile and unstable. 

For example, he labeled gold the “ultimate asset bubble” in 2010 but was actively buying it, only to sell most of his holdings for a nice profit before the price dropped. 

What are some ways to identify fading bubbles?

Some examples:

Momentum deceleration

This is where price advances weaken, shorter-term moving averages flatten, and intraday rallies start to fade quickly.

Here, it becomes more evident that the easy money is at least taking a break, if not over.

Liquidity thinning

Here, you start to see bid-ask spreads widen.

Volume drops are common on up-days, which means there’s simply less selling rather than demand.

And spikes on down-days become increasingly common.

Positioning extremes

Leverage growth stalls.

Options skew goes hyper-offensive, then shifts (calls increase over puts at first, then shift once prices struggle to keep climbing).

Retail inflows peak, then reverse.

Narrative fatigue is common

Bullish catalysts stop generating upside. Good news is sold off instead of continuing to generate new price gains.

Insider behavior is a tell

Insider selling rises while corporate buybacks slow. This shows that those with the most information become more skeptical.

Cross-asset warnings

Credit spreads widen or funding costs rise, which signals stress under the surface.