The Market Has Stopped Pricing Reality
When capital is too cheap, speculation stops looking dangerous.
In the late stages of every great boom, reality starts bending in small ways first.
In 2007, Las Vegas taxi drivers were giving condominium investment advice between airport runs. In 1929, shoeshine boys were discussing stock tips outside Wall Street offices. During the railway mania of the 1800s, entire towns appeared on maps before a single resident had arrived — financed purely on the assumption that prosperity would inevitably follow steel tracks into the wilderness.
At the peak, none of it feels irrational.
That’s the dangerous part.
Speculation rarely feels like speculation while liquidity is abundant. It feels like inevitability. It feels modern. Intelligent. Even responsible.
Until suddenly the system stops asking whether something is profitable… and only asks whether enough capital still exists to keep the story going.
That’s where markets appear to be heading now.
The AI boom is being framed as a technological revolution.
But structurally, what we’re witnessing looks far more like a global credit event disguised as innovation.
And history suggests those two things rarely end the same way.
Cheap Capital Changes Behaviour Everywhere
Something unusual has happened over the last eighteen months.
Markets have stopped valuing companies primarily on realised cash flow and begun valuing them on projected ecosystem dominance inside an assumed future economic architecture.
That sounds abstract.
But it has very concrete consequences.
It means:
debt becomes easier to justify
duration assumptions stretch
depreciation schedules become flexible
profitability becomes narrative-dependent
and capital expenditure stops behaving cyclically
The current AI infrastructure race demonstrates this perfectly.
Technology firms are now committing hundreds of billions toward compute expansion despite the underlying economics remaining deeply uncertain.
Not because the returns are visible.
Because the market currently rewards scale more than discipline.
That changes behaviour across the entire system.
Server farms are financed before demand fully exists.
GPU hardware is funded using assumptions extending years beyond realistic technological obsolescence.
Private companies with minimal profitability attract sovereign-scale capital inflows.
And revenue increasingly circulates within tightly interconnected financing loops.
In several cases, the same firms providing infrastructure are also indirectly funding the customers purchasing it.
That’s not traditional demand expansion.
That’s reflexive liquidity.
And reflexive liquidity always looks strongest near the end.
This Is What Mispriced Money Looks Like
Most investors still think bubbles are valuation events.
Historically, they are usually credit events first.
Valuations are just the visible symptom.
The deeper issue is that when capital remains too cheap for too long, financial systems lose their ability to distinguish between productive investment and speculative acceleration.
Everything begins to clear the hurdle rate.
Projects that would normally be rejected suddenly appear viable.
Leverage stops looking dangerous because refinancing appears permanently available.
That environment creates an illusion:
that growth itself eliminates risk.
It doesn’t.
It usually concentrates it.
And the concentration becomes most visible when liquidity conditions begin tightening while speculative positioning remains extended.
That’s increasingly what today’s market structure resembles.
Not the beginning of a clean expansion cycle.
But the late-stage phase of a liquidity-saturated one.
Why The AI Narrative Matters More Than The AI Technology
The danger here is not that artificial intelligence fails technologically.
In many ways, AI will almost certainly transform industries over the next decade.
The real issue is the speed at which financial markets are trying to monetise that future today.
That distinction is critical.
Historically, revolutionary technologies create two separate cycles:
the innovation cycle
the speculation cycle
The innovation cycle can last decades.
The speculation cycle usually burns far faster.
Right now, markets are compressing years of assumed adoption, productivity gains, and future profitability into present-day valuations. Entire business models are being priced as though demand certainty already exists.
Meanwhile, the underlying economics remain highly unstable.
Hardware becomes obsolete rapidly.
Energy requirements continue rising.
Competition intensifies monthly.
And many AI firms remain dependent on external financing rather than internally sustainable cash generation.
The technology may absolutely survive and reshape the economy.
But that does not guarantee current pricing survives with it.
History Repeats The Same Financial Script
This isn’t new.
History has seen this sequence repeatedly.
Different technologies.
Different decades.
Same financial architecture.
The railways transformed commerce.
Radio transformed communication.
The internet transformed information.
Each innovation changed the world permanently.
But in every case, speculative capital dramatically overestimated the pace, profitability, and durability of the initial expansion phase.
That’s because liquidity changes human behaviour.
Once capital becomes widely available, markets stop financing reality and begin financing possibility.
The sequence is remarkably consistent:
A transformative technology emerges
Cheap capital accelerates deployment
Competition floods the space
Infrastructure expands faster than demand
Leverage accumulates invisibly
Credit conditions tighten
Valuations collapse toward economic reality
Importantly, the underlying technology often succeeds long term.
But investors entering near the peak frequently suffer catastrophic losses because the financing assumptions proved unsustainable.
That’s the part modern markets repeatedly forget.
A correct technological thesis does not automatically produce a correct investment outcome.
The Stress Beneath The Surface Is Growing
Today, we’re seeing multiple signs that liquidity conditions are becoming increasingly unstable beneath the surface.
Bond yields remain elevated despite slowing economic momentum.
Gold continues making new highs.
Speculative assets have become increasingly volatile.
And market leadership has narrowed dramatically into a small cluster of mega-cap technology firms.
At the same time, corporate leverage tied to AI infrastructure continues accelerating aggressively.
That combination matters.
Because bubbles rarely collapse when optimism disappears.
They collapse when financing conditions tighten enough to expose how dependent growth was on cheap capital in the first place.
And increasingly, markets appear vulnerable to exactly that transition.
The real risk isn’t necessarily recession.
The real risk is repricing.
A world where investors suddenly demand sustainable returns instead of narrative-driven growth multiples.
That would force markets to reconsider assumptions around:
future profitability
infrastructure demand
financing durability
and long-duration valuation models
And once that process begins, it tends to feed on itself.
Because leveraged systems do not correct gradually.
They unwind mechanically.
Liquidity, Not Innovation, Is Carrying This Market
If this framework is correct, then the most important variable over the next several years won’t be AI adoption rates.
It will be the cost of capital.
Because modern markets have become structurally dependent on abundant liquidity remaining permanently available.
That assumption now looks increasingly fragile.
Inflation pressures remain persistent.
Governments continue running historically large deficits.
Bond markets are demanding higher yields.
And central banks have far less flexibility than investors became accustomed to during the last decade.
That changes the entire environment.
For years, liquidity expansion masked poor capital allocation decisions because financing remained cheap enough to refinance almost indefinitely.
But if capital becomes expensive again, markets will suddenly begin differentiating between:
productive growth
and speculative expansion
That repricing process could be far more violent than most investors currently expect.
Especially in sectors where future expectations have become disconnected from present-day economic reality.
This Isn’t About Technology
This isn’t about whether AI works.
It’s about whether markets have dramatically overpaid for the speed of its monetisation.
That’s an entirely different question.
The most dangerous bubbles are never built around fake ideas.
They’re built around real innovations attached to distorted financial conditions.
And once enough liquidity enters the system, investors stop evaluating risk rationally.
They start assuming scale itself guarantees inevitability.
History suggests otherwise.
The railway towns eventually emptied.
The radio giants eventually collapsed.
The internet bubble eventually imploded despite changing the world permanently.
The innovation survived.
The excess surrounding it didn’t.
And that’s what today’s market increasingly resembles:
not the birth of sustainable equilibrium but the late-stage acceleration phase before financial gravity returns.
What Happens When Momentum Stops Replacing Valuation?
The question is whether today’s financial system can survive the scale of speculative capital currently attached to that belief.
Because once markets stop measuring value properly, they begin measuring momentum instead.
And momentum works right up until the moment liquidity disappears.
That’s why the old stories matter.
The railway towns.
The empty property developments.
The shoeshine stock tips.
At the peak of every cycle, people stop believing they’re participating in speculation.
They believe they’re participating in the future.
Right before the pricing mechanism breaks.
In a sea of noise, your work is a rare signal. Appreciate the high-quality macro mechanics. Always enjoy reading (or listening to) the material you provide!
Wow, great article, lots of truth bombs here.
Well done!