crypto 2029

Lukas (@MiyaHedge)

Published June 2026 · To be scored in 2029

You are sitting in front of the biggest change in the history of crypto and if you want to remain in this industry, you need to pay attention to what's happening right now.

There are three core questions this industry has right now. What makes a token valuable? How do we translate different technology frontiers into blockchain? What happens when crypto stops being an asset class and becomes financial infrastructure for traditional finance?

I could argue each of these in the abstract. Plenty of people do, every day, and the arguments never resolve anything.

So I'm going to do something different. I'm going to write down, period by period, what I think actually happens in this industry between now and 2029 with names, numbers, and dates, concretely enough that you can come back in three years and check whether I was right. This is one future out of many, and parts of it will be wrong. But vague claims about the future are unfalsifiable, and unfalsifiable claims are worthless. I'd rather be specific and wrong than vague and safe.

This forecast is informed by where I sit: I'm working inside the intersection of crypto startups, regulations and venture and am talking to alternative asset managers and capital allocators every week. That doesn't make me immediately right — sure. It means my guesses are priced against real constraints.

Mid 2026: The Only Good Assets Aren't Tokens

It's mid 2026, the pre-IPO perpetual markets find product-market fit before anyone agrees on what a token should be worth.

It starts on Hyperliquid. The SpaceX pre-IPO perp, dismissed initially after a big Ventuals liquidation manipulation,¹ becomes the most-watched price feed in private and public markets. By July, banks and hedge funds reference it when marking their own private positions, and consumer apps like Robinhood use it to orient post-IPO pricing. In the weeks before each major listing, the perp converges on the opening print with embarrassing accuracy — embarrassing, that is, for the underwriters who charged seven figures to arrive at the same number. OpenAI and Anthropic perps draw even larger open interest. For a brief window, a crypto-native exchange is the closest thing the world has to a live price for its most important private companies.

Retail, meanwhile, is asking a more basic question: why does anything else on-chain trade at all? The altcoin market has spent eighteen months bleeding out, founders and funds exiting through DATs and TWAP sales, while $HYPE, the one token with a working value-accrual loop, outperformed everything. A dozen value-accrual mechanisms were proposed, and most never reached escape velocity, because they all shared the same flaw: the mechanics were attached to worthless companies. The industry has solved how a token could capture value before acquiring any assets worth capturing value from.

That inversion is the quiet engine of the pre-IPO boom. The demand was never for perps, but for quality and the only quality available on-chain, in mid-2026, is a synthetic claim on companies that have nothing to do with crypto.

Late 2026: AI Doesn't Need Crypto

Anthropic and OpenAI hit technological escape velocity, the foundation model battle and AGI is starting to get priced in early, with that one of the consequences are that everyone who is not one of the core foundation model companies and working in this sector is starting to bleed. Capital is starting to price AGI as something a company holds on its balance sheet and not something commoditized for everyone surrounding it.

AI x crypto dies quietly in this environment, not because the thesis was refuted but because nobody had time to refute it. The x402 payment rails ship and find no payers; the agent economy that was supposed to need on-chain money never arrives at scale, and the agents that do exist settle in dollars through APIs, the way software always has. The honest sentiment in venture now is that AI did not need crypto, and the venture class stops pretending otherwise.

The single AI x crypto product with genuine product-market fit right now are prediction markets, foundation model predictions hit escape velocity since they are the most accurate financial vehicle to express the thing that single handily moves the largest amount of capital: Who has the best model in 1 month.

Away from the tape, something less entertaining is happening. After the CLARITY Act was considered a nothing burger by most traders after it passed the senate mid 2026 and nothing rallied immediately, tokenization programs are accelerating during the end of 2026. Large asset managers move from pilot to production with no fanfare, because fanfare is precisely what their compliance departments are paid to prevent: money-market funds, private credit, the unglamorous and boring middle of the balance sheet. None of this trades. There's no price chart, neither are there KOLs on CT who bullpost a transfer agent filing.

By the end of 2026, crypto has two economies that barely acknowledge each other. The loud one prices the AI race. The quiet one is being absorbed into the financial system, one filing at a time. Almost everyone watches the first.

Early 2027: The Foundations Pick a Side

The general-purpose chains run out of road for ambiguity.

For years, every major foundation told two stories at once, consumer adoption on stage, institutional readiness in the data room, and the stories never had to meet. By early 2027 they meet.

The consumer story is extremely concentrated on the tape: the only retail product that found demand concentrated its volume on a handful of venues. The institutional story, meanwhile, is the only one with a paying customer attached. One by one, the foundations take conviction, mostly in the same direction: enterprise sales teams, compliance help, network wide compliance SDKs for tokenised transfer agents and broker dealer licensing, Wall Street relationships and privacy features.

Every pivot gets interpreted the same way by the media and on CT: as a choice, institutions over consumers, the serious customer over the casino.

Inside the foundations, almost nobody actually believes the framing and they internally double down on consumer crypto, just through a different lens. Accreditation thresholds have been loosening for years, and the population that qualifies keeps widening, which means the “institutional” products the foundations are building rails for are aimed, on a short delay, at consumers who simply haven't been called that yet. The people building the rails know it. It's just not something you put in an announcement. The teams building compliance infrastructure talk about banks, because banks are who's paying.

But the quiet economy from late 2026 has just acquired something it never had before: a future retail customer. The two economies that barely acknowledged each other now share a membrane, and the membrane is an accreditation check.

Mid–Late 2027: Three Ceilings

A new generation of companies makes the private markets frothy again. Bio x AI, physical AI, humanoid robotics, the rounds are oversubscribed, the valuations are vertical, and every one of them is years from any public listing. The perp venues list them within weeks; synthetic open interest on companies that barely have revenue sets records. The pattern from 2026 repeats at higher stakes: the world's most wanted assets are private, and the only version of them you can touch has a 8h funding rate.

The real thing exists now, and it grows the way private markets have always grown — through verified channels, compounding quarter over quarter, invisible to a timeline that only registers vertical lines. The gap to the perps' growth rate has a specific cause: private securities cannot be generally solicited, so the one distribution engine crypto has ever mastered — someone posts a chart, the crowd arrives — can't legally touch this asset class yet. The perps, meanwhile, have a ceiling of their own, and theirs is structural: a perpetual needs a catalyst close enough to price, which confines synthetics to late-stage names with a listing in sight. Everything earlier, think the mid-stage rounds, the Bio x AI and robotics names years from any exit, has no viable synthetic expression at all. For most of the private market, regulated ownership isn't the slow alternative; it's the only instrument that can exist. It just isn't allowed to publicly introduce itself yet.

Stablecoins hit a different ceiling. Supply keeps grinding higher, it never stops, but the expansion plans quietly shrink. The midterms changed committee math, the 2028 field is forming, and several of the loudest voices in it are running against private dollar issuance. The statutes passed in 2025 and 2026 still stand, but statutes are implemented by administrations, and every bank treasurer modeling a ten-year settlement build now has to price a scenario where the next one is hostile. Nobody cancels anything. They extend timelines, shrink pilots, wait for November 2028. The dollar moves on-chain at exactly the speed of political certainty, which in mid-2027 is low.

The same caution spreads through the rest of the quiet economy. Tokenized private credit and fund shares keep launching, and keep landing in the same place: production-grade, compliance-approved, and deliberately small, because no one wants to be the case study in next year's senate hearings. The pattern across all three threads is identical even though the causes aren't, the products work, the demand is proven, and the throttle is held by something outside the industry's control.

On any chart but crypto's own, 2027 is a strong year. The industry just spent a decade training itself to read anything linear as failure.

2028: Permission Stops Being Scarce

(From here, the resolution drops. The nearer periods of this paper were forecast in quarters; the rest is forecast in years, and the error bars widen accordingly. One assumption is made explicit now: this scenario assumes the Democratic candidate wins in November 2028. Under the alternative, the timing of what follows shifts, the structure mostly doesn't.)

The casino finishes deflating, and almost nobody marks the date. The extraction machinery had become too efficient for its own survival: every storm of new liquidity in 2026 and 2027 was smaller than the last, drained faster, by fewer and more concentrated actors. There is no crash to point to — memecoin storms still arrive, charts still go vertical for an afternoon, but somewhere in the first half of 2028, the casino quietly stops being the industry's center of gravity. Its volume becomes a statistic instead of a culture. Some of its traders migrate to prediction markets, which inherit the energy; some stay in the shrinking pit; a surprising number have spent the past year doing something nobody predicted of them in 2026 — passing accreditation checks.

The political fear unwinds the same way it arrived: as a repricing, spread across the year. The likely nominees take the industry's money and say the same sentence in different accents — regulated, not banned. The actors who treated the previous administration as an extraction window face investigations, and the industry slowly realizes the cleanup is the bullish signal, not the bearish one: a government distinguishing extraction from infrastructure is precisely what makes infrastructure safe to fund. The bank treasurers who shrank their pilots in 2027 begin un-shrinking them, quietly, before November — by the time the result is in, most of the fear-premium is already gone.

The year's real lesson is delivered by the market everyone could see. Early in 2028, on one of the largest venues, a position big enough to move the mark unwinds through the most crowded pre-IPO perps, and the cascade the structure had been promising since Ventuals finally arrives at scale. Billions in open interest deleted in hours, positions auto-deleveraged, losses socialized, winners paid in haircuts. The postmortems never settle whether it was manipulation or an accident, and the ambiguity is itself the finding: in a market with no anchor, there is no true price to deviate from, so manipulation can't even be defined, let alone proven. A perpetual on a public stock is disciplined by the spot market underneath it. The pre-IPO perps had nothing underneath — the real shares existed, trading quietly on regulated venues, but couldn't be generally distributed or referenced at scale, so every mark was a venue's guess, and a guess can be moved. The cascade wasn't the synthetic market failing. It was the synthetic market working exactly as designed, in the absence of the real one.

For a decade, the solicitation ban was defended as investor protection. What the wreckage demonstrates is that it had been protecting people from the enforceable version of the trade while leaving them alone with the leveraged, unanchored one. The dividing line that matters was never synthetic versus real. It was enforceable versus not.

The relief arrives in the aftermath, and it reads less like a reform than like market plumbing: guidance permitting general solicitation of resales of private securities, secondary markets, not primary raises, to verified accredited buyers, whose ranks have been widening for years. The logic is almost dull: the synthetic markets need an anchor, and the cheapest anchor is a real market people are allowed to know exists. A ninety-year-old speech rule, narrowed in an afternoon, as a derivatives fix.

The first week is the memecoin launch all over again, except the charts are real companies. Resale listings get posted, screenshotted, shilled, legally, for the first time in the asset class's history. The discourse splits on contact: half the timeline calls it the new primitive, half asks whether retail just became exit liquidity for venture. The second half has the right instinct and the wrong decade — the question was fair when the assets were tokens attached to nothing. These are claims on the companies the perps spent two years proving everyone wanted. The flow that ignites first is exactly where the synthetic markets said it would be: the late-stage names everyone already knows, then — because ownership has no funding rate and no catalyst requirement — outward into the mid-stage frontier the perps could never reach. The perps don't die; they become the late-stage annex of a market that no longer needs them to be the whole thing.

By December, the industry has its bull market, and it runs on the oldest primitive in finance, newly allowed to introduce itself.

2029: The Market Is the Only Thing Left to See

The bull market's first full year doesn't look like the previous ones, and the difference is the point. The assets going vertical have real companies building real things that actually benefit humanity. The new primitive ordinary people trade is the private company — the biotech name three trials deep, the robotics firm everyone's seen the demo of, the AI lab whose perp they traded in 2026 and whose stock they can now actually hold. Accreditation, loosened in steps for a decade, has quietly produced a retail class that buys what only institutions could touch five years earlier, and most of them never think of any of it as crypto.

The token complex splits along the line this paper opened with. The chains that became settlement and issuance infrastructure for the new market capture real flow, and their tokens trade like claims on that flow. Everything else faces a market that has become brutally literal: a token without an enforceable claim or a working value-accrual loop doesn't bleed out over eighteen months anymore — it simply never trades. The value-accrual debate of 2026 didn't get won by any mechanism. It got obsoleted by the arrival of assets that never needed the debate.

Stablecoins do in 2029 exactly what they did every year of this scenario: compound, meaningfully, without a hockey stick. Supply roughly doubles from mid-2027 levels by year-end — call it 20% a year, sustained — and the ceiling above that pace is not a market failure but a policy choice that both parties, in different vocabularies, keep making: private dollar issuance grows at whatever speed keeps it useful and stops it from competing with the sovereign balance sheet. The dollar moved on-chain at the speed of political certainty, which by 2029 is moderate, and permanent.

The casino still exists. It runs in the corner it deflated into, storms occasionally, and matters roughly as much as any other corner of the entertainment economy. Its alumni are distributed across prediction markets, the new secondaries, and — the part nobody predicted in 2026 — accreditation paperwork.

And the third question of this paper, the one about crypto becoming financial infrastructure, resolves the only way it ever could: by becoming impossible to ask. There is no event. Settlement lives somewhere — purpose-built rails, public chains, combinations nobody outside an operations team could diagram — and the ordinary participant neither knows nor cares, the way nobody knows which clearinghouse stands behind their brokerage account. The absorption that began one filing at a time in late 2026 completes itself by disappearing from view. The infrastructure won by becoming boring. What's left visible is the thing this industry was always actually building, underneath every cycle of pretending otherwise: a market.

So, the three questions, answered the way the scenario answered them.

What makes a token valuable — the same thing that was always going to: an enforceable claim on something real, now enforced by a market harsh enough to delete everything without one. How do technology frontiers get translated into blockchain — through private markets: the frontier companies never needed tokens, they needed venues, and once the venues could speak in public, the frontier listed itself. What happens when crypto becomes infrastructure for traditional finance — nothing happens; it abstracts, and the question stops parsing.

Parts of this will be wrong. That was the deal in the introduction. The spine of it fails under one observation: if, by the end of 2028, retail demand for private companies has not found a legal channel — no relief, no widened access, the offshore synthetic perpetuals and wrappers still carrying the flow — then the central claim of this paper, that the bottleneck was law and not technology, is broken, and you should discount everything built on it.

Watch that one variable. Score the rest in 2029. I'd rather lose specifically than win vaguely.