On this page
How to read this page
Bitcoin attracts two kinds of bad writing: the kind that says it will fix everything, and the kind that says it is worthless. Both skip the interesting part. The serious criticisms of bitcoin are made by serious people, some of them Nobel laureates, some of them bitcoin developers, and several have never been fully answered.
Each section below states a concern the way its strongest advocates would, then the best response, then where things actually stand. Where there is no good answer, it says so.
| Risk | In a sentence | Status |
|---|---|---|
| Price | It can lose half its value in months. | Certain to recur |
| No underlying value | It earns nothing, so its price rests on belief. | Permanent, philosophical |
| Custody | Exchanges fail; keys get lost or stolen. | The most common way people actually lose money |
| Concentration | A few custodians, pools and holders loom large. | Growing |
| Regulation | Rules can change, and differ everywhere. | Improving in the US, unsettled |
| Energy | Mining uses about as much power as a country. | Real; mix getting cleaner |
| Security budget | Miner rewards keep halving; fees have not filled the gap. | Unresolved, long-term |
| Quantum computing | Future machines could break today’s signatures. | Not imminent; preparation contested |
| Illicit use | Ransomware, scams and sanctions evasion use it. | Real but a small share |
| Governance | It may be too hard to change, or changed by the wrong people. | Live debate |
Price risk
The concern. Bitcoin has fallen by more than three-quarters four times, and by more than half as recently as the first half of 2026. People who bought near any of its peaks waited years to break even, and people who borrowed to buy, or who needed the money during a slump, were wiped out. Nothing prevents a deeper or longer fall than any seen so far.
The response. Every one of those falls was followed by a new record, and long-term holders have done extraordinarily well. Volatility has declined as the market has grown. Position size and time horizon are in the holder’s control.
Where it stands. The volatility is not a bug that will be patched. It follows from a fixed supply meeting fickle demand. The historical recoveries are a fact about the past, not a law. Anyone telling you bitcoin cannot go to a small fraction of its current price is selling something. The numbers are on the economics page.
“It produces nothing”
The concern. This is Warren Buffett’s objection, and Paul Krugman’s, and that of most academic economists. A farm produces food, a company produces profits, a bond pays interest. You can estimate what each is worth from what it yields. Bitcoin yields nothing. Its only return comes from selling to someone who will pay more, which is how a bubble works. With no anchor, there is no price at which it is cheap, and no reason it could not go to zero.
The response. The same is true of gold, which has been worth something for five millennia, and of every currency. Money is valuable because it is useful as money: scarce, durable, transferable and widely accepted. Bitcoin’s “product” is a monetary network that no government controls, and people demonstrably pay for that. Seventeen years and five crashes without dying is evidence of something more than a mania.
Where it stands. This is a disagreement about what kind of thing bitcoin is, and no data will settle it. If it is a speculative asset, the critics are right that it has no fundamental value. If it is an emerging money, valuing it by cash flows is a category mistake. What both sides should agree on is that its price depends on continued collective belief to a greater degree than a share or a bond does.
Custody: the risk that actually bites
The concern. The bitcoin network has an excellent security record. The businesses around it do not. Mt. Gox lost about 850,000 BTC in 2014. Bitfinex was robbed of about 120,000 in 2016. In 2022 Celsius, Voyager, BlockFi and FTX all failed with customer funds locked or missing. In February 2025 the exchange Bybit lost about $1.5 billion of ether in the largest crypto theft on record, a reminder that the problem is not confined to the past. On the other side, people who hold their own keys lose them, are phished out of them, or die without telling anyone where they are.
The response. Regulated US exchanges and ETF custodians are far more robust than the offshore venues that failed, hold customer assets separately, and are audited. Hardware wallets and multisig make self-custody safe for anyone willing to learn. The tools have improved enormously.
Where it stands. There is no risk-free way to hold bitcoin. There is only a choice of which risk to take: trusting an institution, or trusting yourself. Most real-world losses come from this category and from outright scams, not from anything wrong with bitcoin. The practical guide covers how to reduce both.
Concentration
Bitcoin is designed to have no centre. In several places, centres have formed anyway.
- Custody. One company, Coinbase, is the primary custodian for most US spot ETFs, which together hold about 1.29 million BTC. A serious failure there, technical, legal or criminal, would be a systemic event. Issuers have begun adding second custodians, but the concentration remains.
- One corporate holder. Strategy owns about four percent of the eventual supply and finances its purchases with debt and preferred shares. If it were ever forced to sell in size, the effect on the market would be severe. The company says its debt is structured to avoid that; 2026 was a partial stress test, which it passed while reporting very large paper losses and selling a small amount of bitcoin for the first time.
- Mining pools. Three pools produce more than 60 percent of blocks. Pools do not own the machines and miners can leave, but pool operators currently decide which transactions are included.
- Hardware. Most mining machines come from two or three manufacturers, chiefly in China.
- Software. The large majority of nodes run one implementation, Bitcoin Core, maintained by a small group of developers funded by a few companies and non-profits.
- Wealth. Early adopters hold very large amounts. Distribution has broadened over time but is still highly unequal.
The response. None of these parties can change the rules, create coins or spend anyone else’s. ETF investors can switch funds; miners can switch pools; anyone can run different software. Concentration of custody is not concentration of control over the protocol.
Where it stands. The protocol remains decentralised where it counts, in who validates the rules. The economy built on top of it has become noticeably more concentrated since 2024. Whether institutional adoption strengthens bitcoin or slowly domesticates it is an open and important question. See who owns bitcoin.
Regulatory and political risk
The concern. Governments cannot switch bitcoin off, but they control the doors: banks, exchanges, tax rules. China has banned trading and mining. Other countries restrict or heavily tax it. In the United States, policy swung from enforcement-led hostility before 2025 to open support afterwards, which shows how quickly it could swing back. The Strategic Bitcoin Reserve rests on an executive order, not a law. The main market-structure bill failed a Senate vote 49 to 50 in September 2026. And bitcoin’s close association with one political movement, and with a president whose family has its own crypto businesses, makes it a partisan target in a way it was not five years ago.
The response. Spot ETFs held by millions of voters, bitcoin on corporate and state balance sheets, and a large domestic mining industry all make a hostile reversal politically costly. Bitcoin itself has been treated as a commodity by US regulators under both parties. And the network has already survived an outright ban by the world’s second-largest economy.
Where it stands. Regulatory risk for bitcoin specifically is lower in the US than at any time in its history, and still real. Tax treatment, self-custody rules and privacy tools are the areas to watch. Elsewhere the picture varies country by country.
Energy and environment
The concern. Mining consumes roughly 138 terawatt-hours a year by Cambridge’s 2025 estimate, with associated emissions of about 40 million tonnes of CO₂-equivalent. Because miners compete, a higher bitcoin price leads directly to more energy use. Natural gas is now the largest single source.
The response. More than half of the energy comes from renewable or nuclear sources, up from just over a third in 2022; coal’s share has collapsed. Miners are flexible loads that can support grids and monetise energy that would otherwise be wasted.
Where it stands. Both are true. The fuller treatment, with the study data, is under the energy debate.
The security budget
The concern. This is the criticism bitcoin developers themselves worry about most. The network is secure because attacking it would cost more than it could gain, and that cost is set by what miners are paid. Today almost all of their pay is the block subsidy. But the subsidy halves every four years, by design, on its way to zero. The plan was always that transaction fees would take over. They have not: through much of 2026, fees were under one percent of miner revenue.
If the bitcoin price doubles every four years, the dollar value of the subsidy holds steady. It cannot do that for ever. Eventually either fees must rise a great deal, meaning the base chain becomes expensive to use, or the amount spent on security will fall relative to the value being secured. A cheaper-to-attack network holding trillions of dollars is an uncomfortable combination. Meanwhile miners are finding that AI companies will pay more for their electricity than bitcoin does.
The response. The subsidy will still be meaningful for decades. Security needs to exceed the profit from an attack, not track market value, and a successful attacker would destroy the worth of what they stole. As bitcoin becomes a settlement layer for large transfers and for second layers, users will pay substantial fees for scarce block space. Demand for inscriptions showed that fee pressure can appear suddenly.
Where it stands. Unresolved. Nobody knows what a mature fee market will look like, and it will not be tested for many years. Proposals to change the 21 million cap to fund security are considered unthinkable by nearly all participants. This is a slow-moving risk, but it is real, and anyone who says it is solved is guessing.
Quantum computing
The concern. Bitcoin’s signatures rely on elliptic-curve mathematics. A large, error-corrected quantum computer running Shor’s algorithm could work backwards from a public key to its private key. Machines of that size do not exist. But in March 2026 Google researchers estimated that fewer than 500,000 physical qubits would be enough, about a twentieth of earlier estimates, and Google has set 2029 as its own deadline for moving to quantum-safe cryptography. US standards bodies plan to retire the vulnerable algorithms by 2035.
The exposure is uneven. Coins in modern addresses that have never been spent from reveal only a hash of the public key, and are safe until the moment they are spent. But research published in 2026 estimated that roughly a third of all bitcoin, between 6.5 and 6.9 million coins, sits in outputs whose public keys are already visible: very old formats, including about 1.7 million coins from the Satoshi era, plus reused addresses.
The response. Quantum-safe signature schemes already exist and were standardised by NIST in 2024. Bitcoin can add them by soft fork, and BIP-360 is a concrete first step. The same threat applies to almost all internet security, so bitcoin will not be facing it alone or without warning.
Where it stands. The engineering is tractable; the politics is not. Quantum-safe signatures are far larger, which means fewer transactions per block. Migration would take years of coordination in a community that has just demonstrated how hard it finds change. And there is no agreement on what to do about coins that never migrate, above all Satoshi’s: leave them for a quantum thief, or freeze them and break the promise that no one’s bitcoin can be confiscated. Expect this to be the defining technical argument of the next few years.
Illicit use
The concern. Bitcoin’s first real market was Silk Road. It remains the standard currency of ransomware, is used in darknet markets and by sanctioned states, and is the tool of choice for investment scams, because victims can be talked into sending payments that cannot be recalled. North Korean state hackers have stolen billions in crypto.
The response. Blockchain analytics firms consistently estimate illicit activity at a small share of total crypto volume. Bitcoin is a poor tool for crime: every transaction is public and permanent, and investigators have traced and seized enormous sums, including most of the Bitfinex proceeds and the coins that now make up the US reserve. Much criminal activity has shifted to stablecoins and other networks.
Where it stands. Criminals use bitcoin, as they use dollars and phones. The scale is real in absolute terms and modest in relative terms. The harm that reaches most ordinary people is fraud, covered under common scams.
Governance: too rigid, or not rigid enough?
The concern, version one. Bitcoin has not changed its consensus rules since 2021. Proposals with years of work behind them, such as covenants, have no path to activation. A system that cannot adapt may be unable to fix its security budget, respond to quantum computing or scale beyond custodial services. “Ossification” might be another word for paralysis.
The concern, version two. A small number of developers maintain the software nearly everyone runs, and their funding comes from a short list of companies. In 2025 they changed a long-standing default over the objections of a vocal part of the community. If a few maintainers can do that, critics ask, how decentralised is the project?
The response. These two criticisms answer each other. The 2025 change was to a relay default, not a consensus rule, and nobody was forced to adopt it; by 2026 up to a fifth of reachable nodes were running the alternative client. The 2026 attempt to change consensus rules without broad agreement went nowhere. That is the design working: easy to opt out, hard to impose. A monetary system whose rules are difficult to change is doing what its holders want.
Where it stands. Bitcoin’s resistance to change is its central feature and its central risk, and you cannot have one without the other. The events are described under current debates.
The bottom line
Bitcoin is an experiment that has run far longer and grown far larger than almost anyone predicted, including many of the people on this page. That is not proof that it will keep working. The honest position holds two thoughts at once: it has survived everything thrown at it so far, and several of its hardest tests are still ahead.
If you are considering exposure, the risks most likely to hurt you personally are, in order: being scammed, losing access, buying more than you can afford to see halve, and panic-selling when it does. None of those is a flaw in the protocol. All of them are avoidable. This is general education, not advice; for decisions about your own money, speak to a qualified adviser.