Crypto: An Evidence-Based Guide

Crypto attracts two kinds of writing: promotional and dismissive. Both skip the hard parts. This guide explains the technology accurately, shows the actual volatility and drawdown record, catalogues how people lose coins, and gives a harm-reduction framework for anyone who decides to own some.

By Karanvir Singh, founder of Quantix Invest

Where this guide stands

Two honest statements can be true at once. First: Bitcoin and its descendants solved a genuinely hard computer science problem, and the systems work as designed, running continuously for over fifteen years without a central operator. Second: none of that tells you what a coin should be worth, and the price history is one of the most violent of any traded asset in modern financial history.

So this page does not tell you crypto is the future of money, and it does not tell you it is worthless. It tells you what the thing is, what the record shows, and what goes wrong. If after reading it you want to hold some, the sizing section exists so that being wrong is survivable. If you want to hold none, that is a completely defensible portfolio, and you will not have missed anything you needed.

How a blockchain actually works

Strip away the jargon and a blockchain is a shared ledger that thousands of independent computers maintain identical copies of, with rules that make it very expensive to rewrite history.

The ledger. There are no coins in the sense of objects. There is only a list of transactions. Your balance is whatever the ledger says is spendable from addresses you control. Nothing else exists.

Keys. Ownership is proved by a private key, a very large secret number. The matching public key generates your address. To spend, you sign a transaction with the private key. Anyone can verify the signature; nobody can forge it without the key. This is why the phrase "not your keys, not your coins" is not a slogan but a literal description of the system: whoever holds the key can move the funds, and there is no appeal.

Blocks and the chain. Transactions are gathered into blocks. Each block contains a cryptographic fingerprint (a hash) of the block before it, so blocks form a chain. Change any old transaction and every subsequent fingerprint breaks, which is what makes the record tamper-evident.

Consensus. The hard part is getting strangers to agree on which block comes next without a referee. Two dominant answers:

• Proof of work (Bitcoin). Miners race to find a number that makes the block's hash fall below a target. Finding it requires enormous trial-and-error computation, so proposing blocks costs real electricity. Rewriting history would require redoing all that work faster than the honest network, which is prohibitively expensive at Bitcoin's scale. The system adjusts difficulty roughly every two weeks so blocks arrive about every ten minutes regardless of how much hardware joins.
• Proof of stake (Ethereum since 2022, and most newer chains). Validators post the network's own token as collateral. They are selected to propose and attest blocks, earn rewards for honest behavior, and lose part of their stake for provable misbehavior. This cut Ethereum's energy use by roughly 99.9%, and it changes the security model from "attacking costs electricity" to "attacking costs you your collateral."

Supply rules. Bitcoin's issuance is fixed in code: the block reward halves roughly every four years, and the total supply asymptotically approaches 21 million. That schedule is enforced by every node independently, which is what "no one can print more" means in practice. Other chains have different and sometimes changeable rules; do not assume scarcity because one system has it.

Smart contracts. Ethereum generalized the idea: instead of only recording transfers, the network runs programs. A smart contract is code deployed to the chain that executes exactly as written when called, holding and moving funds according to its logic. This enables lending protocols, exchanges, and tokens, and it also means bugs are financial. Hundreds of millions of dollars have been drained from contracts through logic errors, because the code is the rule and there is no undo.

Bitcoin, Ethereum, and everything else

Bitcoin does one thing: move and store a scarce digital token with a fixed supply schedule and the strongest track record of continuous operation. The pitch is monetary, not technological. Its block space is deliberately limited, which keeps running a full node cheap enough for individuals, at the cost of low throughput. Payment layers built on top (the Lightning Network being the main one) handle fast small payments off-chain.

Ethereum is a general computing platform. Its token pays for computation. Most tokens, stablecoins, lending protocols, and NFTs live on Ethereum or on chains that copy its design. It changes more often than Bitcoin, which is both its strength and an added risk.

Everything else. Tens of thousands of tokens exist. A small number are serious competing platforms. The overwhelming majority are speculative instruments with concentrated insider ownership, thin liquidity, and no mechanism connecting the token to any cash flow. The empirical pattern across every cycle since 2013 is the same: the long tail of tokens outperforms wildly during the mania and then loses 90% or more, with a large fraction going effectively to zero and never recovering. This is not a prediction. It is a description of what has already happened three times.

Memecoins deserve a plain sentence: they are explicitly not backed by anything and their creators generally say so. Buying one is a bet that someone else will buy it from you at a higher price. That is a legitimate thing to know you are doing, and a terrible thing to do accidentally.

What makes it worth anything

A stock has a claim on earnings. A bond has contractual payments. A rental property has rent. What does a token have?

The honest answer is: expectations, plus whatever utility the network provides. A crypto asset produces no cash flow, so standard valuation tools do not apply. Its price is set entirely by what the next buyer will pay. That is not automatically damning. Gold has behaved the same way for millennia and holds a real place in the world's financial system. But it does mean nobody, including confident people with charts, can compute a fair value. Models based on stock-to-flow ratios or halving cycles have been publicly wrong by large margins.

The arguments that have some substance:

• Censorship-resistant settlement. For people under capital controls, in unstable currencies, or cut off from banking, being able to hold and move value without permission has demonstrable value. This is the use case with the clearest evidence, and it is not one most readers of this site have.
• Credibly fixed supply. If you believe governments will keep expanding money supplies, an asset whose supply cannot be expanded by decree is a plausible hedge. The 2022 experience complicates this: inflation surged and Bitcoin fell about 65% in the same year, behaving like a risk asset rather than an inflation hedge.
• Programmable settlement infrastructure. Stablecoin transfer volumes running into the trillions annually suggest real demand for moving dollars on open networks quickly. Whether that demand accrues value to any particular token is a separate question.

The arguments that do not hold up: "adoption is inevitable" is not evidence, network-value formulas borrowed from telecoms are not valuation, and the fact that a price went up is not a reason it should have.

Volatility and the drawdown record

This is the section people skip and should not. Bitcoin's annualized volatility has typically run in the range of roughly 50% to 80%, several times that of the S&P 500. More usefully, look at the peak-to-trough declines.

Four drawdowns of 75% or more in about a decade. For comparison, the S&P 500's worst decline since the Great Depression was roughly 57%, and that one is treated as a generational event. Altcoins have generally fallen further and recovered less, and many have not recovered at all.

Worked example: what a 5% allocation does to a portfolio. Start with $100,000: $95,000 in a diversified stock and bond portfolio, $5,000 in Bitcoin.

• Crypto goes to zero. You are down $5,000, or 5%. Annoying, recoverable, roughly one ordinary bad quarter in stocks.
• Crypto falls 80%, the historically typical bear. You lose $4,000, 4% of the portfolio. Meanwhile the other $95,000 is doing whatever it was going to do.
• Crypto goes up 5x. Your $5,000 becomes $25,000, adding $20,000 to your net worth, a 20% gain on the total portfolio from one small sleeve.

Now run the same three cases at a 50% allocation: minus $50,000, minus $40,000, plus $200,000. The upside case is life-changing, and so is the downside case. The asymmetry that makes a small position sensible is the same asymmetry that makes a large one reckless, because you have to survive the path, not just the destination. An 80% drawdown on half your net worth ends most people's investment plans, and the historical record says you should expect to face one.

The diversification claim, tested

Crypto was widely marketed as an uncorrelated asset. The early data supported it, mostly because the asset was small and traded by a separate crowd. As institutional ownership grew, the correlation with risk assets rose. Through the 2022 tightening cycle, Bitcoin fell hard alongside technology stocks, with rolling correlations to the Nasdaq reaching levels that made the diversification pitch difficult to defend precisely when it was needed.

The lesson generalizes beyond crypto: correlations measured in calm markets tend to rise toward one in a panic, because in a panic everyone sells whatever they can. Any asset whose diversification case rests on a short and recent correlation estimate deserves skepticism.

The residual case is weaker but not empty: crypto's return drivers are not identical to equities', and a very small allocation adds a return stream that does not move in lockstep every year. That is a modest claim, and it justifies a modest position.

Custody: the risk unique to this asset

With stocks and bonds at a regulated US broker, records are held by a custodian, SIPC covers certain broker failures, and if you lose your password you can prove who you are and get back in. None of that is native to crypto. Three failure modes have destroyed enormous amounts of value.

1. Exchange failure. The list is long: Mt. Gox in 2014, which held a substantial share of all Bitcoin trading and left creditors waiting over a decade; Celsius and Voyager in 2022; and FTX in 2022, where customer assets were reportedly commingled and misused, with roughly $8 billion missing at the point of collapse. Deposits at a crypto exchange are generally an unsecured claim on that company. There is no FDIC insurance on crypto, whatever the marketing page implies, and phrasing about insurance usually refers to a hot-wallet policy covering the company, not you.

2. Self-custody failure. Holding your own keys removes counterparty risk and adds operational risk that is entirely yours. Lost seed phrases, dead hardware wallets with no backup, house fires, and deaths without a documented recovery plan have permanently removed a large fraction of all Bitcoin from circulation. Estimates commonly put lost coins in the millions of BTC. Irreversibility cuts both ways: nobody can seize it, and nobody can restore it.

3. Theft. Phishing sites, malicious browser extensions, fake wallet apps, SIM swaps that defeat text-message two-factor authentication, malware that swaps a pasted address for the attacker's, and support impersonators. Transactions are final. There is no chargeback.

Practical custody guidance. If you hold a small amount, a large regulated exchange with strong security practices is a reasonable tradeoff and far better than a badly executed self-custody setup. If you hold an amount you would be devastated to lose, a hardware wallet with the seed phrase written on paper or steel, stored in two physically separate secure locations, never photographed and never typed into any device, is the standard approach. Test the recovery process with a small amount before trusting it with the rest. Use an authenticator app rather than text messages for two-factor authentication everywhere. Write down, in your estate documents, how a trusted person could recover the assets, because an undocumented cold wallet is indistinguishable from destroyed money.

Spot exchange-traded funds, approved in the US in 2024, changed the calculus for many people. They hold the asset through an institutional custodian, trade in an ordinary brokerage account, are eligible for retirement accounts, and remove key-management risk entirely. In exchange you pay an expense ratio, you cannot move the asset off-chain, and you have added the fund sponsor and custodian as counterparties. For a small allocation held inside a long-term portfolio, that trade is often the sensible one.

Stablecoins, yield, and DeFi

Stablecoins are tokens designed to hold a constant value, almost always one US dollar. The credible ones are fully backed by short-term Treasury bills and cash, with regular attestations. They are useful for moving dollars quickly and are genuinely used at scale. They are also, in economic substance, a private money-market instrument with no government guarantee. The reserve composition and the issuer's jurisdiction are the whole risk.

Algorithmic stablecoins tried to hold the peg through trading incentives rather than reserves. In May 2022 the largest of them, UST, and its companion token LUNA collapsed from a combined market value in the tens of billions to essentially nothing within days, taking a chain of lenders down with it. That episode is the reference case for why an unbacked peg is a promise, not a mechanism.

Yield. Any offer of high, stable yield on a crypto deposit deserves the same question you would ask anywhere: who is paying it and from what? Legitimate sources exist, including lending to leveraged traders and staking rewards paid in newly issued tokens. But in 2021 and 2022, several platforms advertising 8% to 18% on stablecoins were funding it with proprietary trading, related-party loans, or token emissions. Most of them are gone and their depositors were unsecured creditors. If a rate is far above short-term Treasury yields with claimed low risk, the risk is somewhere you have not been shown.

Staking pays for genuine work securing a proof-of-stake network. The rewards are real, denominated in the token, and come with lockup periods, slashing risk, and platform risk if you stake through an intermediary. A 4% yield on an asset that can fall 80% is not a bond.

DeFi protocols are useful to understand and hazardous to use casually. Smart contract bugs, oracle manipulation, and governance attacks have drained billions. Audits reduce risk; they do not eliminate it.

How people actually lose money

Very little of the loss in this space comes from thoughtful investors watching a considered position decline. Most of it comes from a short list of repeatable patterns.

• Romance and "pig butchering" scams. A stranger builds a relationship over weeks, then introduces a trading platform that shows fake gains and blocks withdrawals. Losses reported to US authorities from crypto-related investment fraud run into the billions annually, and this category is the largest piece.
• Impersonation. Fake support agents, fake wallet-recovery services, and fake versions of legitimate exchange websites bought as search ads. No legitimate service ever asks for your seed phrase, ever, for any reason.
• Rug pulls. A token launches, insiders hold most of the supply, promoters generate excitement, insiders sell into it, liquidity is withdrawn, the price goes to zero.
• Leverage. Perpetual futures on offshore venues offer leverage of 20x or higher. Liquidation cascades during ordinary crypto volatility routinely wipe out billions of positions in an hour. Leverage on an asset with 70% volatility is a mathematical guarantee of eventual liquidation.
• Buying the top of the cycle. Retail inflows peak with prices and media attention by every measure we have. The average dollar invested in crypto has done considerably worse than the average coin, because most dollars arrive late.

Position sizing and harm reduction

If you want exposure, the following framework keeps a bad outcome from becoming a life event.

1. Fill the boring parts first. Emergency fund funded, high-interest debt gone, employer retirement match captured, and a diversified core portfolio in place. A guaranteed 100% return from a 401(k) match beats any speculative expected value, and paying off a 22% credit card is a risk-free 22%.

2. Cap it as a share of net worth, not of enthusiasm. A common range among advisers willing to include crypto at all is 1% to 5% of investable assets. Above roughly 10%, the asset dominates your portfolio's risk regardless of how it looks in a table, because its volatility is several times everything else's. Write the number down before you buy.

3. Decide in advance about rebalancing. If your 3% position becomes 12% after a run, selling back to 3% is what turns volatility into realized gains and keeps risk where you chose it. Most people find this psychologically hard, which is exactly why the rule should be written before it is needed.

4. Buy on a schedule. With an asset this volatile, spreading purchases across months removes the single worst decision (putting everything in on one euphoric day) at essentially no cost.

5. Prefer the simplest holdings. Concentration in one or two large assets with the longest track records is more defensible than a basket of small tokens, where the historical base rate of permanent loss is high.

6. Keep records from day one. See the next section.

7. Have a written plan for what would change your mind, in both directions. Positions held without an exit thesis tend to get sold at the worst moment or held to zero.

Taxes and record keeping

In the US, crypto is treated as property, not currency. That has consequences people discover too late:

• Selling for dollars is a taxable event. So is trading one token for another, and so is spending crypto on goods. Each is a disposal with a gain or loss.
• Holding periods matter: gains on assets held over a year get long-term rates; under a year, ordinary income rates.
• Staking rewards and mining income are generally taxable as ordinary income at the value when received, and that value becomes the cost basis for later disposal.
• Cost basis reporting from exchanges has improved with newer broker reporting rules, but transfers between wallets and platforms routinely break basis tracking. Keep your own records.
• The wash sale rule has historically been applied to securities, and its application to digital assets has been the subject of proposed legislative changes, so confirm current treatment rather than relying on old advice.

A person who made two hundred small trades across three exchanges and two wallets in a year has created a tax problem that costs real money and time to untangle. Fewer, larger, documented transactions are cheaper in every sense.

Common mistakes and the bottom line

• Sizing by conviction instead of by consequence. Conviction is not a risk measure. The only question that matters is what happens to your life at zero.
• Assuming past cycles repeat on schedule. Four-year cycle narratives are pattern-matching on a handful of observations.
• Leaving meaningful sums on an exchange indefinitely. Every collapse looked fine right up to the week it did not.
• Chasing yield. Double-digit "safe" returns on dollars have a near-perfect record of being something other than what they claimed.
• Buying small tokens after a large move. The base rate of permanent loss in the long tail is brutal and well documented.
• Treating an 80% drawdown as unthinkable. It is the norm for this asset, not the tail.
• No inheritance plan. Self-custodied assets with no documented recovery path die with you.

Bottom line. The technology is real, the systems work, and the honest case for a small allocation is that it is a high-variance asset with a return stream unlike the rest of your portfolio and a genuine use case for people the traditional system does not serve. The honest case against is that it produces no cash flow, cannot be valued, has fallen 75% or more four times in a decade, failed its inflation-hedge test in 2022, and sits in an ecosystem where custody failures and fraud have destroyed tens of billions of dollars of ordinary people's savings.

Both of those are true. Neither justifies a large position. If you hold none, your financial plan is complete without it. If you hold a few percent, keep it small, keep it simple, secure it properly, record it for tax purposes, rebalance it mechanically, and never let it become the thing that determines whether you retire.

This guide is education, not individualized financial advice, and nothing here is a recommendation to buy or sell any asset.