A 7×24 market that never closes, a complete map.

Crypto is the only “newly born” asset class of the past decade, and it’s open 7×24 globally, has no central-bank backstop, and settles on public chains visible to everyone — those three facts mean it plays differently from equities, FX, or commodities. This guide systematically walks through crypto’s market structure, CEX vs DEX, perpetual contracts + funding rate, on-chain data, DeFi yields, MEV, typical strategies, and custody / regulatory / counterparty risks. After reading it, you should be able to follow trader jargon like “BTC funding rate annualized 30%, ETH beacon staking + restaking 5.4%, Uniswap LP impermanent loss,” and know where each kind of money is made and where the risk comes from. Read alongside §09 (crypto quant) of A taxonomy of quant trading schools for best results.

§01 · Framework — the 10-minute crypto framework

In 2009 Satoshi Nakamoto released the Bitcoin whitepaper; by 2025, total crypto market cap is ~$3.5T (about 15% of global gold market cap, or one Apple). This market has four fundamental differences from the equity market you’re used to: 1) 7×24 globally, no close; 2) no issuer / central-bank backstop, almost no intrinsic cash flow; 3) assets can be issued permissionlessly by anyone (memecoins, shitcoins); 4) every transaction is recorded on a public chain, anyone can look at Vitalik’s wallet balance.

These four properties produce its unique playbook: on-chain data becomes the new “fundamentals,” perpetual contracts + funding rate become the dominant trading instruments, MEV (Maximal Extractable Value by miners / validators) becomes its own industry, and custody risk (exchange blowups) far exceeds the price volatility itself.

1. Market size. Total cap ~$3.5T (2025): BTC ~$2T at 55%, ETH ~$500B at 14%, stablecoins ~$220B at 6%, the remaining alts at 25%. Daily volume ~$80-150B; the top 5 exchanges represent 80%+. CoinGecko / DefiLlama / TradingView are the standard free toolkit.
2. Who’s trading. Four forces: 1) retail (50%+ of daily volume, split between Asia and North America); 2) crypto-native funds (Pantera, Multicoin, Galaxy); 3) traditional institutions (BlackRock spot BTC ETF already $40B+ AUM); 4) quant market makers / HFT (Jump, DRW, Wintermute, GSR, Cumberland). SEC’s January 2024 spot-BTC ETF approval was the institutional-entry watershed.
3. What sets the price. No “DCF” — most crypto assets generate no cash flow (BTC has none; ETH gets ~3% via staking). Price drivers: 1) USD liquidity (Fed hikes / QT); 2) retail sentiment + leverage; 3) regulatory events; 4) on-chain native demand (stablecoin usage, DeFi TVL). Like gold — “valuation” barely exists; price = belief × liquidity.
4. Three product layers. 1) Spot (BTC / ETH / Stablecoin / Alts) 2) Derivatives (perpetuals, quarterly futures, options) 3) DeFi (AMM market making, lending, staking, yield farming). Stack them and the strategy space is huge. 2024 crypto derivatives volume is 4-5× spot.
5. Three infrastructure layers. 1) Public-chain layer (Bitcoin / Ethereum / Solana / various L2s) 2) Protocol layer (Uniswap / Aave / Lido / EigenLayer) 3) Application layer (wallets like MetaMask, aggregators like 1inch, perps like dYdX). Each layer has its own “tokenomics” and investment opportunities. See §02 Market structure + §07 DeFi.
6. Volatility levels. BTC long-term RV 50-80%, alts 100-200%, memecoins 500%+ (annualized). 3-10× SPX. So the same “small position” swings 3-10× more in P&L. Position management matters far more than in traditional markets. Position sizing in Kelly criterion.

Bottom Line · Crypto’s first lesson: survive.

Crypto’s cumulative returns over the past 15 years are the highest in financial history (BTC ~200% annualized), but during the same period 90% of tokens went to zero, and 3 of the 5 largest exchanges have blown up (Mt. Gox, FTX, Celsius / Voyager). The market features extremely right-skewed returns + extremely left-tailed blowups. Before you study “which strategy has the highest alpha,” study “how not to go to zero.”

§02 · Market structure — CEX · DEX · OTC

Crypto trading happens in three places: centralized exchanges (CEX), decentralized exchanges (DEX), and over-the-counter (OTC). The split is roughly 75% / 15% / 10%, and changes every year.

CEX · centralized exchanges

Binance, Coinbase, OKX, Bybit you know belong here. They essentially turn crypto into a row in a database — when you deposit, the asset is no longer yours; it’s an “IOU” (debt) the exchange owes you. Pros: deep liquidity, friendly UI, full derivatives. Cons: if the exchange blows up, your money is gone (FTX lesson).

• Binance — global #1, ~$30-50B/day, 500+ tokens
• Coinbase — US compliance leader, institutional channel Coinbase Prime
• OKX / Bybit — strong derivatives, primary entry for Asian capital
• Kraken — long-running US exchange, conservative, slow listings

DEX · decentralized exchanges

Smart-contract auto-matching + no account + self-custody of assets. The largest category is AMM (Automated Market Maker) — Uniswap, Curve, PancakeSwap; the rising category is order-book DEXes — dYdX, Hyperliquid. Pros: censorship-resistant, can’t be frozen, lots of long-tail assets. Cons: gas fees, slippage, sandwich-attack risk.

• Uniswap — Ethereum AMM king, ~$1-3B/day
• Curve — stablecoin / homogeneous-asset swap specialist
• PancakeSwap — Uniswap clone on BSC, popular in Asia
• Hyperliquid — on-chain perp dark horse, exploding from 2024

OTC · off-exchange

Large trades (single ticket > $100K) go via OTC desks rather than exchanges to avoid impact. Cumberland (DRW subsidiary), Galaxy Digital, Genesis (bankrupt), Falcon X, B2C2, Wintermute all run OTC. The main battleground for institutions / miners / whales.

CEX vs DEX comparison

Not your keys, not your coins · Assets on a CEX legally aren’t yours.

FTX blew up in November 2022, instantly freezing $8B+ in user assets — most was eventually returned but it took 2 years. Mt. Gox went bust in 2014; users got money back 10 years later. CEX provides convenience but at its core hands the private keys to a third party. Cold wallets (Ledger, Trezor) are the only safe destination for large crypto holdings. Day-trading capital / DeFi operations can use hot wallets (MetaMask).

§03 · Major assets — BTC · ETH · Stable · Alt · Memecoin

Crypto has tens of thousands of tokens, broadly split into the 5 categories below. Their risk-return profiles are completely different; sort that out before sizing positions.

1. BTC · Bitcoin / digital gold. Released by Satoshi in 2009, fixed cap of 21M, PoW consensus. $2T cap, 55% of crypto. Narrative: “hard asset hedging fiat debasement.” Annualized RV 50-70%; long-run the steadiest crypto asset. Institutions buy via spot ETFs (IBIT, FBTC).
2. ETH · Ethereum / smart-contract platform. Live since 2015, switched to PoS in 2022. $500B cap, 14% of crypto. Narrative: “global settlement layer.” Stake to earn 3-4%. EIP-1559 burns part of gas → partially deflationary. RV 60-90%, higher β but higher α too.
3. Stablecoin. Tokens pegged to $1. USDT (Tether) $140B + USDC (Circle) $60B = 95%. The “USD checking account” of crypto — base currency for trading, DeFi, cross-border payments. See CRCL earnings analysis.
4. L1 / L2 chains. Non-Ethereum smart-contract L1s (SOL · AVAX · TON · APT · SUI) + Ethereum scaling solutions (Arbitrum · Optimism · Base · zkSync). Narrative: “Ethereum is too expensive and slow; we’re faster and cheaper.” Bigger swings than ETH.
5. DeFi / GameFi · application tokens. UNI (Uniswap), AAVE, CRV, LDO, ENA, etc. Essentially “stocks of DeFi protocols” — but the vast majority pay no cash dividend, only governance rights. Most are down 80-95% from 2021 peaks.
6. Memecoin. DOGE · SHIB · PEPE · BONK · WIF · TRUMP. No “use case” or cash flow, purely narrative / community driven. Can 10-100× short term, 90%+ go to zero within a year. The casino floor of crypto.

Market-cap / liquidity ladder

§04 · Spot / Perps / Options — spot · perp · options

Crypto derivatives volume is 4-5× spot — anyone serious about crypto trading needs to understand this layer.

1. Spot. Hold the coin directly. BTC ETFs (IBIT / FBTC) are “wrapped spot.” Simple, but no leverage, no shorting. The main institutional entry.
2. Perp · perpetual contract. Crypto-native product invented by BitMEX in 2016 — a futures contract with no expiry. Anchored to spot via the funding-rate mechanism. The main battleground of crypto trading, $40-80B/day. See §05.
3. Quarterly · quarterly futures. Traditional-style futures with explicit expiries (Mar / Jun / Sep / Dec). Simpler than perps; common for institutional cash-and-carry. OKX / Binance / Deribit all list them.
4. Options. Mostly on Deribit (80%+ liquidity). BTC / ETH option IV runs 50-90% long term. Like equity options, but the skew shape differs — right-side calls are also often expensive (rally expectations). See A guide to volatility trading.
5. Margin · margin spot. Borrow tokens to buy/sell. 10-20× common. Simpler than perps but margin management is finer. Binance / OKX / Kraken support it.
6. Inverse vs Linear · coin-margined vs USDT-margined. Inverse (coin-margined): use BTC as both margin and unit of account. Linear (USDT-margined): use USDT. Beginners should use USDT-margined — P&L is intuitive, no mental math.

Perp vs quarterly

§05 · Funding rate — the funding rate

Perpetuals have no expiry — so how does the price not drift from spot? The answer is the funding rate: every 8 hours, longs and shorts pay each other a fee, pulling the perp back to spot.

Mechanism in detail

Simplified formula: Funding = Premium + Interest (the actual formula is more complex, but premium is the core). Premium = (perp price − spot) / spot.

• Perp > spot (longs aggressive) → funding positive → longs pay shorts
• Perp < spot (shorts aggressive) → funding negative → shorts pay longs
• Settled every 8 hours (00:00, 08:00, 16:00 UTC)
• Typical range ±0.01% / 8h (annualized ±11%); bull-market tops can hit +0.1% / 8h (annualized 109%)

Reading market sentiment from funding

Funding-rate arb (basis trade)

Cash & Carry · crypto-native arb · Buy spot + sell perp = lock in funding rate.

When funding is positive (say annualized 20%): buy 1 BTC spot + sell 1 BTC perp. Price moves don’t affect P&L (long/short cancel); you only earn funding. This was the institutions’ favorite “risk-free” arb in the 2020-2022 bull, annualized 20-40%.

After 2022, with Jump, Jane Street and other traditional HFTs flooding in, funding compressed to 5-10%. At the 2024 cycle top it returned to 30-50%. When the market overheats, this is a stable “signal + arb” two-fer.

Note: this strategy on CEXes carries counterparty risk (if the exchange fails, the spot-ETF wrapper is a safer multi-leg, but most retail can’t go fully onshore). See §10.

§06 · On-chain — on-chain analytics

Stocks have P/E and cash flows; crypto assets barely have cash flow — but they have every transaction recorded on a public chain. That’s why on-chain analytics was born. Glassnode / Nansen / Arkham / Dune are the most-used tools.

1. MVRV · market value / realized value. Market cap divided by sum of all BTC at the price of their last move. > 3 → market overheated (top); < 1 → market in loss (bottom). Macro bull/bear signal for BTC.
2. NVT Ratio · network value / on-chain volume. Akin to a P/E. High NVT = price too high relative to network usage. Popularized by Willy Woo. But after the perp era, on-chain no longer represents all volume; the signal has weakened.
3. Exchange Inflow / Outflow. Users moving coins to exchanges → preparing to sell → bearish. Conversely, moving out → preparing to hold → bullish. Multi-month sustained exchange outflow = bullish (2023-2024).
4. Active Addresses. How many unique wallets send / receive transactions per day. BTC long-run 600K-1M / day, ETH 400-600K / day. Network-usage fundamentals.
5. SOPR · Spent Output Profit Ratio. Average profit/loss across all UTXOs that moved that day. SOPR < 1 → network-wide capitulation selling (bottom signal). Glassnode’s signature indicator.
6. Hash Rate / Hash Ribbon. Total BTC network hash rate. Hash rate falling → miner capitulation → historical bottoms. Hash Ribbon is a 30/60-day moving-average crossover signal.
7. Stablecoin Mcap. USDT + USDC + DAI total. Continued rise = new money waiting to enter (bullish dry powder). 2023-2024 stablecoins from $120B → $220B underpinned the BTC rally.
8. Realized Cap. Sum of all UTXOs valued at the price of last move. More stable than market cap; reflects “cost basis.” The denominator of MVRV.
9. Smart Money tags. Nansen / Arkham label top on-chain addresses (VC / fund / whale / smart money). Tracking their flows can front-run direction. But beware reverse manipulation by adversaries.

How to use on-chain data · It’s confirmation, not a single signal.

On-chain indicators are nearly all lagging and easily bypassed by perps / OTC (institutions now do 80% of trading off-chain on CEXes; that flow is invisible to on-chain). Use on-chain data as “long-term macro signal + extreme-threshold alert,” not as an intraday tool. Reading where MVRV is at tops and bottoms is ten times more useful than watching it move daily.

§07 · DeFi yield — AMM · lending · staking · restaking

DeFi (Decentralized Finance) was the new frontier that exploded in 2020 — moving traditional finance’s lending, market making, and derivatives on-chain, with all logic written as open-source smart contracts. Three main yield categories: AMM market making, lending interest, staking / restaking.

1. AMM market making (Liquidity Provision)

Deposit two tokens (e.g. ETH + USDC) at 50:50 value into a Uniswap pool — you become an LP (liquidity provider). Each swap through the pool earns you a proportional fee (0.05% / 0.30% / 1%).

The risk is “impermanent loss” (IL): if ETH rallies, the pool mechanism auto-decreases ETH and increases USDC → you “miss out” — that gap is IL. The bigger the price swing, the bigger the IL. Fee income must beat IL to be profitable.

• USDC / USDT pool (homogeneous): IL ≈ 0, APR 2-8%
• ETH / USDC pool (high vol): material IL, APR 10-30% but net of IL
• Uniswap V3 concentrated liquidity: provide liquidity within a chosen price band, higher APR but more management

2. Lending protocols

Aave, Compound, MakerDAO / Sky are the leaders. Deposit tokens into the contract to earn interest; or post collateral to borrow other tokens (a common play: post ETH, borrow USDC, take leverage).

• Stablecoin deposit APR: USDC ~ 3-8%, USDT ~ 4-8%
• ETH deposit APR: 0.5-2%
• LTV: ETH usually 75-83%, breach = forced liquidation
• Liquidation penalty: 5-13% (paid to the liquidator keeper)

Most-common user move: recursive borrowing → post ETH, borrow USDC → buy more ETH with USDC → post and borrow again. Effectively 3-5× long ETH. In the 2022 LUNA crash, this kind of recursive loan got mass-liquidated — direct cause of Celsius / Three Arrows blowups.

3. Staking · restaking

After ETH switched to PoS in 2022, locking ETH to participate in consensus earns “staking yield.” Native staking requires 32 ETH and running a node, so most people use liquid staking (LST):

• Lido (stETH) — largest LST, ~$30B TVL, APR 3-4%
• Rocket Pool (rETH) — decentralized LST, APR 3-3.5%
• Coinbase ETH (cbETH) — centralized LST, compliance-friendly

Since 2024, EigenLayer Restaking has driven a new concept — “re-staking” already-staked ETH to other protocols (AVS) for stacked yield. stETH + ezETH (Renzo) / rsETH (KelpDAO) total APR can reach 5-8%. But stacked risk also rises in step — smart-contract risk, AVS slashing risk, and liquidity risk all stack.

“High APY = free lunch” is wrong · 20%+ APR almost always hides a price.

2020-21 saw a flood of 1000%+ APY “yield farms” — essentially the project inflating token rewards (printing money for you) + token price dumping immediately. Farmers got 1000% APY in a token that fell 99% → net loss. Real, sustainable DeFi yield is generally in the 3-10% range. When you see 50%+ APY, default skepticism, check the source, watch TVL trends, audits, and the token-emission schedule.

§08 · MEV — Maximal Extractable Value

MEV (Maximal Extractable Value, formerly Miner Extractable Value) is the extra value block producers extract by choosing transaction order. In crypto-native finance, this is its own industry — annual scale of hundreds of millions to billions of dollars.

Three main MEV types

1. Sandwich attack. See your pending large buy of token X → front-run by buying first → your trade executes (price gets pushed up) → sell immediately (back-run). You eat the slippage.
2. Arbitrage. ETH = $3,000 on Uniswap, $3,010 on SushiSwap → atomic buy-and-sell in one transaction → harvest the $10 spread. “Benign MEV” with no victim.
3. Liquidation. Monitor collateralized positions on Aave / Compound; once they breach the liquidation line, trigger liquidation → collect 5-13% bonus.

Participant roles

• Searcher — writes bots to find MEV opportunities, builds bundles
• Builder — packages bundles into a full block, bids to validators
• Validator — picks the highest-paying builder block to include, takes a cut of MEV

Scale and impact

How to avoid getting sandwiched

• Set low slippage (0.5% rather than 3%)
• Use MEV-protected RPCs (Flashbots Protect, MEV Blocker)
• Route large orders via OTC desks, not Uniswap
• Use aggregators (1inch / CowSwap) — internal MEV-protected routing

MEV philosophy · It’s the price of public-chain “openness.”

Every pending tx is public in the mempool → anyone can see and front-run. That’s the cost of public-chain verifiability. Eliminating MEV entirely is nearly impossible; only privacy mechanisms (SUAVE, Penumbra) or encrypted mempools mitigate it. MEV is partly “malicious” and partly “benign” — arbitrage and liquidation actually make the market more efficient; only sandwich attacks are pure negative-sum.

§09 · Strategies — crypto trading strategies

The strategies below are ordered low-to-high by “barrier + capital.” Beginners start with 1-3; institutions run 5-10 simultaneously.

1. DCA BTC / ETH. Buy BTC / ETH spot in fixed amount monthly, hold in cold wallet. Best risk-adjusted strategy of the past 10 years. No research, no timing — fits 99% of retail.
2. Funding-rate arb · Cash & Carry. Buy spot + sell perp, lock funding rate. Annualized 5-30%. Requires CEX execution, carries counterparty risk. Most-used institutional strategy. See §05.
3. Cross-exchange arb · CEX price spread. Binance BTC = $50,000, Coinbase = $50,030 → simultaneous buy/sell. Annualized 100%+ in 2017-2020, compressed to 5-15% by 2024. Requires multi-exchange capital + low latency.
4. LST yield · Lido / Rocket Pool. Buy stETH / rETH and hold; APR 3-4%, liquid for DeFi stacking. 3-4% extra yield over plain ETH. Passive strategy.
5. Stablecoin lending · USDC lending. Deposit USDC into Aave / Compound for 3-8%. Low-risk, low-yield, but bears smart-contract risk + stablecoin de-peg risk. 5× a bank deposit.
6. LP market making · Uniswap V3. Provide liquidity in an ETH/USDC pool with a chosen price band. APR 10-50% (net of IL). Active management + IL math required. Mid-level DeFi strategy.
7. Trend / momentum · Crypto CTA. Use moving-average / breakout rules to follow BTC / ETH trends. Similar to traditional CTA but bigger swings. Moonshot / Goldfinch / SwissBorg-class funds run this.
8. On-chain stat-arb · mean-reversion across majors. Relative strength reversion among BTC / ETH / SOL → long-short hedge. Sharpe 1-2 but small capacity. Mainstream crypto-quant strategy.
9. Options strategies · Deribit Vol. BTC / ETH covered call / put selling / iron condor. BTC IV runs 50-90% long term, sell-vol carry is huge. But BTC moves are also huge — use carefully. See Volatility guide.
10. MEV / Searcher · on-chain market making / arb bot. Write bots to harvest on-chain MEV. Extremely high barrier (smart contracts + Solidity + real-time infra). Million-dollar income but fierce competition. See §08.

Strategy mix · by account size

§10 · Six risks — six unique risks

1. Exchange blowup (counterparty). Mt. Gox (2014, $450M), QuadrigaCX (2019), FTX (2022, $8B+), Celsius (2022), Voyager (2022), BlockFi (2022). Anything held on a CEX / CeFi can theoretically go to zero at any time. Cold wallets are the only safe destination for size.
2. Smart-contract bugs. Poly Network ($600M, 2021), Wormhole ($320M, 2022), Ronin ($625M, 2022), Curve ($70M, 2023). Audits don’t immunize. The first 3 months of TVL exposure on a new protocol = high risk.
3. Stablecoin de-peg. UST (LUNA, 2022) burned $40B to zero. USDC briefly de-pegged to $0.87 during the SVB collapse (2023-03). USDT has had repeated reserve-quality questions. “Stable” coins aren’t always stable — trust only 1-2 leaders (USDC, USDT) + diversify holdings.
4. Lost / stolen private keys. Estimated ~20% of total BTC supply (~4 million coins) is permanently lost (Satoshi, lost hard drives, etc.). Phishing attacks have cleaned countless wallets. “Writing the seed phrase on paper and photographing it to the cloud” is the most common form of suicide. Hardware wallet + multisig + offline backup are baseline.
5. Regulatory uncertainty. SEC vs. Coinbase / Ripple lawsuits, IRS crypto-tax policy, country-level “exchange bans,” differing stablecoin regimes. A strategy legal today may need re-evaluation tomorrow because of new rules. Always track jurisdictions.
6. Sudden liquidity evaporation. Small-coin tops → dump → during a 99% drop, the orderbook is empty → your stop fills at the floor. LUNA went from $80 to $0 in a week; many stops never had a chance to trigger. Big coins (BTC/ETH/SOL) have deep liquidity; small coins can’t be relied on.

The lesson FTX taught everyone · “Our assets are with a third party. They told us so.”

The week before FTX collapsed, SBF was still tweeting “Funds are SAFU.” Celsius’s CEO was doing marketing videos a week before its collapse. Never believe a CEX / CeFi / project’s “we’re safe” claim — they have no incentive to tell you the truth. The only judgment criteria: 1) Proof of Reserves (on-chain verifiable reserves); 2) Bitcoin / Ethereum mainnet withdrawals are unblocked; 3) fiat rails covered by regulators. Everything else is noise.

§11 · How to start — how to start in crypto

1. L1 · complete beginner. Start with BlackRock IBIT spot BTC ETF (any US equity account works). 1-3% allocation, DCA over a year. Don’t touch any alts / DeFi / derivatives. Goal: first feel “what holding crypto feels like” (can you sleep through a 50% drawdown?).
2. L2 · on-chain entry. Open Coinbase / Kraken account, buy small BTC + ETH, transfer to a Ledger / Trezor hardware wallet. Learn to use MetaMask with Uniswap, Aave. Use very small amounts ($100-500) to experience one on-chain swap and one DeFi deposit, learn gas / slippage / wallet mechanics.
3. L3 · derivatives entry. Open OKX / Bybit, try low-leverage (2-5×) perps. Study funding rate, try cash-and-carry arb. Position < 5% of total capital. Focus on margin management + liquidation mechanics.
4. L4 · systematic strategies. Use Hummingbot / 3Commas / your own Python to run 1-2 quant strategies (funding-rate arb + trend + LP market making). Build daily PnL + risk reports; run 6-12 months to see actual Sharpe. $50K-500K is the right size for this.
5. L5 · pro / startup. Build a team + multi-exchange prime brokerage + own on-chain infra (nodes / RPC / MEV). $1M+ to start → $10M to mature. You’ll likely lose 30%+ in some black swan (exchange fail / protocol hack); that’s the tuition.
6. Required reading / common resources.
   • Bitcoin Standard (Saifedean Ammous) — monetary philosophy
   • Mastering Ethereum (Andreas Antonopoulos) — technical primer
   • The Infinite Machine (Camila Russo) — ETH history
   • On-chain data: Glassnode / Nansen / DefiLlama / Dune
   • Live news: The Block · CoinDesk · Bankless · Laura Shin Podcast

One-line summary · Crypto is the youngest, wildest, and most humility-requiring market in finance.

It’s the highest-returning asset class of the past 15 years, and also the asset class that’s wiped the most people. Both are true. That means how big a position you can win isn’t decided by your insight, but by your position management + custody security + counterparty selection. Citadel’s Ken Griffin took 30 years to build a quant empire; FTX’s SBF took 30 months to wipe every user. Both stories ran simultaneously in 2022 in crypto. Always put “survive” before “earn”.