Side Lesson 09: Crypto in a Portfolio — Bitcoin, Ethereum, and the 1-5% Allocation Case

---

1. Why This Is Important

Bitcoin is sixteen years old. Over those sixteen years its price has gone from a fraction of a cent to north of one hundred thousand dollars, with two complete crashes of about minus-eighty-five per cent along the way. That is not a normal asset class life cycle. It is the life cycle of a brand-new monetary experiment running in real time, on a public ledger, in the open. Whether or not it is "real" is a store-of-value question (every store of value rests on belief), but the size of the market, the liquidity, and the fact that you can now buy it inside a brokerage account through an SEC-registered spot ETF means that ignoring it is no longer a neutral position — it is an active allocation decision.

Four reasons this lesson exists:

**The asset class is now investable from a regular brokerage

account, by name, without ever touching a private key.** The January 2024 spot Bitcoin ETF approvals (IBIT, FBTC, ARKB and eight others) and the July 2024 spot Ethereum ETF approvals (ETHA, FETH) are the single most important regulatory event in crypto's short history. The default for serious investors is US-listed names; spot ETFs collapse the crypto-as-asset-class question into the same wrapper as everything else in this tutorial. Before 2024, crypto required a separate exchange, a wallet, and a custody decision. After 2024, it is a ticker.

**Bitcoin's volatility is roughly seventy per cent annualised —

four times the S&P 500.** That single number is what makes sizing the position a math problem, not an opinion problem. You do not "buy crypto" the way you buy an index fund; you buy a tiny sliver of it, sized so that a minus-eighty-per-cent drawdown is a tolerable hit to the whole portfolio. The vol-tail-wags-the-dog frame is the cleanest one: a five-per-cent BTC sleeve at seventy-per- cent vol contributes about three-and-a-half percentage points to total portfolio vol, which is most of what a 60/40 portfolio's stocks contribute.

**The 2017 / 2021 / 2025 three-cycle pattern is now visible

enough to discuss honestly.** Each cycle ran about four years peak-to-peak, each blew off the top with leverage and altcoins, and each washed out in an eighty-per-cent drawdown that took eighteen-to-twenty-four months to recover from. The 2024–2025 cycle, the first one with US institutional money inside the spot ETFs, is the first cycle where the peak did not require an exchange leverage flush — it ended with a more orderly fade. Three cycles is not a regime the way the 1970s inflation episode is a regime, but it is enough to retire the "BTC has no track record" argument.

**The tax and product wrappers create real edges and real

traps.* The IRS treats crypto as property*, not as a security: that means there is no qualified dividend, no Section 1256 60/40 blended rate, no wash-sale rule (yet), and no automatic 1099-B cost-basis tracking on direct-held coins. The spot ETFs partially fix this — they trade like ETFs and cost-basis is tracked — but the underlying remains property. Tax via the wrapper is more of an edge here than in equities precisely because the tax code is still half-finished. Get the wrapper right and the net-of-tax outcome is a different asset class than getting it wrong.

The goal of this side lesson is to give you the half-dozen numbers and the one sizing rule that turn "should I own Bitcoin" from a philosophical argument into a sleeve-sized position in the four- tranche portfolio — small, deliberate, and in the right sub-account.

---

2. What You Need to Know

2.1 The Bitcoin Thesis in One Page — Twenty-One Million and Halvings

You can argue about Bitcoin's price for the next decade. You cannot argue about its supply schedule, because the supply schedule is the protocol. Two numbers carry the entire thesis:

• The maximum supply is 21,000,000 coins, ever. The cap is

enforced by every node on the network, by consensus. As of April 2026, roughly 19.85 million have been mined. The remaining ~1.15 million will trickle out, asymptotically, over the next hundred-plus years. There is no committee that can vote to raise the cap; doing so is a hard fork that the network would, by fifteen years of revealed preference, reject.

• Every four years the new-coin issuance halves. The block

reward started at 50 BTC per block in 2009, fell to 25 in November 2012, to 12.5 in July 2016, to 6.25 in May 2020, to 3.125 in April 2024, and will fall to 1.5625 in 2028. New supply is currently about 0.85% of the existing float per year — already lower than gold's roughly 1.5% per year mine-supply growth, and falling.

That is the entire "digital gold" thesis: a fixed-cap commodity-like asset whose new-supply rate is mechanically engineered to fall below gold's, with a globally-distributed validation network and no central issuer. Every other crypto debate is downstream of whether you accept that thesis or not. The honest framing is that this is a belief asset, the same as gold. The only question is the durability of the belief, and the only available evidence is fifteen years of price discovery, three completed boom-bust cycles, and a $2-trillion-ish market capitalisation that the legal and institutional system has chosen not to suppress.

The same framework applies, with a discount, to Ethereum (ETH). Ethereum is not fixed-supply — its issuance rate is variable and was even briefly negative after the 2022 Merge — but it has the second- largest network effect, the second-largest validator set, and the only spot ETF approval outside Bitcoin. Treat ETH as a smaller, more volatile, more "tech-flavoured" satellite to a BTC core if you hold either at all. Everything else (the so-called "altcoins") has not demonstrated equivalent durability and is, for the purposes of a portfolio lesson, noise.

2.2 The 2017 / 2021 / 2025 Cycles — Pattern, Not Coincidence

Bitcoin's price history is best read as three completed four-year cycles, each anchored by a halving, each ending in a roughly minus- eighty-per-cent drawdown:

• Cycle 1 (2013-2015) and the warm-up. Price ran from about $13

in January 2013 to $1,150 in late 2013, then bled to $200 by early 2015 — a minus-eighty-three-per-cent drawdown. This was the Mt. Gox era, with most of the float on a single Tokyo exchange that promptly collapsed.

• Cycle 2 (2017-2018). Price ran from $1,000 in January 2017 to

$19,800 in December 2017 (the original retail mania), then bled to $3,200 by December 2018 — a minus-eighty-four-per-cent drawdown. Futures launched at the top (CBOE/CME, December 2017), which gave institutions their first short-able instrument; that is not a coincidence with the timing of the top.

• Cycle 3 (2020-2022). Price ran from $9,000 in early 2020 to

$69,000 in November 2021, then bled to $15,500 by November 2022 amid the FTX/Three Arrows/Luna implosions — a minus-seventy-seven- per-cent drawdown. This was the cycle that broke crypto's worst intermediaries (centralised lenders, leveraged "yield" platforms) and pushed the survivors toward the regulated wrapper.

• Cycle 4 (2024-2025). Price ran from $42,000 at the January

2024 spot ETF launch to roughly $145,000 in late 2025, then faded to about $112,000 by April 2026 — a milder minus-twenty-three-per- cent drawdown so far, the first cycle with material US institutional flow inside the regulated wrapper. Whether this cycle ends with a conventional minus-eighty-per-cent flush or with a structurally shallower drawdown is the live question of 2026.

Three things to take from the pattern. First, the peaks are roughly four years apart, anchored on the halving. That regularity will eventually break, but it has not yet, and the calendar is information. Second, the drawdowns are catastrophic. Minus eighty per cent is the modal outcome at the end of each cycle, not the tail. A retail investor who cannot stomach watching a position lose four- fifths of its value, on schedule, in a window of twelve to eighteen months, should not own this asset class. Third, each cycle's recovery has taken the price to a higher high than the previous cycle's peak. A buyer who entered at the absolute worst tick of the 2017 top (December 2017, $19,800) was underwater for three years and then doubled. A buyer at the 2021 top ($69,000) was underwater for about three years and then doubled. The pattern is brutal but ergodic — if the network thesis continues to hold.

2.3 The Volatility and the Drawdowns Are the Whole Sizing Problem

The number that decides whether crypto belongs in a portfolio at all is its annualised volatility, and it is unforgiving:

• Bitcoin daily volatility, fifteen-year average: roughly 70%

annualised. For comparison: SPX is 16-19%; the Nasdaq-100 is 22-26%; gold is 15%; long Treasuries (TLT) is 14%; meme stocks at their worst are 60-80%. BTC sits at the top of the volatility ranking of any liquid asset class.

• Bitcoin maximum drawdown, fifteen-year history: -85% (2014)

and -84% (2018) and -77% (2022). Three drawdowns deeper than any single drawdown the S&P 500 has produced since 1932.

• Ethereum daily volatility, eight-year average: roughly 90%

annualised, with drawdowns deeper than Bitcoin's at every cycle trough.

The math you cannot escape: a 5% BTC sleeve at 70% vol, held alongside a 95% portfolio of stocks-and-bonds at, say, 11% vol, contributes roughly $\sqrt{(0.95 \times 0.11)^2 + (0.05 \times 0.70)^2}$ ≈ 11.5% portfolio vol — about 50 basis points above the no-crypto baseline. The same arithmetic at a 10% BTC sleeve crosses 13% vol — you have meaningfully changed the risk profile of the whole portfolio to make room for the sleeve. *Above ten per cent, the BTC position is no longer a satellite — it is a co-driver of the portfolio's risk*, which violates the four-tranche logic for almost every investor's risk budget.

This is the entire reason the canonical retail allocation is 1-5% of investable assets, with 3% the centre of gravity. It is not a guess; it is the size at which BTC adds expected return without swamping the rest of the portfolio's variance contribution. In barbell terms: the BTC sleeve sits on the right edge of the barbell, with T-bills on the left, and the index-fund middle absorbs neither's volatility.

2.4 The Kelly Logic Behind 1-5%

The 1-5% range is not arbitrary. Run a fractional-Kelly calculation and the same answer falls out. Take Bitcoin's historical Sharpe (roughly 0.7-0.9 over the full sample, lower over rolling five-year windows), assume forward expected return of 15-25% per year (lower than the realised, which is the standard adjustment for an asset that has tripled its market cap during your sample), and assume vol holds at 60-70%. Full-Kelly says size the sleeve at *expected excess return divided by variance*: about (0.20 - 0.04) / 0.65² ≈ 38%.

Nobody runs full-Kelly on a single asset. The standard discount is quarter-Kelly to one-eighth-Kelly, which lands at roughly 5% to 10% of the portfolio. Apply a second discount for parameter uncertainty (your expected-return number is not a fact, it is a guess on a fifteen-year sample), and the practical sleeve falls to 1-5%. That is the same number the institutional consultants (BlackRock, Fidelity) landed on independently, and the same number a fractional-Kelly calculation produces. It is not three different answers; it is one answer arrived at three ways.

The other discipline is rebalancing. Without rebalancing, a 5% BTC sleeve that triples in a cycle becomes a 13% sleeve at the top — exactly when the four-cycle history says you should be trimming. With quarterly or annual rebalancing back to target, you mechanically sell into the cycle peaks and buy into the troughs, and the sleeve stays at the size you actually budgeted for. Most of the long-term contribution of a 1-5% BTC sleeve to a portfolio's CAGR comes from the rebalancing flow, not from buy-and-hold.

2.5 Spot ETFs vs Self-Custody — IBIT, FBTC, ETHA, and the Wrapper Choice

Before January 2024, owning Bitcoin through a US brokerage required either GBTC (a closed-end fund that traded at premiums/discounts of ±40%) or a direct holding through an exchange (custody risk, key management). The spot ETF approval changed that overnight.

The four wrappers worth knowing:

• IBIT (iShares Bitcoin Trust) — BlackRock's fund. Largest by

AUM (about $58 billion as of April 2026). Expense ratio 0.25% (initial 12-month waiver expired). Custodian: Coinbase. The de facto institutional default.

• FBTC (Fidelity Wise Origin Bitcoin Fund) — Fidelity's fund.

Self-custody (Fidelity Digital Assets, in-house). About $20 billion AUM. ER 0.25%. The choice for investors who already have a Fidelity-house preference.

• ETHA (iShares Ethereum Trust) — BlackRock's spot Ethereum ETF,

approved July 2024. About $5 billion AUM, ER 0.25%. The largest of the spot ETH funds.

• GBTC (Grayscale Bitcoin Trust) — the legacy product, converted

to ETF in January 2024. ER is 1.50% — six times the iShares fee. The only reason to hold it is if you have an embedded long-term capital gain you do not want to crystallise. Otherwise, switch to IBIT.

The wrapper choice matters for three reasons. Tax treatment: all four are 1099-B reporting, cost-basis tracked, and trade like an ETF in any taxable or retirement account — none of the headache of direct-held coin tax accounting. Custody: you outsource the key- management problem to a regulated custodian; for 99% of investors this is the right answer and replaces a meaningful tail risk (self-custody loss, exchange failure) with a small ongoing fee. Wrapper purity: spot ETFs hold the underlying coin one-to-one, unlike futures-based BITO (the 2021 product) which suffered a roughly 8-10% per year contango drag. Spot ETFs are the right wrapper for buy-and-hold; futures wrappers are for traders.

The case for direct self-custody is narrow: large positions (>$1 million) where the 0.25% fee is meaningful, ideological alignment with the original Bitcoin thesis ("not your keys, not your coins"), or use cases that require on-chain transactions (DeFi, payments). For the 1-5% sleeve in a tax-advantaged retirement account, the spot ETF is the unambiguously correct wrapper. The US-listed wrapper is the default.

2.6 Tax Treatment — Property, Not Security

The IRS classifies cryptocurrency as property (Notice 2014-21, unchanged through 2026). The consequences:

• No qualified dividend rate. Crypto pays no dividend in the

first place, so this only matters for staking rewards (taxed as ordinary income at receipt) and for miners (ordinary income on the block reward at fair value).

• No Section 1256 60/40 split. Unlike SPX options or futures,

crypto gains are pure short-term-or-long-term capital gains based on the standard one-year holding period.

• No wash-sale rule (yet). Section 1091 covers "stocks or

securities." The IRS has, as of April 2026, declined to extend it to crypto. That means you can sell BTC at a loss in December for tax-loss-harvesting and re-buy it on January 2nd — a play that is illegal in equities. Multiple legislative proposals to close this loophole have been introduced; do not assume it persists.

• Cost-basis is your problem if you hold direct coins. Direct

exchanges issue 1099-MISC for staking income but not always 1099-B for trades; the IRS will eventually mandate cost-basis reporting (the 2025 Form 1099-DA) but until then the burden is on you. Spot ETFs avoid this entirely — they are 1099-B-reported securities inside the ETF wrapper.

• No automatic IRA eligibility for direct coins. Most IRA

custodians do not allow direct cryptocurrency holdings. *Spot ETFs are eligible* in any standard brokerage IRA. Location matters: hold the BTC sleeve inside a Roth IRA where any future gains are tax-free, given the volatility profile makes the asset a particularly clean Roth candidate.

The single most actionable tax insight: **the wash-sale-rule absence makes crypto the cleanest tax-loss-harvesting asset in the US tax code right now.** Sell at any loss; rebuy immediately; realise the loss against ordinary income up to $3,000 per year and unlimited offset against capital gains. That advantage will not last forever. Use it while it does.

2.7 What Crypto Is Not — No Cash Flow, No Floor, No Yield

The honest catalogue of what crypto cannot do for a portfolio:

• There is no intrinsic floor. Stocks have earnings, real

estate has rents, bonds have coupons and a maturity payment. Bitcoin has none of these. The price is supported entirely by belief, and a sufficiently coordinated belief shift could in principle send it to zero. The fifteen-year out-of-sample test is encouraging but not exhaustive.

• There is no qualified dividend or coupon income. A retiree who

needs cash flow gets none from BTC directly. The income solution is not "BTC pays a yield"; it is "trim the BTC sleeve back to target on rebalance and the trim cash flows the retirement budget."

• There is no obvious correlation hedge. BTC was supposed to be

uncorrelated. In the 2022 risk-off, BTC fell -65% as the S&P fell -25% and TLT fell -31%; the correlation was high and positive. Recent samples (2023-2025) have shown lower correlation (around 0.3-0.4), but BTC has not earned a long-volatility role the way trend-following or long-vol options have. Treat it as a high-vol return enhancer, not a diversifier.

• There is no consumer-protection backstop. Direct-held crypto

is not SIPC-insured. Stablecoins are not FDIC-insured. Self-custody loss is permanent. Spot ETFs are SIPC-insured at the brokerage level (against broker failure, not against the underlying coin going to zero).

The right mental model is: BTC is a high-volatility, high-expected- return return asset. It belongs in the Stores-of-Value tranche at 1-5% portfolio weight, in a tax-advantaged wrapper when possible, sized for a tolerable drawdown contribution, and rebalanced on a fixed schedule. It is not a hedge. It is not a cash-flow source. It is not a savings account. It is, on the fifteen-year evidence, an investable asset class — and the position size that respects that evidence is small.

---

3. Common Misconceptions

"Bitcoin is a scam / has no value." Bitcoin has a $2-trillion

market cap, fifteen years of continuous price discovery, and is held by US-listed ETFs custodied at SEC-registered firms. The "scam" framing was tenable in 2014, less so in 2018, and not at all in 2026. It is a belief asset — like gold and fiat — and the only honest question is the durability of the belief.

"Bitcoin is a hedge against inflation / the dollar." The

2022 evidence rejects this directly. CPI hit 9.1%, the dollar strengthened (DXY +8%), and BTC fell 65%. Whatever BTC is, it is not a mechanical inflation or dollar hedge. Belief asset, not structural hedge.

"You should put 20-50% of your portfolio in crypto." No.

At 70% vol, a 20% sleeve makes BTC the dominant contributor to portfolio variance. The 1-5% range is what the math (and the barbell logic) supports. Larger positions are concentrated bets, not diversified holdings.

"You need to buy direct coins to be a real investor." For

the 1-5% sleeve, the spot ETF (IBIT, FBTC, ETHA) is the better wrapper: lower friction, cost-basis tracked, IRA-eligible, no self-custody tail risk. Direct coin ownership is for amounts where the 0.25% fee is material or for use cases that need on-chain transactions.

"Crypto is uncorrelated with stocks." It was advertised that

way; the evidence is mixed at best. In 2020-2022 the rolling correlation with the Nasdaq ran at 0.5-0.7. In 2023-2025 it has averaged 0.3-0.4. Some diversification benefit, far less than the marketing suggests.

"Ethereum is just a smaller Bitcoin." No. Ethereum has a

variable issuance schedule, a fundamentally different security model (proof-of-stake since 2022), and a meaningfully different thesis (it is a smart-contract platform, not a fixed-supply commodity). It is more volatile, has had deeper drawdowns, and has not demonstrated Bitcoin's network durability. Treat it as a smaller satellite, not as a substitute.

"Halvings always cause new highs." They have, three times,

roughly twelve to eighteen months after the halving event. That is not a guarantee; it is a sample of three. Each cycle has had a different driver (retail mania 2017, institutional + COVID 2021, spot ETF 2024). The next one may not.

**"There is no wash-sale rule, so I can game the system

forever."** The IRS has signalled multiple times that it intends to extend §1091 to digital assets. Several legislative proposals are live in 2026. Use the loophole while it exists; do not build a long-term plan around it.

"Bitcoin's volatility will compress as it matures." It has,

from 100%+ in the early years to 60-70% recently. That is still four times equity vol. Expecting it to converge to equity-like vol in a five- or ten-year window is a faith statement, not an evidence-based one. Size it for the vol it has, not the vol you wish it had.

"Stablecoins are safe cash." Some are; some are not. USDC

(Circle) and USDT (Tether) carry counterparty and reserve-quality risk that bears no resemblance to FDIC-insured cash. The 2023 USDC depeg (Silicon Valley Bank exposure) is the canonical cautionary example. Treat stablecoins as a payment rail, not as a savings account.

---

4. Q&A

**Q: I am thirty-five with a 401(k) and a Roth IRA. Should I own Bitcoin?** A: A 1-5% sleeve, sized to your overall risk budget, held in the Roth IRA via IBIT or FBTC, rebalanced annually back to target. The Roth-and-spot-ETF combination means any future gain is tax-free, and the spot ETF is custodied by a regulated firm. Start at 1%, see how you feel through one cycle, scale to 3-5% if you can sit through a minus-eighty-per-cent drawdown without selling.

Q: Why a Roth IRA specifically, and not a taxable brokerage? A: Location matters. BTC is the highest-expected-return, highest-volatility asset most retail investors have access to. In a Roth, the upside is tax-free; in a taxable account, you owe long- term capital gains on the eventual exit, and the no-wash-sale-rule benefit only matters in the taxable account. The optimal split: hold the core BTC sleeve in the Roth, hold a smaller trading sleeve in the taxable account where you can harvest the no-wash-sale-rule advantage on each cycle's drawdown.

Q: IBIT vs FBTC — does it matter? A: Functionally, no. Both are spot Bitcoin ETFs at 0.25% expense ratio with regulated custody. IBIT has more AUM and tighter bid-ask spreads (a fraction of a basis point); FBTC uses Fidelity's in-house custody if you prefer that. For a buy-and-hold Roth IRA position, either works. Avoid GBTC unless you have an embedded capital gain — its 1.50% ER is six times the iShares fee.

Q: Should I dollar-cost-average into BTC or buy a lump? A: The Vanguard lump-vs-DCA evidence (Side 5) generalises — lump wins about two-thirds of the time on equities. Crypto's volatility makes the sample less stable, but the same principle applies: if you have the cash and you have decided on the allocation, deploy it. The exception is if your conviction is genuinely lower; in that case DCA across six to twelve months is a behavioural tool that prevents you from front-loading regret if the drawdown comes early.

Q: What about Ethereum? Bitcoin only? Both? A: For the first 1-3% of allocation, Bitcoin only — it has the longest track record, the cleanest thesis, the deepest liquidity, and the largest spot-ETF wrapper. If you want to push to 5%, add Ethereum via ETHA at roughly a one-quarter to one-third weight of the BTC sleeve. So a 5% crypto sleeve might be 4% BTC / 1% ETH. Beyond ETH, nothing else has demonstrated portfolio-grade durability.

**Q: My friend has 30% of his net worth in BTC and is up 5x. Should I do that?** A: Survivorship bias. Your friend who is up 5x is the one talking; the friends who sized similarly in 2017 or 2021 and got cut to a fifth are not posting. The 1-5% rule is sized for the *full distribution* of outcomes, not for the lucky tail. Alpha is rare, and concentrated alpha in a single high-vol asset is also a concentrated way to lose.

Q: What signals would tell me the cycle has topped? A: The four-cycle pattern suggests: (a) parabolic price action with double-digit weekly moves, (b) altcoin season — non-BTC tokens massively outperforming BTC for sixty to ninety days, (c) leverage in the futures market (open interest >$30B, funding rates persistently above 0.05% per eight hours), and (d) retail-survey indicators (Coinbase in the App Store top 10). All four were present at the 2017 and 2021 peaks. None of them are precise enough to time, but the combination is a credible "trim back to target" signal.

Q: What would change my mind on holding any BTC at all? A: Three regime breaks would close the case: (1) a successful fifty-one-per-cent attack on the network or a major protocol failure; (2) a coordinated US/EU/UK ban on regulated crypto wrappers; (3) a ten-year drawdown from which the price does not recover (i.e. the network effect breaks). None of these is impossible; none is currently indicated. Until one of them happens, the 1-5% sleeve is a reasonable exposure for an investor who has already maxed out their 401(k) and built an emergency fund.

Q: Is "Bitcoin is digital gold" a real thesis or a marketing line? A: Both. The thesis — fixed-supply, network-effect-protected, non-sovereign store of value — is real and has held up over fifteen years. The marketing line conflates the thesis with a reliable inflation hedge, which is not what gold has been either (see Side 6). Treat both gold and BTC as belief assets in the Stores-of-Value tranche, sized to the risk budget, and do not expect either to do the work of TIPS during an actual inflation regime.

**Q: How does a futures-based product like BITO compare to spot ETFs like IBIT?** A: BITO holds CME Bitcoin futures, which are typically in contango. The roll cost — selling expiring near-month futures and buying further- out futures at a higher price — has run 8-12% per year on average, which mostly destroys the long-term return. BITO is an acceptable tactical vehicle (briefly long for a few weeks); for buy-and-hold, spot ETFs (IBIT, FBTC) are the right wrapper.

Q: What is the interactive at the bottom of this lesson useful for? A: For sizing the position before you buy it, not after. Move the "BTC % of portfolio" slider from 0 to 10. Watch what happens to the expected return, the volatility, the Sharpe ratio, and — most importantly — the maximum drawdown. The sweet spot for most retail investors is 2-4%; above 5%, the BTC sleeve dominates the risk; below 1%, it does not move the needle. The lab is built to make that trade-off legible at the scale at which it actually compounds.