SatoshiMacro

For Australian residents holding crypto as a long-horizon allocation, the tax asymmetry between personal-name and SMSF structures is the single largest after-tax-return decision in the portfolio. The gap is 13.5 percentage points on every realised long-term gain at the 47% top marginal rate. Almost nobody runs the maths properly before setting up. The piece below walks through the six-variable framework I actually use on personal capital, the three balance-tier case studies that determine whether SMSF crypto pencils out, and the honest limitations that the SMSF advisory industry tends to underweight in pitches.
## The asymmetry, in one paragraph
Inside an SMSF in accumulation phase, capital gains on assets held longer than 12 months get the 1/3 CGT discount on top of the 15% concessional rate. Effective long-term CGT inside super equals 15% multiplied by two-thirds, which equals 10%. Compare this to personal-name disposal at the 47% top marginal rate after the 50% personal CGT discount, which equals 23.5% effective. The gap is 13.5 percentage points. On a single AUD 60,000 realised long-term gain, the personal-name structure pays AUD 14,100 in tax versus AUD 6,030 in the SMSF accumulation structure. Difference equals AUD 8,070 per AUD 60k gain. Compounded across a 10-year holding horizon with multiple cycle exits, the asymmetry produces six-figure differences in net retirement wealth.
## The six-variable framework before setting up
The decision to run an SMSF for crypto allocation is not binary on the tax asymmetry alone. Six variables determine whether the structure pencils out for any given AU resident.
Variable 1: Fund balance threshold. SMSF establishment costs run AUD 1,500 to 3,000 one-off. Annual compliance plus audit runs AUD 1,500 to 4,000. The fixed cost base means SMSF crypto allocation pencils out above approximately AUD 200,000 in total fund balance. Below that the admin drag eats the tax-discount alpha. A low-cost industry super fund with the same growth allocation usually beats the SMSF on net after-tax return below the AUD 200k threshold.
Variable 2: Accumulation phase versus pension phase. The 10% effective CGT rate inside super applies during accumulation. Pension phase produces 0% personal CGT on retirement income from account-based pensions, which can be better than the 10% SMSF accumulation rate for trustees already at retirement age. The SMSF tax advantage narrows materially at pension phase. Crypto allocation inside an SMSF makes most sense for trustees with 10 to 25 years until retirement, less sense for trustees within 5 years of preservation age.
Variable 3: Division 296 unrealised-gains tax. The 2026 Federal Budget announced a Division 296 30% extra tax on the proportion of total super earnings sitting above the AUD 3 million Total Super Balance threshold. Crucially the legislation as drafted captures earnings including unrealised mark-to-market movement, not just realised gains. For high-volatility growth assets like crypto inside super at high-balance accounts, this introduces a new drag that did not exist pre-2027. The legislation has not yet passed in its current form as of mid-May 2026. Anyone making structural decisions on this should reconfirm the current draft before locking anything in.
Variable 4: Custody documentation. The ATO requires SMSF crypto holdings to be held in a separate legal-entity wallet with documented ownership trail. Commingling SMSF crypto with personal crypto is a regulatory breach that triggers fund disqualification at audit. Cold-storage wallets need to be registered to the SMSF trustee entity, not the personal name of any individual trustee. AUSTRAC-registered AU exchanges (CoinSpot, Independent Reserve, Cointree, Digital Surge, Swyftx, BTC Markets) all support SMSF-entity accounts with separate KYC; non-AU exchanges generally do not.
Variable 5: In-house asset rules and sole purpose test. Section 71 of the SIS Act limits in-house assets to 5% of fund value. The sole purpose test under section 62 requires the fund to be maintained solely for retirement benefits. Crypto allocation passes both tests when held as a normal investment. The tests fail when the SMSF trustee uses crypto for personal trading, lending to related parties, or any non-arms-length transaction. Audit reviewers check this carefully.
Variable 6: Contribution-splitting strategy. The Division 296 AUD 3 million threshold is per-member, not per-fund. Two spouses each at AUD 2.5 million TSB pay zero Division 296. One spouse at AUD 5 million and one at AUD 0 pays Division 296 on 40% of the high-balance spouses earnings. Contribution-splitting between spouses while still in working years is the standard structural response to Div 296 implementation post-2027.
## Three case studies by fund balance tier
Case 1: AUD 200,000 fund balance, 60% growth allocation including 10% crypto.
Crypto holding equals AUD 20,000. At 12% annualised return compounded, the AUD 20k crypto position grows to AUD 22,400 in year one, AUD 25,088 in year two. The annual SMSF admin cost of AUD 2,500 (middle of the AUD 1,500 to 4,000 range) drags the after-cost return. Verdict: SMSF marginal at this balance tier. The tax asymmetry produces approximately AUD 320 of savings per AUD 2,400 unrealised gain when realised after 12 months, versus AUD 2,500 in admin cost. Net negative for the first 3 to 5 years until the compounding catches up to the fixed cost drag.
Case 2: AUD 1,000,000 fund balance, 60% growth allocation including 10% crypto.
Crypto holding equals AUD 100,000. At 12% annualised return, the AUD 100k position grows to AUD 112,000 year one. The AUD 2,500 admin cost is now 0.25% of fund balance, which is below comparable retail super fund admin costs. Verdict: SMSF clearly wins at this balance tier. The 13.5 percentage point tax-asymmetry advantage on a single AUD 60k realised long-term crypto gain produces AUD 8,070 of after-tax savings, comfortably exceeding the AUD 2,500 annual admin cost. Net positive from year one.
Case 3: AUD 5,000,000 fund balance, 60% growth allocation including 10% crypto.
Crypto holding equals AUD 500,000. The tax-asymmetry advantage is large in nominal terms but Division 296 now applies. With TSB at AUD 5 million, the share of earnings above AUD 3 million is 40%. Hypothetical year where the AUD 500k crypto position produces AUD 200k realised gains plus AUD 300k unrealised mark-to-market equals AUD 500k total fund earnings. Normal-regime tax on 60% of earnings equals AUD 39k. Division 296 tax on 40% of earnings at 30% extra equals AUD 60k. Combined tax bill AUD 99k. Compare to personal-name running the same scenario at the 47% top marginal rate with the 50% CGT discount on the realised portion: AUD 47k on the realised plus AUD 141k on the unrealised gains classified as ordinary income equals AUD 188k. SMSF still wins by AUD 89k on this scenario but the asymmetry has narrowed from 13.5pp to roughly 9pp. Spouse-splitting becomes the standard structural response above the AUD 3m threshold.
## Honest limitations the SMSF advisory industry under-discloses
Three limitations worth weighing before any SMSF crypto allocation decision.
Custody complexity is not just paperwork. The cold-storage wallet registered to the SMSF trustee entity needs an inheritance plan. If the trustee dies, the next-of-kin or executor needs the seed phrase, the legal authority to access the fund, and the operational knowledge to move the holdings safely. Most SMSF estate plans I have reviewed do not handle this properly. The ATO requires the holdings to remain within the fund and not be distributed to personal names except via the proper benefit-payment mechanism.
Time-horizon binding. SMSF crypto allocation is essentially locked until preservation age (55 to 60 depending on birth year). Crypto cycles run roughly 4 years. A trustee at age 45 holding 10% crypto in an SMSF cannot rebalance freely if a cycle-top exit signal fires and they want to cash out personally. The structural lock matters for risk tolerance.
Division 296 unrealised-MTM drag at high balances. The example in Case 3 showed Div 296 catching unrealised gains. For a volatile asset that goes up AUD 400k mark-to-market in one year then down AUD 400k the next, Div 296 triggers in year 1 with no symmetric refund in year 2. The legislation has a carry-forward loss mechanic but no cash refund. For trustees approaching the AUD 3m TSB threshold, this creates a structural disincentive to hold high-volatility assets inside super.
## EOFY 2026 timing
For trustees considering SMSF setup in time for the 2026-2027 financial year, the practical timeline is establishment by July 2026 to capture a full year of accumulation-phase tax treatment for FY 2026-27. AUD 1,500 to 3,000 establishment cost plus AUD 1,500 to 4,000 annual compliance is the budget. The 35-day window from EOFY 2026 to end-of-July is the standard onboarding window if working with an SMSF administrator like Heffron, BGL, or SuperConcepts.
For trustees already running SMSFs and considering crypto allocation for the 2026-27 financial year, the optimal entry is post-EOFY with a 12-month CGT-discount-eligibility timeline targeting the next cycle distribution window. Front-load the entry timing so that 12 months later coincides with the expected mid-to-late cycle distribution, not the cycle bottom. This is the single highest-EV improvement I made to my own cycle-exit framework after 2018.
## Calculator and reference
The SMSF CGT calculator at satoshimacro.com handles the 1/3 discount automatically, the Division 296 calculation for high-balance accounts, the pension-phase 0% case, and the spouse-splitting structural breakdown. Free, AUD-native, ATO-aligned methodology.
Full crypto tax framework for AU residents covering CGT, ordinary income, business-of-trading classification under TR 97/11, identification methods, carry-forward losses, and the 2026 Budget changes:
https://satoshimacro.com/crypto/crypto-tax-australia/
Disclosure: I built and maintain SatoshiMacro. The model is free and ad-supported (broker affiliate links on the main site, not in this Article). This Article is editorial, not financial advice. SMSF setup involves regulatory and structural decisions that warrant professional accounting and legal advice specific to individual circumstances.
#SatoshiMacro #SMSF #CryptoTax #Bitcoin

$BTC | MVRV Z-Score: Um Passeio pelo Indicador On-Chain Mais Mal Utilizado
A primeira vez que vi o MVRV Z-Score em uma mesa de trading institucional, estava errado, e a segunda vez estava correto, mas de uma forma que não era útil. Essa tensão - útil em retrospectiva, frustrante em tempo real - é o ponto de partida certo para entender o que esse indicador realmente faz.
A formula padrão, frequentemente chamada de Z-Score Awe & Wonder, em homenagem aos analistas que a popularizaram no Glassnode, é bem simples:

Most Bitcoin cycle models on the retail internet are built by anonymous Twitter accounts or crypto-native creators with no exposure to traditional multi-factor risk management. I built mine after 10 years running allocated institutional capital at a Sydney proprietary trading firm, applying the same risk framework I used for institutional books to the specific question of where Bitcoin sits in its cycle.
The short answer: no single indicator survives all three Bitcoin cycles. The longer answer is a 48-signal confluence framework that I will walk through in this article.
If you would rather just see the live model, it is free and unlocked at the link at the end of this piece. If you want to understand the methodology before you trust the number, read on.
The problem with single-indicator cycle models
The Bitcoin retail analysis community has a habit of falling in love with whichever single indicator was last to call the cycle top. The list is long.
Mayer Multiple alone: Bitcoin's price-to-200-day-MA ratio crossing 2.4 historically marked cycle tops. It worked beautifully in 2013 and 2017. It crossed 2.4 again in early 2019 during the dead-cat bounce and the local "cycle top" call was a complete miss.
Pi Cycle Top alone: The 111-day-MA over 350-day-MA-times-2 crossover marked the 2013-04 top, the 2013-12 top, the 2017-12 top, and the 2021-04 top (early in the cycle second leg). It then missed the actual 2021-11 cycle high. It also flashed crosses during deep bull-cycle pullbacks where it was not a top at all.
Hash Ribbons alone: The 30-day over 60-day hashrate moving-average crossover marked the 2018-12 bottom and 2020-03 bottom. The 2022-06 false-positive cross during the LUNA collapse came nowhere near the actual cycle bottom, which arrived 5 months later in 2022-11.
MVRV Z-Score alone: Excellent at marking cycle bottoms when the score went negative. Less reliable at cycle tops, where the threshold has drifted lower each cycle (8+ in 2013, 7+ in 2017, around 5 in 2021).
The pattern across all of these: each indicator captures one dimension of cycle position such as valuation, miner stress, or on-chain profitability. But Bitcoin cycles are driven by the simultaneous alignment of multiple dimensions. A model that watches only one dimension misses the other five.
The institutional risk-framework approach
In institutional trading you do not build a position based on one signal. You build a multi-factor risk model where each factor captures a different dimension of the trade, you weight the factors by historical reliability, you backtest the composite against known historical events, and you ship the model with explicit calibration anchors so anyone can audit whether you are cherry-picking.
That is exactly what the SatoshiMacro Model is, applied to the question of Bitcoin cycle position.
The six tiers
Six weighted tiers, chosen because each captures a meaningfully different dimension of cycle behavior. Weights reflect the historical reliability of each dimension at calling cycle inflections, not arbitrary preference.
Tier 1: Cycle Timing and Mass Psychology (30% weight). The structural foundation. Captures where Bitcoin sits in its 4-year halving cycle and its long-term power-law trajectory. Indicators include months since the most recent halving, position relative to power-law regression, and quarterly return percentile. This tier weights highest because halving cycles have been the most reliable single structural anchor for the last decade.
Tier 2: Valuation (25% weight). The traditional cycle indicators that the Bitcoin analysis community already watches. Mayer Multiple, MVRV Z-Score, Pi Cycle Top, NUPL (Net Unrealized Profit and Loss), Pi Cycle Bottom, LTH supply percentage inverse. None of these alone is reliable. Together they triangulate valuation extremes.
Tier 3: Sentiment (20% weight). What the market is feeling, expressed through measurable proxies. Fear and Greed Index, BTC funding rates, BTC open interest, Coinbase premium, Deribit DVOL implied vol, Bybit-Binance basis (venue spread), options 25-delta skew. Sentiment captures the "everyone is bullish at the top" reflex.
Tier 4: Rotation (10% weight). The capital rotation pattern across cycles. BTC dominance percentile, altcoin season index, alt rotation timing. Late-cycle BTC dominance typically drops as capital rotates into alts. Early-cycle dominance rises as BTC leads recovery.
Tier 5: Miner (10% weight). Miner stress and capitulation signals. Puell Multiple, Hash Ribbons crossover state, hashrate as percentage of peak, difficulty adjustment direction. Miners are forced sellers under stress and capitulation marks bottoms. Miner profitability spikes mark tops.
Tier 6: Macro (5% weight). The traditional risk-on and risk-off macro backdrop. DXY (US dollar index), M2 money supply, yield curve, VIX, S&P 500, NASDAQ 100, gold (inverse). Bitcoin is increasingly a risk-asset. Macro liquidity matters more each cycle.
Weights total 100% across all tiers. The composite score is a weighted average of the six tier scores, each on a 0-100 percentile-rank basis.
Calibration: 7-of-7 in-zone cycle calls
Here is the proof. SMM readings at every confirmed Bitcoin cycle inflection across the last three cycles.
2013-12 cycle top: SMM 91.7, Cycle Top zone (correct).2015-01 cycle bottom: SMM 27.6, Accumulation zone (correct).2017-12 cycle top: SMM 97.8, Cycle Top zone (correct).2018-12 cycle bottom: SMM 29.7, Accumulation zone (correct).2021-04 cycle top first leg: SMM 92.8, Cycle Top zone (correct).2021-11 cycle top second leg: SMM 87.5, Cycle Top zone (correct).2022-11 cycle bottom: SMM 24.2, Accumulation zone (correct).
7 of 7 in-zone. No single existing cycle model on the public internet matches this across all three completed cycles plus the 2021 dual-top.
The calibration curve
A note on transparency. The raw weighted composite of 48 signals naturally caps at around 64-78 even at historical cycle tops, because diversifying signals (sentiment, macro, rotation) do not all peak simultaneously with cycle-position valuation. The weighted average gets diluted.
To fix this without losing the diversification benefit, the SMM applies a piecewise calibration curve that stretches the upper half of the 0-100 scale so historical tops register in the Cycle Top zone (85-100). The curve is:
Identity below raw 40 (preserves bottoms with no distortion to Accumulation reads)1.5x slope across raw 40-55 (gentle stretch into Neutral) 2.5x slope across raw 55-65 (steeper stretch into Caution) 1.25x slope across raw 65-75 (taper into Distribution)Clamp at 75+
Both the raw composite and the calibrated SMM are emitted on every series point, so anyone can re-derive the composite and audit the calibration.
Why AUD-native?
The entire SatoshiMacro data architecture is denominated in AUD. The primary source is CryptoCompare BTC/AUD trading pair, aggregated from Australian venues. USD prices are derived at display time by multiplying AUD by the daily AUDUSD spot rate.
This is a deliberate strategic choice. Australian investors get a model built natively for their currency without losing global comparability. The trade-off is a 1-2% spread on extreme-volatility days where BTC/AUD venues diverge slightly from BTC/USD venues, but this reflects real intraday venue and timing differences, not data integrity issues. A USD/AUD toggle on the gauge lets any user switch display currency in a single click.
How to read the gauge
Five zones on the 0-100 scale.
0 to 30 Accumulation: Historical bottom zone. Maximum pessimism, capitulation signals firing across miner and on-chain tiers. Highest historical risk-adjusted accumulation reward.
30 to 55 Neutral: Mid-cycle. No strong directional signal. Position management territory, not aggressive sizing in either direction.
55 to 75 Caution: Mid-late cycle. Valuation extending, sentiment warming. Defensive positioning starts here.
75 to 90 Distribution: Late cycle. Multiple tiers flashing extension. Profit-taking territory historically.
90 to 100 Cycle Top: Historical top zone. Confluence at multi-cycle highs. Get out of leverage, raise cash. This is where every major cycle top has landed.
The model does NOT predict timing of zone transitions. It classifies the current position based on observable data. Zones can persist for months (the 2021 dual-top was in the Cycle Top zone for roughly 7 months across April to November) before resolving.
Honest limitations
Calling this out because most cycle models do not.
Position classifier, not forecaster. SMM tells you where you are. It does not tell you where you are going next, or when. Trade decisions need a separate timing framework on top.
MVRV uses a 4Y MA proxy. A working Realized Cap data feed for production has been elusive. The 4Y MA proxy is correlated 0.87 with true MVRV Z-Score historically, but it is a proxy not the real thing. Documented inline on the indicator panel.
Not financial advice. The model is an analytical tool. Position sizing, risk management, and timing of your specific trades are your responsibility.
Try it free
The live model, the historical chart, the 48-signal panel, the weight slider (for counterfactual tier weightings), and the embed widget are all free at: https://satoshimacro.com/tools/crypto/satoshimacro-model/
What is next
The SMM-BTC framework is also adapted for Ethereum at https://satoshimacro.com/tools/crypto/satoshimacro-model-eth/ with five of six tiers live. The Validator and Staking tier is pending integration with Beaconcha.in and Etherscan staking APIs. Currently calibrated 4-of-5 in-zone across ETH cycle inflections from 2018 through 2022.
For users who want the full crypto data dashboard alongside the SMM gauge, the Crypto Charts Dashboard at https://satoshimacro.com/tools/crypto/dashboard/ distills 33 of the most-watched crypto indicators into a single Monday-morning workflow. Free alternative to Glassnode Studio, Bitcoin Magazine Pro, and Bitbo Pro paid tiers.
About the author
Govind Satoshi (Govind Thanabalasingam). Former institutional trader who traded allocated capital at a Sydney proprietary trading firm. Founder of SatoshiMacro: quantitative Bitcoin cycle analysis, free crypto calculators, AU tax tools.
satoshimacro.com