Smooth Treasury at a Glance
Treasury companies typically do not put to work their assets, it is like owning real estate and benefiting from capital appreciation but without generating rental income. The Smooth Treasury has proprietary strategies to generate income on its digital assets while ensuring proper risk management.
Every figure on this page is read from the treasury accounts and refreshed automatically.
Note: the current live numbers run significantly above the engine's backtested average, which is expected so soon after a bear-market recovery: that phase of the cycle is historically the most productive for the books. The backtested figures in the Bot Management Engine section are the calmer, cycle-average expectation.
Outperform Bitcoin
The strategy is built to beat holding bitcoin outright. Grid yield and leveraged trading income are earned on top of the assets the treasury holds, and both are reinvested, so over the same window and from the same starting capital the treasury ends with a larger balance and more bitcoin than a pure buy and hold of bitcoin would have held.
Treasury value in dollars, every recorded day from Aug 10, 2026: $3,345 now against $2,592 committed, +29.0%. The blue line is the Bitcoin Buy-and-Hold baseline, the same capital committed on day one and held in bitcoin alone, which would be $3,031 now. The gold line is the account, so it moves with both the income and the market price of what it holds.
Reported, not illustrative: what the whole treasury is worth measured in bitcoin, at each day's bitcoin price. The Bitcoin Buy-and-Hold baseline, the same $2,592 held in bitcoin from Aug 10, 2026 and left alone, would still be 0.0414 BTC, the flat line. The treasury holds 0.0441 BTC, 6.5% more.
Grid yield, week by week
Reported from the daily record, not a model. Grid yield is the matched-pair component: it comes from price choppiness rather than direction, and in every one of the recorded days it has been positive.
One bar per seven day block, anchored to the start of the record, so the first and last blocks cover fewer than seven days. Each bar sums its own block, nothing accumulates between bars. The record covers 99 days, Aug 1, 2026 to Nov 7, 2026. Hover a bar for the dollars it earned and the return that represents on the capital committed. Grid yield is the matched-pair component: it comes from price choppiness rather than direction, and it has been positive in every one of the 99 recorded days, which is the property that makes it the base of the compounding rather than the upside. Monthly totals August $24, September $26, October $22, November $6. The leveraged trading income is reported in the headline figures and deliberately left off this chart: it is an order of magnitude larger and swings both ways, which would flatten the grid bars into the axis and hide the part that never goes negative.
Where the capital sits
Capital is spread across digital assets and equities so no single position determines the outcome. Sleeve weights move with performance and are rebalanced through new deployments rather than forced sales.
What the return cost in risk
Every sleeve runs against an automated rule set, and the board is regenerated continuously. The fair question is what the return cost in risk, so both columns below are the same capital over the same recorded days: the treasury as reported, against that capital held in bitcoin and left alone.
| Measured over 91 recorded days | Smooth Treasury | Bitcoin, held |
|---|---|---|
| Return over the window | +25.2% | +16.8% |
| Sortino ratio, downside risk only | 2.74 | 3.06 |
| Volatility, annualised | 77.5% | 41.6% |
| Downside deviation, annualised | 43.9% | 23.4% |
| Deepest drawdown | -36.0% | -12.6% |
| Worst day | -8.6% | -5.7% |
Both columns are the same capital marked daily at the same prices. The ratio is return per unit of downside risk, with zero as the minimum acceptable return, so a bad day counts and a good one does not. The treasury runs at higher volatility and a deeper drawdown than holding bitcoin, and earned materially more return on that downside risk. The window is 91 days, short enough that these are indicative rather than settled, and the treasury holds the coins it trades, so part of its correlation with the benchmark is by construction.
Smooth Crypto Market Indicator v1
The treasury's performance comes from better tooling and better strategy, and this is one of the base models: it tracks where bitcoin sits in its long-run overbought and oversold structure, which tells the engine when the market offers the cheap end of the band and when it asks for patience at the top. Updated daily; hover any day for its reading.
What the engine is built to deliver
Backtested across April 2022 to October 2026 on live-recorded data, with exchange-executable parameters (integer grid leverage, daily margin discipline). Two books: the growth engine and the income sleeve. The benchmark column is the do-nothing alternative: buy BTC on day one of the same window and hold it, no grids, no leverage, no income. These are cycle-average expectations, not promises; the current window sits above them because the market is coming out of a bear.
| Metric | Growth engine | Income sleeve | Buy & hold BTC |
|---|---|---|---|
| Backtested CAGR | 70.7% | 51.6% | 13.9% |
| Deepest drawdown | -12.4% | -8.0% | -66.1% |
| Average grid yield, per month | 1.9% | 1.6% | 0% (no income) |
| Sortino ratio | 5.75 | 5.95 | 0.40 |
| Calmar ratio | 5.71 | 6.46 | 0.21 |
Backtest window: 1,663 days (Apr 1, 2022 to Oct 19, 2026). The growth engine is the aggressive 2x tier (it43), the income sleeve the conservative tier (it16); both are the adopted production configuration and both recompute directly from the production harness (stcore/backtest/variant_backtest.py) in the treasury repo, running the Oct 6 rainbow-band checkpoint design: the SBTC de-risk ruler for cycle-top protection and rainbow-band gates for the confirmed-bottom 2x deployments. The design validated at 78.5-81.6% CAGR on the Oct 5-6 checkpoint windows; the current window (which includes the recent SOL/BTC slide) nets to the figures above. Engine v1 reads the Market Indicator v1 bands: the accumulation gate deploys leverage at confirmed oversold extremes, and the de-risk ruler works off the same long-run structure. The backtest applies the same execution constraints as the live books: integer grid leverage, margin-rotation reserves, and the market-structure de-risk rules described in the indicator above. The benchmark column uses the identical window and the identical daily BTC price series: buy on the first day, hold to the last. Its deepest drawdown is the 2022 bear market (-66%), which the books sidestep through the de-risk rules, and it produces no income along the way. Both models are version 1; later versions target better KPIs at equal or lower risk.
Run your own assumptions through the measured rates
Preloaded with what the books actually earned and what the assets actually did, every field editable. The grid rate is a percentage of the base asset value, so as the balance grows the income it produces grows with it. Leveraged trading income is an input you set rather than a number projected for you.
Current KPIs restores what the record measured, all three rates over the same 100 day window, and it is the preset that opens first. Realistic sets 1% a month grid yield, 25% a year asset growth and 1% a month leveraged trading income, and leaves your capital as typed. Every field below stays editable, and the schedule updates as you type.
| Year | Income that year | Cumulative income | Capital at end | Return |
|---|---|---|---|---|
| Year 1 | $3,648 | $10,197 | 293.4% | |
| Year 2 | $18,001 | $40,119 | 1447.8% | |
| Year 3 | $74,469 | $157,837 | 5989.4% | |
| Year 4 | $296,626 | $620,965 | 23857.0% | |
| Year 5 | $1,170,638 | $2,443,009 | 94151.9% |
$2,592 compounds to $2,443,009 over 5 years, 94151.9% on the capital. $1,170,638 of that is income earned along the way and reinvested; the balance is the assets appreciating, counted both on the original capital and on the income that has already gone back in.
The same 3.05% a month pays $150.34 in month one and $141,694.51 a month by year five, because the rate is charged on a balance that grows: $2,592 of assets at 100.0% a year is $82,944 by then. That is the part a flat extrapolation of the grid rate misses.
Most of the difference from the grid rate alone is the leveraged trading line, 2.75% a month on the current balance. On the record that line is marked, it was negative on 53 of the 100 days measured, and it is the youngest book's directional profit rather than grid chop. Those three rates are measurements of 100 days in one market, so the five year table above is arithmetic on them, not a forecast.
Illustrative schedule, not a forecast and not a performance claim. The grid rate and the leveraged trading rate are both measured on the record between Jul 31, 2026 and Nov 7, 2026 (100 days); the asset growth is that same window's price move annualised, and its whole life figure plus the per asset moves are in the notes above. All three are short window marks, the trading rate can be negative, and none of them is a rate to carry for five years. Income is reinvested every month, asset appreciation and income are compounded together, and the capital at the end therefore counts both. A larger allocation will not necessarily earn the same rate, since grid yield depends on how much the price chops inside the book's range.