Simulated · pre-launch
Algotoria Spike Diversified
Algotoria’s dormant-by-design spike-capture strategy, run by the Algotoria Spike neural-network engine: it trades up to 40 high-liquidity perpetual futures only when model conviction is exceptional. Its capital is held in a collateral book built to reduce the account’s dependence on US Treasuries and the US dollar as a store of value: 50% at all times outside US-dollar-backed assets, in tokenised gold, a tokenised S&P 500 ETF and bitcoin weighted by inverse volatility, and 50% in two yield-bearing US-dollar stablecoins, backed largely by US Treasury bills, so their yield rises and falls with Treasury-bill rates. It is delivered through non-custodial Separately Managed Accounts (SMA) or Managed Sub-Accounts (MSA), so the investor keeps custody. Simulated results: no client money has been managed under it yet.
Out of sample from 1 July 2026: the models and every rule and parameter were fixed before that date, so later months were never used to choose anything; earlier months are a walk-forward simulation whose settings were chosen on that same history. The collateral rule is the exception: it was chosen on data to 30 June 2026 but adopted in October 2026, after these months had been observed.
Signal: 18-model ensemble
Simulated execution for a constant $10 M book on OKX: every order sized and capped as if the book were $10 M, filled on the historical 1-second order book 30 seconds after each signal, entries and exits alike, under the production execution loop, taker fees included; returns compounded. Restated in early November 2026 to a 30-second signal-to-order latency applied inside the engine (ML-0025 / ML-0026); the figures of 20 October 2026 are superseded.
| Year | Jan | Feb | Mar | Apr | May | Jun | Jul | Aug | Sep | Oct | Nov | Dec | Year | Since simulation start |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2024 | +12.4% | +4.1% | +5.3% | +0.2% | +2.1% | +0.4% | +0.2% | +4.7% | +2.1% | +1.4% | +5.1% | +7.5% | +55.3% | +55.3% |
| 2025 | +3.3% | −1.5% | +1.5% | +2.6% | +2.1% | +1.4% | +1.6% | +0.6% | +3.4% | +9.3% | −0.2% | +0.3% | +26.9% | +97.0% |
| 2026 | +1.9% | +19.3% | −2.6% | +4.0% | +0.9% | −3.1% | +1.0% | +11.9% | −0.7% | −1.6% | +32.8% | +161.7% |
| Year | Jan | Feb | Mar | Apr | May | Jun | Jul | Aug | Sep | Oct | Nov | Dec | Year | Since simulation start |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2024 | +12.4% | +4.1% | +5.3% | −4.5% | +2.1% | +0.4% | −0.5% | +4.7% | +2.1% | −0.3% | +5.1% | +7.5% | +44.5% | +44.5% |
| 2025 | 0.0% | −1.5% | +1.5% | +1.7% | +2.1% | +1.4% | +0.2% | +0.6% | +3.4% | +7.8% | −0.2% | +0.3% | +18.5% | +71.3% |
| 2026 | −0.3% | +19.3% | −2.6% | 0.0% | +0.9% | −3.1% | +0.6% | +11.9% | −0.7% | −4.2% | +21.1% | +107.4% |
| Year | Jan | Feb | Mar | Apr | May | Jun | Jul | Aug | Sep | Oct | Nov | Dec | Year | Since simulation start |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2024 | +12.4% | +4.1% | +5.3% | −5.5% | +2.1% | +0.4% | −0.6% | +4.7% | +2.1% | −0.6% | +5.1% | +7.5% | +42.4% | +42.4% |
| 2025 | −0.6% | −1.5% | +1.5% | +1.6% | +2.1% | +1.4% | −0.1% | +0.6% | +3.4% | +7.5% | −0.2% | +0.3% | +16.8% | +66.4% |
| 2026 | −0.7% | +19.3% | −2.6% | −0.8% | +0.9% | −3.1% | +0.5% | +11.9% | −0.7% | −4.8% | +18.8% | +97.8% |
| Metric | Algotoria Spike Diversified (simulated)simulated · Gross | Bitcoinbenchmark | BITA Crypto 10benchmark | Gold (PAXG)benchmark | S&P 500benchmark |
|---|---|---|---|---|---|
| 1 month | −1.57% | −7.28% | −13.05% | −6.38% | −1.11% |
| 3 months | +9.29% | +23.28% | +22.41% | −3.68% | +1.73% |
| 6 months | +7.82% | +1.56% | +2.73% | −15.66% | +5.35% |
| 12 months | +32.93% | −29.26% | −32.65% | −2.51% | +10.97% |
| 24 months | +90.50% | +10.28% | +2.16% | +41.05% | +33.10% |
| CAGR | +40.48% | +21.97% | +10.27% | +24.87% | +18.11% |
| Total return | +161.73% | +75.46% | +31.88% | +87.53% | +60.13% |
| Metric | Algotoria Spike Diversified (simulated)simulated · Net 25 | Bitcoinbenchmark | BITA Crypto 10benchmark | Gold (PAXG)benchmark | S&P 500benchmark |
|---|---|---|---|---|---|
| 1 month | −4.23% | −7.28% | −13.05% | −6.38% | −1.11% |
| 3 months | +6.34% | +23.28% | +22.41% | −3.68% | +1.73% |
| 6 months | +4.50% | +1.56% | +2.73% | −15.66% | +5.35% |
| 12 months | +21.18% | −29.26% | −32.65% | −2.51% | +10.97% |
| 24 months | +62.15% | +10.28% | +2.16% | +41.05% | +33.10% |
| CAGR | +29.39% | +21.97% | +10.27% | +24.87% | +18.11% |
| Total return | +107.38% | +75.46% | +31.88% | +87.53% | +60.13% |
| Metric | Algotoria Spike Diversified (simulated)simulated · Net 30 | Bitcoinbenchmark | BITA Crypto 10benchmark | Gold (PAXG)benchmark | S&P 500benchmark |
|---|---|---|---|---|---|
| 1 month | −4.76% | −7.28% | −13.05% | −6.38% | −1.11% |
| 3 months | +5.75% | +23.28% | +22.41% | −3.68% | +1.73% |
| 6 months | +3.83% | +1.56% | +2.73% | −15.66% | +5.35% |
| 12 months | +18.91% | −29.26% | −32.65% | −2.51% | +10.97% |
| 24 months | +56.92% | +10.28% | +2.16% | +41.05% | +33.10% |
| CAGR | +27.24% | +21.97% | +10.27% | +24.87% | +18.11% |
| Total return | +97.78% | +75.46% | +31.88% | +87.53% | +60.13% |
| Metric | Algotoria Spike Diversified (simulated)simulated · Gross · 5‑minute | Bitcoinbenchmark · 5‑minute | BITA Crypto 10benchmark · daily | Gold (PAXG)benchmark · daily | S&P 500benchmark · daily |
|---|---|---|---|---|---|
| Current drawdown[8] | −2.93% | −37.26% | −43.82% | −29.64% | −2.99% |
| Adjusted max drawdown[4] | −4.24% | −52.78% | — | — | — |
| Raw max drawdown[5] | −7.66% | −53.84% | −57.84% | −29.67% | −19.00% |
| Max drawdown duration[9] | 94 days | 389 days | 389 days | 275 days | 127 days |
| Ulcer index[10] | 0.97 | 24.07 | 31.41 | 11.88 | 3.33 |
| Calmar ratio[11] | 5.28 | 0.41 | 0.18 | 0.84 | 0.95 |
| Adjusted Calmar[12] | 9.55 | 0.42 | — | — | — |
| Sortino ratio[13] | 8.33 | 0.69 | 0.19 | 1.26 | 1.22 |
| Sharpe ratio[14] | 2.55 | 0.45 | 0.13 | 1.00 | 0.93 |
| Annualised volatility[15] | 14% | 39% | 46% | 20% | 15% |
| Metric | Algotoria Spike Diversified (simulated)simulated · Net 25 · 5‑minute | Bitcoinbenchmark · 5‑minute | BITA Crypto 10benchmark · daily | Gold (PAXG)benchmark · daily | S&P 500benchmark · daily |
|---|---|---|---|---|---|
| Current drawdown[8] | −5.55% | −37.26% | −43.82% | −29.64% | −2.99% |
| Adjusted max drawdown[4] | −6.37% | −52.78% | — | — | — |
| Raw max drawdown[5] | −8.62% | −53.84% | −57.84% | −29.67% | −19.00% |
| Max drawdown duration[9] | 165 days | 389 days | 389 days | 275 days | 127 days |
| Ulcer index[10] | 1.97 | 24.07 | 31.41 | 11.88 | 3.33 |
| Calmar ratio[11] | 3.41 | 0.41 | 0.18 | 0.84 | 0.95 |
| Adjusted Calmar[12] | 4.62 | 0.42 | — | — | — |
| Sortino ratio[13] | 3.67 | 0.69 | 0.19 | 1.26 | 1.22 |
| Sharpe ratio[14] | 1.70 | 0.45 | 0.13 | 1.00 | 0.93 |
| Annualised volatility[15] | 15% | 39% | 46% | 20% | 15% |
| Metric | Algotoria Spike Diversified (simulated)simulated · Net 30 · 5‑minute | Bitcoinbenchmark · 5‑minute | BITA Crypto 10benchmark · daily | Gold (PAXG)benchmark · daily | S&P 500benchmark · daily |
|---|---|---|---|---|---|
| Current drawdown[8] | −6.08% | −37.26% | −43.82% | −29.64% | −2.99% |
| Adjusted max drawdown[4] | −7.12% | −52.78% | — | — | — |
| Raw max drawdown[5] | −9.64% | −53.84% | −57.84% | −29.67% | −19.00% |
| Max drawdown duration[9] | 165 days | 389 days | 389 days | 275 days | 127 days |
| Ulcer index[10] | 2.43 | 24.07 | 31.41 | 11.88 | 3.33 |
| Calmar ratio[11] | 2.82 | 0.41 | 0.18 | 0.84 | 0.95 |
| Adjusted Calmar[12] | 3.82 | 0.42 | — | — | — |
| Sortino ratio[13] | 2.93 | 0.69 | 0.19 | 1.26 | 1.22 |
| Sharpe ratio[14] | 1.53 | 0.45 | 0.13 | 1.00 | 0.93 |
| Annualised volatility[15] | 15% | 39% | 46% | 20% | 15% |
Drawdowns of the strategy and of Bitcoin are measured on five-minute paths with the adjusted definition, peak and trough each confirmed over 24 hours, with the raw maximum drawdown beside it. BITA Crypto 10, gold and the S&P 500 have only daily closes over this window, so their columns show the raw drawdown, which also sets their duration, ulcer index and Calmar ratio. The tiles above read the same figures.
Drawdown is read together with duration. Half of the collateral is held outside US-dollar-backed assets to diversify away from the dollar, not to make drawdowns shallower: that half adds market risk of its own, and the simulated drawdown is deeper than Algotoria Spike Stable’s.
Collateral[16]
Yield-bearing stablecoins[17] for 50% of the collateral and, for the other 50%, tokenised gold, a tokenised S&P 500 ETF and bitcoin held outside US-dollar-backed assets, weighted by inverse volatility within fixed limits; both halves held at all times. Algotoria manages the whole book; clients deposit and withdraw in USDT.
| Asset type | Assets | Share | Limit[18] |
|---|---|---|---|
| RLUSDRLUSD | RLUSD | 25.2% | No cap |
| USDGUSDG | USDG | 25.2% | No cap |
| Tokenised goldXAUT, PAXG | XAUT, PAXG | 13.8% | 30% |
| Tokenised S&P 500 ETFxSPY | xSPY | 25.9% | 40% |
| BitcoinBTC | BTC | 10% | 30% |
Rules, limits and exclusions
- Half outside the US dollar, at all times
- The collateral has two halves. 50% is held in the yield-bearing US-dollar stablecoins RLUSD and USDG, in equal parts. 50% is held outside US-dollar-backed assets, in tokenised gold (XAUT and PAXG in halves), xSPY, a tokenised S&P 500 ETF, and bitcoin. The purpose is to reduce the account’s dependence on US Treasuries and the US dollar as a store of value, as protection against the dollar’s debasement: this is diversification, not a hedge or a guarantee. The basket is held at all times: it is not switched on or off by market conditions or by the account’s drawdown. Every instrument is held on OKX and accepted by the venue as margin, so the whole book sits in the account that margins the trading book.
- Inverse-volatility weighting within fixed limits
- Inside the non-dollar half each asset is weighted in inverse proportion to its realised volatility over the previous 90 days, so that no single asset dominates the basket’s risk: the more volatile an asset has just been, the smaller its share. The targets are capped at 30% of the collateral for gold, 40% for xSPY and 30% for bitcoin; any excess is spread over the other assets. The rule is non-discretionary.
- Re-weighting and the share held outside the dollar
- The basket is re-weighted once a quarter, at the first day-end of each calendar quarter (UTC). Between quarters a leg is traded back to its target only once it has drifted more than 20% of its own weight away, so the share held outside US-dollar-backed assets moves with prices between re-weights; the mandate keeps it between 45% and 55% on average. In the simulation it averaged 50.5% and ranged from 44.8% to 54.5%. The shares shown above are those held on 30 September 2026, from the targets set at the re-weight of 1 July 2026.
- US equities still carry US risk
- xSPY is exposure to the US stock market, so it still carries US geopolitical and policy risk, although many of the companies in the S&P 500 are globally diversified and earn a large share of their revenue outside the United States. The diversification away from the US dollar therefore rests mainly on gold and bitcoin.
- The yield given up
- Gold, xSPY and bitcoin earn no yield, so the book earns the stablecoin rewards on only half of its collateral: about half of the yield of a stablecoin-only book, roughly 2 percentage points of NAV a year.
- Liquidity
- Execution is staged against venue liquidity: no single day’s trade in any instrument exceeds 10% of that pair’s 30-day median turnover on OKX, and any remainder carries forward.
- What the basket does not hold
- No single stocks, no crypto-linked equities, no leveraged or inverse tokens and no altcoins: equity exposure is index exposure, and bitcoin is the only crypto-asset in the collateral.
- Performance measure
- Gains and losses on the collateral, including the basket, count in the performance measure (NAV-Based Method).
- Margin treatment
- Each instrument is accepted as margin at a discount set by the exchange, which can change it. At the first tier the venue currently counts 98% of the value of the stablecoins, 98% of bitcoin, 95% of XAUT, 85% of PAXG and 97% of xSPY, and less at larger holdings. With half of the book in these assets, a sharp fall in them reduces the margin available to the strategy during a shock; the margin headroom is monitored, and new positions may be declined when the stressed margin is thin.
Trading signal
A neural-network signal rather than a rules book: one conviction score per instrument, a deliberately high entry bar and a full exit stack on every position. Beside it: the strategy’s capacity and how it executes.
How the signal works
Each instrument is analysed continuously, around the clock, by a proprietary neural network that reads horizons from order-flow detail up to multi-week context and reduces them to a single directional conviction score. Every instrument has its own direction model, deployed as a multi-seed ensemble so that no individual training run can drive a trade. A separate model forecasts near-term volatility; it never opens a position and feeds the capital-preservation overlay below. Algotoria Spike Diversified changes only the collateral: the signal is the same as Algotoria Spike Stable’s.
Capacity[22]
Investment terms
Every investor in Algotoria Spike Diversified selects a risk tier during onboarding. The tier sets the drawdown budget, average and maximum leverage, and the minimum allocation.
Terms
| Investment form | SMA / MSA |
|---|---|
| Investment assets[21] | Up to 40 USDT-margined perpetual futures; the simulated record trades 7: Bitcoin, Ethereum and five major altcoins |
| Minimum investment[19] | 50,000–150,000 USDT |
| Management fee | 0% |
| Performance fee[3] | 25–30% quarterly |
| PnL method[26] | NAV-Based Method |
| Hurdle rate | None |
| High watermark | Rolling, post-fee |
| Lock-in period | None |
| Recommended horizon | ≥ 1 year |
| Redemptions | 5-day notice |
| Deposits and withdrawals | USDT only; Algotoria converts into and out of the collateral |
| Trading venues | OKX, Binance, Bybit |
| Custody | Investor retains; trade-only API |
Risk tiers[20]
| Tier | Max DD | Avg leverage | Max leverage | Min allocation |
|---|---|---|---|---|
| High | 35% | 100% | 300% | 50,000 USDT |
| Medium | 25% | 67% | 200% | 100,000 USDT |
| Conservative | 15% | 33% | 100% | 150,000 USDT |
Methodology and sources
- Returns
- Daily time-weighted returns in USDT over UTC days, from 1 January 2024 to 31 October 2026. Monthly, yearly and rolling figures compound the daily returns; CAGR annualises the full period.
- Fee bases
- Net 30 and Net 25 are modelled from the gross simulated curve by deducting a 30% or 25% fee on profit above the rolling, post-fee high watermark, settled quarterly. The assumed collateral yield is included, as it is in the fee base of the NAV-Based Method. They are not taken from a fee-paying account.
- Simulation
- The production trading engine on exchange market data under execution profile close-5m-v1: decisions only on completed five-minute candles, stop-losses and take-profits tested against the bar’s close, every fill at the following boundary, and no stop or take-profit resting at the exchange. Execution is simulated for a constant $10 M book on OKX, taker fees and market impact included, at 3× gross exposure.
- Walk-forward
- Fold models are trained on all data from January 2021 up to each fold’s cut (the same convention as the deployed models), the last cut being 1 July 2026; after it the models of the same walk-forward run with a training cut of 1 July 2026, fixed since. The signal is the mean of 18 models trained from different random seeds, none of which was used to select any parameter. The entry threshold was selected on data to 30 June 2026 and frozen on 21 August 2026. From 1 July 2026 the record is out of sample: the models and every rule were fixed before that date, nothing was re-selected on the later months, and those months were scored once, in a pre-registered reading.
- Collateral basket
- The collateral is simulated day by day: 50% in RLUSD and USDG in equal parts, and 50% in tokenised gold (XAUT and PAXG in halves), xSPY and bitcoin, each weighted in inverse proportion to its realised volatility over the previous 90 days with the targets capped at 30% of the collateral for gold, 40% for xSPY and 30% for bitcoin, re-weighted at the first day-end of each calendar quarter (UTC) and, between quarters, traded back to target only when a leg drifts more than 20% of its own weight away. Basket trades are filled at the OKX day-end close and charged the cost measured in Algotoria’s collateral execution study of 27 September 2026 (patient maker orders on OKX spot, at the sleeve sizes of a $10 M book, queue limits included); before xSPY and PAXG listed on OKX the simulation prices the S&P 500 and gold. The book is credited through the collateral ledger, including the basket’s own gains and losses.
- Collateral yield
- Idle collateral earns the venue’s stablecoin rewards, credited day by day through a collateral ledger together with the gains and losses of the volatility-weighted basket; for Algotoria Spike Stable, on stablecoins alone, the credit is about 4.2 percentage points a year over the window. A series without this yield is never published.
- Drawdown
- Adjusted drawdown with the 24-hour peak and trough tests, on five-minute paths: the strategy’s equity path for every fee basis, and Bitcoin’s OKX five-minute closes; the raw maximum drawdown is shown beside it, and the drawdown chart plots each day’s lowest reading. BITA Crypto 10, gold and the S&P 500 have no five-minute history over the window: daily closes, raw drawdown. Durations count calendar days from the first day below the adjusted peak to the day it is regained, or to 31 October 2026 where still open.
- Ratios
- Sharpe and Sortino use the US 3-month Treasury yield (FRED GS3M), 4.45% on average over the period; volatility scales daily returns by √252.
- Data and updates
- Figures as of 31 October 2026, generated 3 November 2026; updated monthly. The simulation is extended at every month-end. Account and term facts are dated where they appear.
Important information
Regulatory status
Algotoria Limited · BVI Registration No. 2161048 · Approved Investment Manager under the Securities and Investment Business Act, 2010 · Certificate No. IBR/AIM/25/2214 · Registered with the BVI Financial Services Commission. The approved-manager regime does not provide the protections of retail fund regulation; clients have no recourse to a statutory ombudsman or compensation scheme.
Eligibility and jurisdiction
This page is for professional investors (as defined by SIBA) only. It is not directed at United States persons or at residents of any jurisdiction on the Restricted Jurisdictions register kept by the firm’s MLRO, and it is not intended for use where its distribution would be contrary to local law or regulation. Each client confirms professional-investor status in the Asset Management Agreement.
No offer, no advice, no tax advice
This page is published for information only. It is not an offer or solicitation to buy or sell any security or service, and not investment, legal or tax advice or a personal recommendation. Algotoria does not provide tax reporting; each client is solely responsible for the tax arising from its account. Any engagement is governed by the Asset Management Agreement alone.
Simulated performance basis
Performance is simulated: a research backtest of a frozen configuration on the production trading engine, at 3× gross exposure, from 1 January 2024 to 31 October 2026, with simulated execution for a constant $10 M book on OKX, taker fees and market impact included. The trading signal is generated walk-forward, and the return on the collateral, the stablecoin rewards and the gains and losses of its volatility-weighted basket, is credited through a collateral ledger. The collateral rule was chosen on data within this window, so the figures state a hypothesis rather than evidence of what the rule will do. No client capital has been managed under Algotoria Spike Diversified and no account achieved these results. Returns are daily time-weighted returns in USDT over UTC days, shown gross and with modelled 30% and 25% fees.
Hypothetical performance
Simulated results are prepared with hindsight and do not reflect actual trading. They cannot account for the effect of the orders themselves on the market, for the operational frictions of managing a real account, or for an investor’s own decisions during a drawdown. Fold models are trained only on earlier data, but the final configuration was selected on results to 30 June 2026 and frozen on 21 August 2026; a re-trained set of models would give a different record. Simulated and past performance are not indicative of future results.
Drawdown definition and gross basis
Drawdowns use Algotoria’s adjusted definition: peak and trough each confirmed over 24 hours, measured on the strategy’s five-minute equity path and gross of performance fees. The raw maximum drawdown, which counts every five-minute low in full, is shown beside the adjusted figure. Benchmarks without a five-minute history are shown on daily closes with the raw drawdown only. A drawdown can exceed the investor drawdown budget.
Benchmarks
Bitcoin and the BITA Crypto 10 Index are the primary references; the S&P 500 and gold are context. BITA Crypto 10 Index, BITA GmbH; not sponsored or endorsed by BITA. Benchmarks are unmanaged and bear no fees or costs. They differ from the strategy in leverage, short exposure and volatility, and comparison with them is illustrative only.
Fees
No management fee and no hurdle. A performance fee of 25–30% of net trading profit, tiered by account balance under Appendix 1 of the Asset Management Agreement, is charged quarterly on the rolling post-fee high watermark under the NAV-Based Method, so collateral gains and losses are in the fee base. No clawback applies. The net views on this page are modelled, not billed. They are applied to the gross simulated curve, including the assumed collateral yield. Commercial terms take effect only if and when the strategy launches, under the applicable Asset Management Agreement.
Collateral
Collateral is 50% yield-bearing stablecoins (RLUSD and USDG) and 50% tokenised gold, a tokenised S&P 500 ETF and bitcoin, weighted in inverse proportion to their volatility within fixed limits, held at all times and managed by Algotoria under a non-discretionary rule. Holding half of the collateral outside US-dollar-backed assets is diversification, not a hedge or a guarantee against the debasement of the US dollar, and that half carries market risk of its own and earns no yield. Rewards depend on third-party instruments, on the account’s VIP status and on prevailing rates, and are not guaranteed. Each stablecoin and token is a claim on its issuer, and the venue accepts each asset as margin at a discount it can change.
Collateral basket risk
The collateral basket adds market risk that a stablecoin-only book does not carry. Gold, equities and bitcoin can fall at the same time as the trading book, deepening or lengthening a drawdown; bitcoin in particular is volatile, and the volatility weighting reduces its share, it does not remove it. Holding these assets is diversification away from the US dollar, not a hedge or a guarantee against its debasement, and xSPY is exposure to the US stock market, which carries US geopolitical risk. The basket earns no yield, so the account gives up about half of the stablecoin yield. The exchange counts these assets as margin at a discount, which reduces the margin available to the strategy during a shock; the margin headroom is monitored, and new positions may be declined when it is thin. Tokenised equities trade on the exchange with limited depth and may take hours or days to buy or sell at the target size. The tokenised instruments carry issuer, redemption, tracking and de-peg risk, and re-weighting and rebalancing incur trading costs.
Tokenised instruments
xSPY listed on OKX in July 2026 and PAXG in October 2025, so before those dates the simulation prices the underlying assets, the S&P 500 and gold, not the tokens. Tracking error, premium or discount to net asset value, issuer risk and 24-hour trading against an index that does not trade around the clock are outside the model; dividends on the equity token are reinvested by the issuer net of US withholding tax of up to 30%, so the total-return basis slightly overstates them.
Yield basis
In production the stablecoins earn what RLUSD and USDG pay on OKX, currently 4.2% and 4.1% at VIP status. Those rates are current, not contractual, depend on the account’s VIP status and the venue’s terms, move, and can be withdrawn; redemption cost, lock-up terms and de-peg risk are not modelled.
Principal risks
Leveraged crypto-asset derivatives can lose all capital committed. Read the full Risk Notice, in particular its sections on leverage and the drawdown budget, drawdown expectations, market volatility, model risk, exchange and stablecoin risk, liquidity and redemptions, and regulatory and legal risk.
Data currency
Figures are as of 31 October 2026 and are updated monthly. Each monthly publication is versioned in the firm’s records; earlier versions are available on request.
Status and what happens next
Algotoria Spike Diversified begins operating only if the Algotoria Spike engine passes its pre-registered forward gate, registered on 21 August 2026 and scored from 1 September 2026 to 30 November 2026; if it does not, there is no product to put the collateral basket underneath.
The collateral rule runs its own paper-forward test, registered on 10 October 2026 before the data it will be read on exists: from 1 October 2026 to 31 December 2026, on the same signal, read once and not before 1 January 2027. To pass, the book must stay within the risk framework over the window (an adjusted drawdown of less than 25%, a raw drawdown of less than 30% and no drawdown lasting 100 days or more), hold 45% to 55% of its collateral outside US-dollar-backed assets on average, keep every target within its limit without capping, re-weight exactly as specified and keep the basket’s turnover at or below 1× net asset value a year; its comparison with Algotoria Spike Stable is reported, not judged. No client capital is managed under the rule before it passes and Algotoria signs its promotion.
The venue’s collateral terms were confirmed from its published documentation on 27 September 2026; the account’s own VIP status and the redemption paths are confirmed on the account before any capital is committed. Contact your Algotoria representative or hello@algotoria.com.
Definitions and notes
The numbered markers on this page refer to these notes.
- How returns are measured Simulated, not traded. The dashed line is a simulation of Algotoria Spike Diversified: hypothetical results that no account achieved. It runs the frozen configuration on the production engine, with the signal generated walk-forward and simulated execution for a constant $10 M book on OKX, taker fees and market impact included; the final configuration was selected on results to 30 June 2026, and the collateral rule was chosen on data within the window. Simulated results have inherent limitations and are not a forecast.Returns are time-weighted, measured daily in USDT over UTC days. USDT is a stablecoin, not the US dollar: its value can move away from 1 USD, and returns measured in dollars or another currency would differ.Every series is rebased to 0% on 1 January 2024, the start of the simulation. Gross figures are before any performance fee. Simulated and past performance are not a guarantee of future results.
- Benchmarks Bitcoin: BTC-USDT spot on OKX. BITA Crypto 10 (B10): price-return index of the ten largest cryptocurrencies by trading volume (BITA Crypto 10 Index, BITA GmbH; not sponsored or endorsed by BITA). These two are the primary references.S&P 500 (SPY-USDT-SWAP perpetual on OKX; Investing.com SPY ETF data before 4 March 2026) and gold (PAXG-USDT spot on OKX; Investing.com XAU/USD data before 15 October 2025) are shown for context.Benchmarks are unmanaged and bear no fees or trading costs; an index cannot be invested in directly. Risk-free rate for Sharpe and Sortino: US 3-month Treasury yield (FRED GS3M), 4.45% on average.
- Gross, Net 30 and Net 25 Gross is the simulated return before any performance fee.Net 30 and Net 25 deduct a 30% or 25% fee on net trading profit above the rolling, post-fee high watermark, settled quarterly for the quarter just closed. The standard schedule is 30% for an account below 100,000 USDT, falling one point per band to 25% at 1,000,000 USDT and above; an executed contract may set a different rate.Both net paths are modelled from the gross simulated curve, not taken from a fee-paying account. There is no management fee and no hurdle. Benchmarks carry no fee and read the same in every mode. All switches on this page move together.
- Adjusted drawdown Drawdowns on this page use Algotoria’s adjusted definition, which tests both ends of a fall over 24 hours.Peak: a new high counts only at the lowest balance of the following 24 hours, and the accepted peak never falls.Trough: a low counts at full depth only if the account is still at or below it 24 hours later; if it has recovered, the level 24 hours on is used.The strategy and Bitcoin are measured on five-minute paths, so a crash candle reversed within 24 hours counts at the level 24 hours later; the raw maximum drawdown beside it counts every low in full. BITA Crypto 10, gold and the S&P 500 have only daily closes here, where a 24-hour test confirms nothing, so they show the raw drawdown alone. Figures are gross; net drawdowns are larger by construction.
- Raw max drawdown The deepest fall below the running peak, with no 24-hour test at either end: every five-minute low counts in full, including a crash candle that recovered within the day. It is shown beside the adjusted figure. For a daily column it is the only drawdown measure, read on daily closes.
- Monthly returns Calendar-month time-weighted returns of the simulated track in USDT, on UTC days. Year compounds the months of that year; Since simulation start is the cumulative return from 1 January 2024 to the year’s last month shown.
- Rolling returns Trailing time-weighted return to 31 October 2026 over each window. CAGR and total return run from 1 January 2024, the start of the simulation.
- Current drawdown Adjusted drawdown at the end of the as-of date, measured from the confirmed peak. A low younger than 24 hours cannot be confirmed yet, so it counts at its full depth. A daily column shows its raw drawdown.
- Max drawdown duration The longest stretch, in calendar days, from the day the series first fell below its adjusted peak to the day it regained it. Where that stretch is still open on the as-of date, it is counted to that date. A daily column counts from its raw peak.
- Ulcer index Root-mean-square of the adjusted drawdown over every five-minute reading (of the raw drawdown over daily closes for a daily column), in percentage points. Unlike maximum drawdown it counts both how deep and how long the series stays under water; lower is better.
- Calmar ratio CAGR divided by the raw maximum drawdown over the same period: return per unit of the deepest fall below the running peak, with no 24-hour test, so a crash candle reversed within the day counts in full. Higher is better.
- Adjusted Calmar CAGR divided by the adjusted maximum drawdown over the same period, shown beside the Calmar ratio for reference. A daily column has no adjusted drawdown, so it shows a dash.
- Sortino ratio CAGR minus the risk-free rate, divided by the annualised standard deviation of losing days (downside deviation, √252 scaling).
- Sharpe ratio CAGR minus the risk-free rate, divided by annualised volatility.
- Annualised volatility Standard deviation of daily returns, scaled by √252.
- Collateral The collateral book is the one thing Algotoria Spike Diversified changes against Algotoria Spike Stable: 50% of the collateral is held at all times outside US-dollar-backed assets, in tokenised gold, a tokenised S&P 500 ETF and bitcoin weighted in inverse proportion to their volatility within fixed limits, and 50% in yield-bearing stablecoins. Its purpose is diversification away from the US dollar as a store of value.The account is funded in USDT, which Algotoria converts into RLUSD and USDG and into the basket. Investors deposit and withdraw in USDT.Collateral yield. The stablecoins earn the venue’s rewards, credited day by day through a collateral ledger together with the basket’s own gains and losses; the basket itself earns no yield. Every Spike series includes this yield, and a series without it is never published.
- Stablecoins other than USDT To open USDT-margined perpetual positions, the account may borrow USDT against stablecoins other than USDT, and the borrowed USDT accrues interest at the exchange’s rate, borne by the account. Yield-bearing stablecoins (RLUSD, USDG) typically earn more than this cost, offsetting it and adding to returns. USDT collateral carries no such cost.
- Limits on the basket The most each asset class of the basket may hold, as a share of the collateral, when the targets are set at a quarterly re-weight: an inverse-volatility target above its limit is cut back to it and the excess spread over the other classes. Between re-weights the shares move with prices, within the 20% drift band. Stablecoins are uncapped: they are how clients fund the account and how trading profits arrive.
- Minimums and risk tiers Three risk tiers are chosen at onboarding: High (35% drawdown budget, maximum leverage 3×, minimum 50,000 USDT), Medium (25%, 2×, 100,000 USDT) and Conservative (15%, 1×, 150,000 USDT). Lower tiers scale leverage down; signal and execution are otherwise identical. They need a larger minimum because per-account operating costs do not fall with leverage. Spike Diversified’s budgets sit five points above Spike Stable’s to allow for the volatility of its collateral.
- Drawdown budget The investor selects the budget at 15%, 25% or 35%. The default is 35% at 3× gross exposure, and the simulated record on this page is not an account run at any of these budgets. It is a monitored ceiling, not an automatic stop. Alerts fire at 60%, 80% and 100% of it; at two-thirds the Investment Committee reviews the account and decides case by case whether to reduce risk; at the full budget the client elects to continue, lower the budget or stop.The Committee may also run the strategy below the budget, down to 0%, and restore it later; returns are then proportionally lower. The budget is not a guarantee and can be exceeded.
- Investment universe The strategy may trade up to 40 USDT-margined perpetual futures, and the Investment Committee may widen the set at any time. The simulated record trades 7: Bitcoin, Ethereum and five major altcoins. All seven legs are long/flat; the strategy never goes short.
- Capacity Capacity $10 million on OKX under the published execution policy: the figures on this page are simulated at a constant $10 million book, every order sized and capped as if the book were $10 million and filled on the historical order book under the production execution loop (Algotoria’s capacity study, October 2026). Larger accounts need their own assessment.
- Leverage and concentration Gross leverage is capped at 3×, enforced at entry; the cap is touched briefly and is not an operating level. It can be selected per account at 3×, 2× or 1×.Concentration limits scale with the leverage selected: on a 1× basis Bitcoin and Ethereum are each held to at most 35% of net asset value (105% at 3×) and any other instrument to 10% (30% at 3×), with the whole book at most 100% (300% at 3×).
- Crash-candle entry guard The guard was sized on the crash candles of 3 February 2025 and 10 October 2025. It blocks new altcoin entries only: it never closes a position and never touches Bitcoin or Ethereum.It is a protection, not a modelled edge: it costs about 13 percentage points of annual return against the single-candle rule it extends.
- Execution Taker: price-protected marketable orders at the decision, completed within minutes. No maker/taker mix is measured, because the strategy has no real account. Stop-losses and take-profits stay inside the strategy.
- NAV-Based Method Net trading profit is the change in the account’s aggregate USDT-equivalent net asset value, adjusted for net inflows and fees. Collateral gains and losses are inside the performance measure, which is the correct treatment when Algotoria manages the collateral.
Due-diligence downloads
Everything to verify this strategy independently
No exchange statements exist for this strategy: its record is simulated and no account has traded it.
For professional investors (as defined by SIBA) only: information, not an offer or investment advice; past performance is not a guarantee of future results, and all capital can be lost.
Diversification away from the dollar costs depth
Half of the collateral is held in gold, equities and bitcoin at all times, including while the account is down, so a sell-off that reaches them and the trading book together adds the basket’s loss to the engine’s and deepens or lengthens a drawdown. The simulation’s deepest adjusted drawdown is −4.2%, against −1.2% for Algotoria Spike Stable, and its Adjusted Calmar 9.55 against 22.73. Its longest spell below a previous peak lasted 94 days, against 78 for Spike Stable. The risk framework limits a drawdown to under 180 days for every strategy, with a target of under 100 days; on this line both are within the limit and under the target. Both spells begin in the second quarter of 2026. In research on a 2020-type market the basket would have deepened the March 2020 drawdown by 4 to 11 percentage points. The investor drawdown budget is 35% at the default 3×, a monitored ceiling, not a guarantee, and it can be exceeded; size the allocation on the budget, not on the simulated figure.
Crypto is volatile, and this strategy is leveraged
The strategy trades leveraged crypto perpetual futures and posts yield-bearing stablecoins and a basket of tokenised gold, a tokenised S&P 500 ETF and bitcoin as collateral; each token is a claim on its issuer. Its annualised volatility in simulation is 14.1%, against 38.8% for Bitcoin and 14.7% for the S&P 500. Losses can arrive quickly and last for months.
A simulation is not a track record
The figures on this page are replays, not results earned by clients. Future results will differ, often materially, and all capital committed can be lost. Read the full Risk Notice before allocating.