11 stock strategies. Your brokerage, your account, your rules.
From treasury income to leveraged momentum. AutoCoin trades during market hours through the brokerage you already use, with trade-only permissions. Run it in your IRA.
Not a rule-based bot. An AI engine.
Most trading bots, 3Commas, Cryptohopper, Pionex, are rule executors. You set a trigger (RSI above 70, a MACD crossover), they fire an order. That is automation, not intelligence.
AutoCoin is a different architecture. The brain runs a consensus voting engine. Seven independent signals each vote bull, bear, or chop on every scan, and the portfolio only moves when the votes converge.
The portfolio only moves when the votes converge. No single oscillator can swing the whole book.
What that looks like in practice:
Strategies lean into momentum: concentrated mega-cap exposure, full equity allocation. The book runs with the trend.
Defensive bots flip to Treasuries (BIL, SHY, IEF), and aggressive bots cut leverage and reduce position counts. No single oscillator flip required.
Hysteresis prevents whipsaws: the regime will not flip until the new vote holds for several consecutive scans, and inverse-volatility weighting shrinks any position where vol spikes.
The brain runs continuously in the cloud, monitoring markets through the trading day and locking in tomorrow's positioning between sessions, so your orders are queued and ready the moment the bell rings. Every order routes through your own brokerage account. You never give us your funds, and we never need them. Your account stays in your name, under your broker's custody.
Why I built AutoCoin
I built AutoCoin because the best trading strategies I respected were all locked behind a $100,000 minimum and a fee structure called 2-and-20, meaning the fund takes 2% of your capital up front, then 20% of any profits they make for you.
The methods themselves are not secret. The research is public. What no one had built was a way for normal traders to actually run these strategies on their own brokerage account, without handing their money over to a fund manager.
Your account. Your funds. The strategies that actually work.
James Warwick, Founder & CEO
Defensive to aggressive. 11 ways to run it.
Arranged from capital preservation to concentrated momentum. Every strategy is plainly described so choosing the mix is you setting your own course.
AtlasStocksThe all-weather 60/40 backbone. Carries the load through every season.
AegisStocksRisk-parity shield. Balances by risk contribution, not headlines.
ApolloStocksSteady index tracking with disciplined rebalancing. The straight path.
OracleStocksDollar-cost averaging into the index. Boring on purpose, powerful over time.
HestiaStocksTreasury income. The hearth fire: capital preservation with yield.
AthenaStocksSmart-beta factor blend. Quality and momentum, weighed with wisdom.
DemeterStocksDividend income harvest. Cash-generating names, compounded patiently.
MidasStocksGold/stocks dual momentum. Holds gold when gold leads, rotates to stocks (up to 3x) when stocks lead.
Nemesis MegacapStocksThe stock-side flagship. Concentrated momentum across the megacaps, regime-aware, cash when the trend breaks.
AresStocksAggressive sector rotation. Goes where the strength is, sector by sector.
HermesStocksMomentum growth engine. Fast-moving growth names with strict position caps.
- Gold versus stocks dual momentum: it holds gold when gold leads.
- When stocks lead, it rotates into the stocks leg with up to 3x leverage.
- Expect long stretches out of gold when gold is weak. The rotation is the strategy.
The published factor research the biggest quant funds rely on, applied to your account.
Packaged for retail accounts. Nothing here is proprietary magic. These are published factor premia, run with discipline on your own brokerage.
Every strategy is built on published, peer-reviewed research, not a black box. The factor premia below are the same families of effects studied for decades and run at the largest quant funds in the world. Each per-strategy backtest below lists its own specific citations under Universe, methodology and provenance.
Winners keep winning over intermediate horizons. The momentum strategies rank and rotate into relative strength, with regime gating to step aside when the trend breaks.
- Carhart (1997), On Persistence in Mutual Fund Performance
- Asness, Moskowitz, Pedersen (2013), Value and Momentum Everywhere
- Moskowitz, Grinblatt (1999), Do Industries Explain Momentum?
Balance a portfolio by each asset's contribution to risk rather than by dollars, so no single asset class dominates the outcome across inflation, deflation, and growth regimes.
- Dalio, Bridgewater, All-Weather framework
- Maillard, Roncalli, Teiletche (2010), Equally-Weighted Risk Contribution Portfolios
- Markowitz (1952), Portfolio Selection
Quality, value, and low-volatility tilts capture premia that plain index exposure leaves on the table, with disciplined, low-turnover rebalancing.
- Fama, French (1993, 2015), three- and five-factor models
- Asness, Frazzini, Pedersen, Quality Minus Junk
- Frazzini, Pedersen, Leverage Aversion and Betting Against Beta
Low-cost broad index exposure and Treasury laddering with duration that adjusts to the rate environment. Boring on purpose, powerful over time.
- Bogle (1976), the case for index investing
- Macaulay (1938), duration theory and the term structure
AutoCoin is software that runs these published methods on your account. It is not investment advice, and the research cited describes general factor premia, not a promise about any individual strategy's future results.
Full numbers, worst drawdowns, and a record you can verify.
Nemesis MegacapBacktestUniverse, methodology and provenance
Universe: Top 30 US mega-caps. Top 4 selected monthly by composite 3m+6m+12m momentum. Risk-off (SPY below 200d SMA + bearish EMA cross) routes to BIL.
Method: Carhart (1997) - On Persistence in Mutual Fund Performance. Asness, Moskowitz, Pedersen (2013) - Value and Momentum Everywhere.
Source: Re-run 2026-05-27 on extended 6.4-year window via AutoCoin's extended backtest harness against yfinance daily closes. Replaces the earlier 2020-2024 backtest.
Backtested results are simulations computed on historical data, not live trading. Past performance does not predict future results. Trading involves substantial risk of loss.
AresBacktestUniverse, methodology and provenance
Universe: SPDR sector ETFs (XLK, XLV, XLF, XLY, XLI, XLP, XLE, XLU, XLB, XLRE, XLC) + leveraged ETF tilt to TQQQ (3× Nasdaq) or SSO (2× S&P) based on top sector. Vol-scaled leverage allocation, capped at 65% in the leveraged sleeve. Risk-off routes to BIL.
Method: Moskowitz & Grinblatt (1999) - Industry Momentum. Asness, Frazzini, Pedersen (2012) - Leverage Aversion and Risk Parity.
Source: Re-run 2026-05-27 on extended 6.4-year window via AutoCoin's extended backtest harness. Drawdown widened vs the 2020-2024 window because 2025 included a deeper leverage-amplified pullback.
Backtested results are simulations computed on historical data, not live trading. Past performance does not predict future results. Trading involves substantial risk of loss.
HermesBacktestUniverse, methodology and provenance
Universe: 30 liquid US large/mid-cap growth + tech names (AAPL, MSFT, NVDA, AMZN, GOOGL, META, TSLA, AVGO, LLY, JPM, V, UNH, MA, ORCL, ADBE, AMD, CRM, NOW, SHOP, PLTR, QCOM, AMAT, INTU, PANW, ANET, NFLX, BKNG, ABNB, UBER, DDOG). Top 10 by composite 3m+6m+12m momentum, equal-weighted at 10% each. Risk-off routes to BIL.
Method: Carhart (1997) momentum. Maillard, Roncalli, Teiletche (2010) - Equally-Weighted Risk Contribution Portfolios.
Source: Re-run 2026-05-27 on extended 6.4-year window. CAGR slightly lower than the prior 2020-2024 number (32.0%) because the 2025 choppy regime hurt momentum rotation strategies, but still very strong absolute return.
Backtested results are simulations computed on historical data, not live trading. Past performance does not predict future results. Trading involves substantial risk of loss.
AtlasBacktestUniverse, methodology and provenance
Universe: SPY, QQQ, VEA (equity sleeve) + BIL, IEI, TLT (Treasury sleeve). Static 60/40 normally, defensive tilt to 50/50 in risk-off regimes.
Method: Dalio / Bridgewater All-Weather framework. Markowitz (1952) - Portfolio Selection.
Source: Re-run 2026-05-27 on extended 6.4-year window. CAGR slightly improved vs the prior 2020-2024 window thanks to the 2025 rally.
Backtested results are simulations computed on historical data, not live trading. Past performance does not predict future results. Trading involves substantial risk of loss.
AegisBacktestUniverse, methodology and provenance
Universe: SPY, TLT, GLD, DBC, VNQ - five non-correlated asset class proxies. Inverse-volatility weighting using trailing 60-day daily-return std. Monthly rebalance.
Method: Maillard, Roncalli, Teiletche (2010) - Risk Contribution Portfolios. Dalio / Bridgewater All-Weather framework.
Source: Re-run 2026-05-27 on extended 6.4-year window via AutoCoin's backtest harness + AutoCoin's extended backtest harness. Drawdown reflects the 2022 stock+bond correlated crisis which hurt risk-parity weighting.
Backtested results are simulations computed on historical data, not live trading. Past performance does not predict future results. Trading involves substantial risk of loss.
ApolloBacktestUniverse, methodology and provenance
Universe: SPY + 11 SPDR sector ETFs. 75% SPY core + 25% spread across top-3 sectors by 3m+6m momentum. Risk-off routes to BIL.
Method: Moskowitz & Grinblatt (1999) sector momentum. Bogle (1976) on index investing.
Source: Re-run 2026-05-27 on extended 6.4-year window. CAGR notably below the prior 2020-2024 number because the 2025 chop misled the sector-momentum signal repeatedly. Strategy is conservative-tier specifically because of this exact dynamic; rotation works in trending markets, drags in choppy ones.
Backtested results are simulations computed on historical data, not live trading. Past performance does not predict future results. Trading involves substantial risk of loss.
AthenaBacktestUniverse, methodology and provenance
Universe: 50-name US large-cap basket spanning sectors. Top 10 selected quarterly by combined factor score (12-1 month momentum z-score + inverse 252-day volatility z-score).
Method: Fama & French (1993) three-factor. Asness, Frazzini, Pedersen (2014) - Quality minus Junk. Frazzini & Pedersen (2014) - Betting Against Beta. Combines momentum and low-vol factors which have strong long-run premiums.
Source: Re-run 2026-05-27 on extended 6.4-year window via AutoCoin's backtest harness + AutoCoin's extended backtest harness. Replaces the earlier estimate.
Backtested results are simulations computed on historical data, not live trading. Past performance does not predict future results. Trading involves substantial risk of loss.
DemeterBacktestUniverse, methodology and provenance
Universe: SCHD, JEPI, JEPQ, VIG, SPHD. Yield-tilted basket with defensive tilt (heavier SCHD/SPHD) in risk-off regimes.
Method: Fama & French (1993) three-factor model. Litzenberger & Ramaswamy (1979) - dividend yield research.
Source: Re-run 2026-05-27 on extended 6.4-year window. CAGR notably improved vs the prior 2020-2024 number - adding 2025 rally compounded the dividend basket nicely.
Backtested results are simulations computed on historical data, not live trading. Past performance does not predict future results. Trading involves substantial risk of loss.
OracleBacktestUniverse, methodology and provenance
Universe: 60% SPY + 20% QQQ + 20% VEA. Set-and-forget buy-and-hold, no exit logic.
Method: Bogle (1976) on index investing. Malkiel (1973) - A Random Walk Down Wall Street.
Source: Re-run 2026-05-27 on extended 6.4-year window. CAGR notably improved vs the prior 2020-2024 number because the 2025 equity rally added meaningfully to the index position.
Backtested results are simulations computed on historical data, not live trading. Past performance does not predict future results. Trading involves substantial risk of loss.
HestiaBacktestUniverse, methodology and provenance
Universe: BIL, SHY, IEI, IEF, TLT. Duration auto-adjusts to the rate environment via 10y yield momentum (^TNX 90-day change).
Method: Macaulay (1938) duration theory; standard fixed-income ladder construction.
Source: Re-run 2026-05-27 on extended 6.4-year window. The CAGR is depressed because the window contains the 2022 US Treasury bond crash (worst since the 1980s) plus continued bond weakness into 2025. Cash-equivalent positioning behaves as expected: low return, low drawdown, low Sharpe.
Backtested results are simulations computed on historical data, not live trading. Past performance does not predict future results. Trading involves substantial risk of loss.
MidasBacktestUniverse, methodology and provenance
Universe: GLD (1× gold), SPXL (3× S&P 500 leveraged ETF), BIL (1-3mo T-Bills). Three-asset rotation with asymmetric leverage - full 3× exposure on the equity side when stocks lead, unleveraged gold when gold leads, T-bills when both are bleeding.
Method: Antonacci (2014) Dual Momentum + leverage-asymmetry tuning from a 31-variant factor isolation sweep. Compare GLD vs SPY 6-month total return at each quarterly rebalance. If SPY wins, deploy 100% into SPXL (3× leveraged S&P). If GLD wins, deploy 100% into GLD. If both negative, BIL. Factor isolation showed: (1) leverage on the stocks side adds CAGR cleanly, (2) leverage on the gold side via NUGT collapses to -94% DD because of miner ETF decay, (3) shorting via inverse ETFs uniformly destroys returns, (4) tight trailing stops cost more than they save, (5) quarterly rebalance beats monthly. C2 stacks only the winning factors.
Source: Backtest harness AutoCoin's Midas factor-sweep harness - 31 variants tested with single-factor isolation followed by stacked combos of the winners. C2_quarterly_3xstocks wins the combo sweep at 17.3% CAGR, Sharpe 0.68. Run 2026-05-27 against yfinance.
Backtested results are simulations computed on historical data, not live trading. Past performance does not predict future results. Trading involves substantial risk of loss.
These are the same backtest numbers shown inside AutoCoin Pro: full windows, methodology, and the worst drawdown printed at equal weight to the return. Estimates are labeled as estimates.
A backtest tells you what a strategy did in a lab. The other half of the story runs in the open: AutoCoin Live shows the system trading in real time, paper and live clearly labeled, with nothing edited after the fact.
Every daily result is hashed, signed, and chained into a record you can verify yourself, instead of taking our word for it.
How we compare to hedge funds and other bots.
Same strategies. Better economics. Your custody.
| AutoCoin | Hedge Fund | Other Bots (3Commas, Cryptohopper, Pionex) | |
|---|---|---|---|
| Decision engine | AI consensus voting: 7 independent signals voting bull, bear, or chop, hysteresis-protected. | Human discretion. Subject to fatigue, emotion, and bias. | Single-rule triggers (RSI above 70, MACD crossover). No AI. One bad signal flips your whole book. |
| Where your money lives | 100% in your own brokerage account, in your name. | You wire your capital to them. They hold it. You have to ask to get it back. | Your exchange. AutoCoin never has it either, but exchange-side risk is real. |
| Cost | Flat $149/mo. No performance fee, ever. | 2% of capital annually plus 20% of every profit. On a $100K account up 20%, the fund keeps $6K. | Tiered subscriptions ($20 to $500/mo). Pay more as your account grows. |
| Minimum capital | $5K recommended. No enforced minimum. | $100K to $1M and up. Out of reach for almost everyone. | Any size, but bot quality is rarely worth it under $1K. |
| Methodology | Fully disclosed. Every strategy lists its factor model plus academic citations (Carhart 1997, Asness 2013, and more). | Black box. Trust us. Most quant funds do not disclose strategy at all. | No methodology. You set your own trigger. There is no underlying research. |
| Asset coverage | Stocks plus bonds plus crypto, all under one subscription. | Single mandate (for example equity-only or crypto-only). | Crypto only. No stocks. No bonds. |
| Lock-up period | None. Cancel any time, withdraw any time. | Quarterly redemption windows. Year-long lockups are not uncommon. | None, but switching costs are real (re-set up every trigger from scratch). |
| Strategy transparency | Every strategy explained, every ticker named, backtested across 2008, 2020, and 2022. | Opaque. You do not know what they hold or why. | Buy when X. No risk model, no regime detection, no drawdown management. |
AI consensus voting: 7 independent signals voting bull, bear, or chop, hysteresis-protected.
Human discretion. Subject to fatigue, emotion, and bias.
Single-rule triggers (RSI above 70, MACD crossover). No AI. One bad signal flips your whole book.
100% in your own brokerage account, in your name.
You wire your capital to them. They hold it. You have to ask to get it back.
Your exchange. AutoCoin never has it either, but exchange-side risk is real.
Flat $149/mo. No performance fee, ever.
2% of capital annually plus 20% of every profit. On a $100K account up 20%, the fund keeps $6K.
Tiered subscriptions ($20 to $500/mo). Pay more as your account grows.
$5K recommended. No enforced minimum.
$100K to $1M and up. Out of reach for almost everyone.
Any size, but bot quality is rarely worth it under $1K.
Fully disclosed. Every strategy lists its factor model plus academic citations (Carhart 1997, Asness 2013, and more).
Black box. Trust us. Most quant funds do not disclose strategy at all.
No methodology. You set your own trigger. There is no underlying research.
Stocks plus bonds plus crypto, all under one subscription.
Single mandate (for example equity-only or crypto-only).
Crypto only. No stocks. No bonds.
None. Cancel any time, withdraw any time.
Quarterly redemption windows. Year-long lockups are not uncommon.
None, but switching costs are real (re-set up every trigger from scratch).
Every strategy explained, every ticker named, backtested across 2008, 2020, and 2022.
Opaque. You do not know what they hold or why.
Buy when X. No risk model, no regime detection, no drawdown management.
Your broker holds your assets. Always.
AutoCoin is a self-directed software tool. It places trades through your existing brokerage account with trade-only permissions, inside the limits you set, and you can stop it instantly.
You connect your brokerage with permissions scoped to placing trades. AutoCoin cannot withdraw, transfer, or move assets out of your account, ever.
Your shares and cash stay at your brokerage the entire time. There is no AutoCoin account to deposit into. We are software directing trades you authorized, nothing more.
Pause or disable any strategy from your dashboard in one click, or revoke access at your broker. Either way trading stops immediately. You stay in control at all times.
The stock bots trade during market hours only, with the capital and strategy mix you configured. Nothing executes outside the rules you switched on.
Run systematic strategies inside your tax-advantaged account.
This is the differentiator most platforms cannot offer: connect a retirement account and let the same disciplined strategies work inside it.
Connect Brokerage, Roth IRA, or Traditional IRA accounts at Public.com and run AutoCoin strategies inside them.
IRA support through Alpaca, a FINRA-member brokerage, alongside standard brokerage accounts.
Long-horizon strategies like dividend income, index tracking, and 60/40 fit naturally inside retirement accounts. Talk to your tax adviser about what suits your situation.
Broker-agnostic by design.
Your account. Your funds. AutoCoin never custodies a single dollar.
US FINRA-member brokerage with SIPC protection. Available today, IRA support included.
Retail-friendly API with low-cost equity execution.
Available in many countries. US users connect through Alpaca.
Brokerage plus Roth IRA and Traditional IRA accounts, the retirement-account path.
Connect Alpaca, Tradier, eToro, or Public.com, including IRA accounts on Public. Alpaca is a FINRA-member brokerage with SIPC protection at the brokerage level, eToro is available in many countries (US users connect through Alpaca), and Public.com covers Brokerage plus Roth and Traditional IRA accounts. AutoCoin requests trading permissions only, never withdrawal or fund-movement access. Availability varies by region, and the app guides you to what works where you live. You can switch brokers or disconnect at any time.
Which brokers work where you live.
Pick where you trade from and we'll show what typically fits. Guidance, never a wall: every broker stays connectable, and each one confirms eligibility when you sign up.
Pick your region above to see if this venue fits.
Pick your region above to see if this venue fits.
Pick your region above to see if this venue fits.
Available worldwide, no regional limits.
Region guidance only, never a block. Every venue stays connectable; the venue's own sign-up confirms who is eligible. Travelers, expats, and cross-border accounts are all fine.
Three ways in. Pick your door.
Pick a strategy and watch it trade simulated capital before you connect anything. Real signals, real timing, demo money. Demo results are simulations.
Watch a bot in demoNo brokerage yet? We walk you through opening an account at one of the four supported brokerages, with region-aware guidance on what works where you live.
Want in before your brokerage is set up? The in-app Solana wallet path runs the Hades crypto bot with no exchange or brokerage account at all. Same membership covers both sides.
One membership. Every strategy.
Stocks and crypto together. No tiers, no upsells, no add-ons.
- 7-day free trial ($1 sign-up fee, card required)
- All 16 strategies, crypto and stocks
- Unlimited bots across your connected venues
- Cancel anytime from your account
- Lifetime access, nothing recurring, ever
- Founding $AC allocation, set up in your account
- All 16 strategies, crypto and stocks
- Hard-capped at 500 people, then it is gone
The hard questions, answered straight.
Can AutoCoin move money out of my brokerage account?
No. When you create your brokerage API key, you enable the Trading permission and leave Withdrawals OFF. AutoCoin can place orders on your account, but cannot move funds out of it, not to our wallets, not anywhere.
Even if our servers were breached, your money stays in your brokerage account, in your name, under your broker's custody.
What happens if I cancel, to my subscription and to my open positions?
Cancellation is one click in your dashboard. The bot stops sending orders immediately. Any positions already open in your brokerage account stay exactly where they are. They are your shares, in your account.
You can hold them, sell them, or keep managing them manually directly with your broker. We never touch them again.
How much capital do I need to make this worthwhile?
Honest answer: the subscription is $149/month, so the smaller your account, the bigger the percentage drag. Below $5K, the fee will likely eat into your returns meaningfully. The strategies are designed to scale comfortably from $10K and up.
If you are starting smaller, you can still run the crypto bots, which size down comfortably, and add the stock strategies once your brokerage account grows. You can also paper-trade the stock side on your broker first to evaluate.
What happens if AutoCoin goes out of business?
Your brokerage account is yours, full stop. If AutoCoin disappeared tomorrow, you would log into your broker directly, revoke the API key, and continue managing your account exactly as before. We cannot take your account with us because we never had it.
Is the trial really $1, or am I getting charged $149 on day one?
$1 today. Nothing else for 7 days. That dollar is a card-verification charge to confirm your payment method is real, not stolen. After 7 days, your subscription rolls to $149/month, unless you cancel before then, which you can do with one click from your dashboard. No hidden charges, no surprise renewals.
Can I run the bot in my IRA or Roth account for tax efficiency?
Yes, wherever your broker supports them. Running higher-turnover strategies inside a Roth IRA is especially tax-efficient, since short-term gains that would normally be taxed as ordinary income are sheltered.
Open the IRA at your broker directly, then connect to AutoCoin the same way. This is the retirement-account differentiator most platforms cannot match.
Can I run multiple bots on the same brokerage account?
Yes. You can run several bots on the same brokerage account at once. Each bot only manages the capital you allocate to it, so they trade side by side without touching one another's slice. You decide how much of the account each bot can use, and you can add or pause bots whenever you want.
Is AutoCoin an investment advisor?
No. AutoCoin is software, not investment advice. AutoCoin LLC is a FinCEN-registered Money Services Business (MSB #31000322777334), but it is not registered with the SEC or FINRA as an investment advisor, and we do not manage anyone's account on a discretionary basis.
Every strategy is pre-built and visible to every subscriber. No one at AutoCoin recommends a strategy to you personally. You are responsible for your own trades, your account, and your local jurisdiction's rules. Past hypothetical backtest performance does not guarantee future returns.
Put discipline behind your brokerage account.
Connect in minutes or start in demo. Trade-only access, your limits, your call.
AutoCoin LLC is a software platform that provides automated execution of pre-built quantitative strategies on third-party brokerage accounts. We are not a registered investment advisor or broker-dealer with the SEC or FINRA. Users hold the brokerage relationship and full custody of their assets at all times. AutoCoin requests only trading permissions on connected accounts, never withdrawal or fund-movement permissions.
Backtest performance shown is hypothetical, not based on actual trading. Past performance does not guarantee future results. Target return and drawdown figures are forward-looking estimates based on each strategy's design and the academic literature, and actual performance will vary. All trading involves risk of loss. Leveraged strategies amplify both gains and losses.
AutoCoin LLC is a Florida limited liability company (document number L25000545515), and is registered with the US Financial Crimes Enforcement Network as a Money Services Business (MSB #31000322777334), St. Petersburg, FL, USA. Questions: admin@autocoin.ai.