Short — profits if the price falls · 9 shares
Entered at $506.74 · now $507.20
Autonomous paper-trading experiment
A rules-based trading bot that scans 8 major US stocks & ETFs three times a day using simulated money — compared, in public, against the funds most people own.
How it's doing
That puts it 3.70 percentage points ahead of the stock market over the same 4 trading days.
Early days. 4 trading days and 4 closed trades is a small sample — a few lucky or unlucky trades can swing these numbers a lot. The 1-year simulation below gives longer-term context.
Paper account value
$100,095.43
+$95.43 (+0.10%)
Started with $100,000.00 of simulated cash
Today
+$42.67
+0.04% since yesterday's close
Profit from closed trades
+$100.02
Locked-in result of 4 finished trades
Win rate
75%
3 wins · 1 loss
Average trade
+0.49R
Profit per trade measured in units of risk (R)
Money at work
4.6%
1 open position · rest in cash
Stocks are trending down, so the bot looks for short (sell) setups or stays in cash. (Based on the S&P 500 and Nasdaq-100 hourly trends.)
Short — profits if the price falls · 9 shares
Entered at $506.74 · now $507.20
Every trade it took — and every one it deliberately passed on — with its reasoning.
Opened a short (sell) trade: 9 shares of MSFT at $506.74
Short MSFT intraday in BEAR regime (SPY/QQQ 1h & 15m bearish_trend): MSFT aligned bearish_trend on 1h (EMA20 514.63 < EMA50 522.27, RSI 33.68) and 15m (EMA20 505.87 < EMA50 511.91, RSI 46.23) with 1.12x RVOL and clean…
Closed MSFT for a +$46.48 gain (+0.71R)
Closed before the closing bell (end-of-day exit)
Closed SPY for a +$44.90 gain (+1.23R)
Closed before the closing bell (end-of-day exit)
Opened a short (sell) trade: 9 shares of MSFT at $519.04
MSFT 1h/15m aligned downtrend with 1.84x breakdown RVOL in BEAR regime; kill-criterion: 15m close above EMA20 (.84)
Passed on a short setup in NVDA
Disqualified by catalyst rubric: abrupt CEO resignation (Jensen Huang), congressional/Commerce China chip probes, and Trump AI 'National Champions' headline whipsaw risk; avoids compounding short tech exposure with ex…
Opened a short (sell) trade: 6 shares of SPY at $745.79
Intraday short aligned with broad BEAR regime (SPY/QQQ 20<50 EMA downtrend) and 15m elevated RVOL (1.6x). Kill-criterion: 15m close above EMA(20) 751.49 or adverse drift > 0.15x ATR.
Passed on a short setup in MSFT
Rejected by risk_guard.py: bid/ask spread too wide (0.825% > 0.20% max).
Passed on a short setup in NVDA
Disqualified by catalyst rubric: abrupt CEO resignation (Jensen Huang), ongoing federal/congressional chip export probes, and imminent earnings report.
Closed META for a +$13.79 gain (+0.17R)
Closed before the closing bell (end-of-day exit)
Opened a long (buy) trade: 7 shares of META at $639.32
Bull Case: Multi-timeframe aligned 20>50 EMA uptrend on 1h/15m with strong 1.54x RVOL and price holding above EMA(20). Bear Kill-Criterion: Invalidate and exit if META closes below 15m EMA(20) (~.29) or if QQQ breaks…
Closed SPY for a -$5.15 loss (-0.14R)
Closed before the closing bell (end-of-day exit)
Opened a long (buy) trade: 6 shares of SPY at $760.55
Bull regime trend continuation; aligned 1h/15m EMA uptrend with 1.18x RVOL. Kill-criterion: 15m close below EMA(20) at 759.93 or market regime break.
Passed on a buy setup in QQQ
Skipped in favor of higher-ranked SPY (1.18x RVOL vs 1.12x) and superior 15m EMA(20) hold; avoids redundant index exposure amid tech headline overhang.
Cumulative return since the bot went live on Nov 11, 2026. Every line starts at 0% so you can compare growth directly, regardless of share price.
Tap a fund to add or remove it. Hover or use arrow keys on the chart for exact values.
Percentage points the bot is ahead (▲) or behind (▼) each fund since launch.
Profit factor = total gains ÷ total losses. Above 1.0 means winners outweigh losers.
| Your bot (live) | Paper bot | +0.10% | −0.02% | — | — | — |
|---|---|---|---|---|---|---|
| SPY · S&P 500 | US Index | −3.60% | −3.60% | — | — | — |
| DIA · Dow Jones | US Index | −3.26% | −3.26% | — | — | — |
| QQQ · Nasdaq-100 | US Index | −4.83% | −4.83% | — | — | — |
| IWM · Russell 2000 | US Index | −4.99% | −4.99% | — | — | — |
| JEPI · JPM Equity Prem Inc | J.P. Morgan | −2.70% | −2.70% | — | — | — |
| JGRO · JPM Active Growth | J.P. Morgan | −5.03% | −5.03% | — | — | — |
| BBUS · JPM BetaBuilders US | J.P. Morgan | −3.81% | −3.81% | — | — | — |
| VTI · Vanguard Total Mkt | Broad Index | −3.72% | −3.72% | — | — | — |
This is a walk-forward simulation of the same rules on past prices — not real trades. Simulations usually look better than reality.
Tap a fund to add or remove it. Hover or use arrow keys on the chart for exact values.
Up = more return. Left = smoother ride. The top-left corner is where you want to be.
Percentage points the 4× simulated bot finished ahead (▲) or behind (▼) each fund over 1 year.
| Bot · 1× cautious sizing | Simulation | +3.47% | −1.00% | 2.03 | 3.11 | 1.77% |
|---|---|---|---|---|---|---|
| Bot · 4× larger sizing | Simulation | +13.26% | −4.20% | 1.86 | 2.88 | 7.12% |
| SPY · S&P 500 | US Index | +10.94% | −9.13% | 0.83 | 1.17 | 13.71% |
| DIA · Dow Jones | US Index | +9.96% | −10.07% | 0.78 | 1.15 | 13.52% |
| QQQ · Nasdaq-100 | US Index | +10.20% | −14.69% | 0.56 | 0.78 | 21.49% |
| IWM · Russell 2000 | US Index | +15.19% | −11.84% | 0.86 | 1.25 | 18.71% |
| JEPI · JPM Equity Prem Inc | J.P. Morgan | +0.59% | −7.70% | 0.11 | 0.15 | 9.56% |
| JGRO · JPM Active Growth | J.P. Morgan | −5.77% | −14.09% | -0.20 | -0.28 | 19.85% |
| BBUS · JPM BetaBuilders US | J.P. Morgan | +10.14% | −9.47% | 0.77 | 1.08 | 13.93% |
| VTI · Vanguard Total Mkt | Broad Index | +11.06% | −9.17% | 0.83 | 1.17 | 13.99% |
The bot writes down its reason for every trade and the condition that would prove it wrong (its "kill criterion").
| Why it traded | ||||||||
|---|---|---|---|---|---|---|---|---|
| MSFT Nov 17 | SHORT | 9 | $506.74 | — | — | — | Open | Short MSFT intraday in BEAR regime (SPY/QQQ 1h & 15m bearish_trend): MSFT aligned bearish_trend on 1h (EMA20 514.63 < EMA50 522.27, RSI 33.68) and 15m (EMA20 505.87 < EMA50 511.91, RSI 46.23) with 1.12x RVOL and clean earnings/binary flags. Kill-criterion: 15m close > .50 or SPY/QQQ reclaiming 15m EMA20 (.53/.20), or 3:50 PM ET EOD flatten. |
| MSFT Nov 16 | SHORT | 9 | $519.04 | $513.88 | +$46.48 | +0.71R | End-of-day exit | MSFT 1h/15m aligned downtrend with 1.84x breakdown RVOL in BEAR regime; kill-criterion: 15m close above EMA20 (.84) |
| SPY Nov 16 | SHORT | 6 | $745.79 | $738.31 | +$44.90 | +1.23R | End-of-day exit | Intraday short aligned with broad BEAR regime (SPY/QQQ 20<50 EMA downtrend) and 15m elevated RVOL (1.6x). Kill-criterion: 15m close above EMA(20) 751.49 or adverse drift > 0.15x ATR. |
| META Nov 13 | LONG | 7 | $639.32 | $641.29 | +$13.79 | +0.17R | End-of-day exit | Bull Case: Multi-timeframe aligned 20>50 EMA uptrend on 1h/15m with strong 1.54x RVOL and price holding above EMA(20). Bear Kill-Criterion: Invalidate and exit if META closes below 15m EMA(20) (~.29) or if QQQ breaks session low on volume. |
| SPY Nov 12 | LONG | 6 | $760.55 | $759.70 | −$5.15 | -0.14R | End-of-day exit | Bull regime trend continuation; aligned 1h/15m EMA uptrend with 1.18x RVOL. Kill-criterion: 15m close below EMA(20) at 759.93 or market regime break. |
Three times every trading day (7:05, 9:05 & 11:05 AM PT) it checks SPY, QQQ, NVDA, AAPL, MSFT, AMZN, META, TSLA for strong trends that agree on both the 1-hour and 15-minute charts, backed by above-average trading volume.
It checks whether the S&P 500 and Nasdaq-100 are trending up, down, or sideways — and only trades with that tide (buying in uptrends, shorting in downtrends).
It skips anything with earnings coming up, lawsuits, CEO departures, investigations, or other surprise risks that could cause a sudden gap.
No single trade can lose more than about 2.0% of the account. Max 5 positions at once, each under 25% of the account.
Every order is placed with an automatic stop-loss (1.5× the stock's typical daily move) and a profit target twice as far away. Everything is closed by 12:40 PM PT, so nothing is held overnight.
Trading with simulated money on real, live market prices. The results are real measurements of the strategy, but no actual dollars are bought, sold, or at risk.
A standard to compare against. Here that's popular funds like the S&P 500 (SPY) — if the bot can't beat simply buying one of these, it isn't adding value.
How much the account grew or shrank, as a percentage of where it started.
The plain difference between two percentages. If the bot is up 3% and the S&P 500 is up 1%, the bot is 2 points ahead.
The biggest fall from a high point to a later low. It answers: “How bad did it get before recovering?” Smaller is better.
Return earned per unit of bumpiness. Above 1 is good, above 2 is very good. It rewards steady gains over wild swings.
Like Sharpe, but it only counts downside swings as risk — upward surprises aren't penalized.
How much daily returns swing around, scaled to a year. The S&P 500 is typically around 15–20%.
Profit measured in units of the risk taken. If a trade risked $100 and made $200, that's +2R. Losing the full planned risk is −1R.
Long = buy now, profit if the price rises. Short = borrow and sell now, profit if the price falls.
Replaying the exact rules day by day on past prices, using only information that was available at the time — no peeking at the future.