A real Grok Bot, built by Joe Apfelbaum
Joe's Trading Desk
The Trading Desk is a bot Joe built with Grok Bot to help him learn options trading the careful way: research first, then backtests, then paper trading, with a short lesson every morning. This page shows what the desk does, how it helps Joe learn, and everything he is learning along the way.
Not financial advice. Educational only. This is not financial, investment or tax advice. Options involve substantial risk and are not suitable for everyone. Backtests and paper trading results do not predict live results. The Trading Desk paper trades only, and no real trade is placed without Joe's explicit OK.
What the Trading Desk does
Think of it as a research assistant and practice partner for options, not a trader. It handles the reading, testing and record keeping so Joe can focus on understanding.
1Backtests in Option Alpha
Turns a trading idea into written rules, then tests those rules on years of historical options data in Option Alpha, a no-code options automation platform. It starts with the simplest version and changes one setting at a time.
2Paper-trading bots
Builds Option Alpha bots on a paper (simulated) account only. Nothing trades with real money, and no live trade happens without Joe saying yes first.
3Research
Checks claims against primary sources. Its first project compared SPX, SPY and XSP options: contract size, settlement, exercise style, fees and tax treatment.
4Risk checks
Before any bot runs, it checks the worst-case loss on each trade, how much risk is stacked on the same market, and a pre-agreed rule for when to pause a bot.
5Trade journal
Logs every paper entry, skip and exit next to what the backtest expected, so differences show up quickly and get explained.
6Daily accountability email
Sends a short evening email to Joe and his brother, who is learning alongside him, covering what was studied and what comes next.
7Daily 6:30 AM lesson
Every morning at 6:30 AM ET it teaches Joe one options concept in plain language and adds the takeaway to a running learning log.
!The guardrail
Paper trading only. The desk researches, tests and explains. Joe makes every decision, and no live trade or money movement happens without his explicit OK.
Built with Grok Bot. The Trading Desk is a Grok Bot with its own instructions, files and daily routines. It keeps its research and logs in a shared folder, runs scheduled lessons and emails, and checks with Joe before anything important.
How it helps Joe learn
The goal is understanding before money. The desk turns learning into a daily habit with a paper trail.
1One lesson a day
A single concept each morning, small enough to absorb before the workday starts, and linked back to the desk's research notes.
2A written learning log
Every lesson becomes one line in a log, newest first, so Joe can see what he has covered and spot gaps. The log below is generated from it.
3Accountability
The evening email keeps a second person in the loop. Learning with someone else makes it far more likely that the lessons actually happen.
4Claims get fact-checked
When Joe hears a trading claim, the desk researches it from primary sources and explains which parts are right, which are wrong, and why.
5Practice before money
Ideas go through backtests and weeks of paper trading first. Joe learns how strategies behave, including losing stretches, without risking capital.
6A weekly review
Once a week the desk compares paper results with the backtest, checks every bot against its written rules, and batches any changes instead of reacting to one bad day.
What Joe is learning
Summaries of the main topics in the desk's research notes so far. These are general concepts, not recommendations.
How Option Alpha works
- Option Alpha is a no-code, cloud-based platform. Bots run scanners, monitors and triggers that open and manage options positions.
- Bots trade defined-risk structures only: spreads, iron condors and similar. Naked options, strangles, calendars and the wheel are not supported.
- The backtester covers a small set of underlyings, such as SPX, SPY, QQQ, IWM and XSP, with intraday data going back to 2013.
- A backtest can be turned straight into a bot, which keeps the live rules identical to the tested ones.
The workflow: idea, backtest, paper bot, review
- Write the idea as rules first: underlying, structure, days to expiration, strikes, entry time, exits, size and the largest acceptable loss.
- Run a baseline backtest with minimal filters and pessimistic fill assumptions.
- Vary one thing at a time and compare the variations side by side.
- Paper trade the bot for 30 to 60 days or more, then review it weekly. Going live is a separate decision that needs Joe's explicit OK.
Backtest hygiene
- Start simple. Every extra filter has to earn its place.
- Look for plateaus, not peaks: good results across nearby settings. A result that collapses when a setting moves slightly is probably overfit.
- Treat small samples, roughly under 100 trades, as anecdotes rather than evidence.
- Be pessimistic about slippage and fees, and hold out the most recent year of data to check the rules on data they were not designed on.
Paper trading before going live
- Option Alpha recommends at least 30 to 60 days of paper trading at the size you would use live.
- Paper trading proves the logic does what you intended. It does not prove the fills, because paper fills near the midpoint and is never assigned.
- The desk uses a graduation checklist: enough trading days and closed trades, every action matching the written rules, and no unrealistic fills.
- Even when the checklist passes, going live still needs Joe's OK.
Sizing and portfolio risk
- Size by maximum loss, not by number of contracts. A common guideline is risking about 1 to 3 percent of an account per position.
- Add up risk across all bots. Several bots on the S&P 500 are really one bet on the same market.
- Set a kill switch before a bot starts, for example pausing it if its drawdown goes past a set multiple of the backtest's worst drawdown.
- Expect live drawdowns to be larger than backtests show.
SPX vs SPY vs XSP
- SPX is an index, not a share. Its options use the same 100x multiplier as SPY, so one SPX contract is about ten times the size of one SPY or XSP contract.
- SPX and XSP options are cash-settled and European-style, so there is no early assignment. SPY options are physically settled and can be assigned early.
- On some expirations SPX has two series at the same strike (AM- and PM-settled). Both legs of a spread should come from the same series.
- Index options and ETF options can be taxed differently in the US. This is not tax advice, and the desk's rule is to confirm anything tax-related with a CPA.
Same-day (0DTE) options
- Daily S&P 500 expirations only go back a few years, so long-history 0DTE results are not meaningful.
- Option Alpha's own research found that only about half of its 0DTE traders were profitable, roughly a coin flip.
- Stops are triggers at the midpoint, not guaranteed exit prices. On fast days, losses can exceed the stop.
- Cash-settled products fit anything held to the close better than physically settled ones.
Common beginner mistakes
- Copying a leaderboard bot and running it live without paper trading or understanding it.
- Trusting a backtest with many filters and few trades, or with zero slippage and no fees.
- Keeping physically settled spreads open into expiration, where bots cannot see an assignment.
- Changing a bot after one bad trade. Option Alpha suggests judging changes on months of data, not days.
The weekly review routine
- Reconcile what the platform shows with what actually happened, including any manual overrides.
- Score each bot against its backtest, its rules and its kill switch.
- Check next week's calendar for market-moving events and confirm each bot's skip settings.
- Write down proposed changes but do not apply them yet. Batch them for a quarterly review.
Lesson log
One line per daily lesson, newest first.
- Glossary
Delta
Delta is how much an option's price moves per 1-point move in the stock, and it is also a rough gauge of the chance of expiring ITM (a 0.16-delta short put is roughly 84% likely to expire OTM)
- Glossary
ITM, ATM, OTM (moneyness)
Moneyness compares strike to stock price; only ITM options have intrinsic value, and OTM options are pure time value that premium sellers hope expires worthless
- Q&A
Strike, bid/ask, time decay, options vs futures
Strike is the contract price; time decay helps sellers; options are rights while futures are obligations
Not financial advice. Educational only. This is not financial, investment or tax advice. Options involve substantial risk and are not suitable for everyone. Backtests and paper trading results do not predict live results. The Trading Desk paper trades only, and no real trade is placed without Joe's explicit OK.
Build your own bots with Grokbot
Joe built the Trading Desk with Grok Bot. In the Grokbot School community on Skool he teaches how to set up bots like this one step by step, with ready-to-use prompts, workflows and live sessions. $29/month.