FAMBF is the framework I built for self-directed traders who already have a strategy they trust on their own account, but spend half their week missing signals or talking themselves out of taking them. A signed risk mandate before any code gets written. Real execution that doesn't fall over on reconnects. Logs you can read without a translator. A kill switch on your phone. The money never moves off your exchange, and you walk away with the source code.
By the time you've been at it for a few years, it's rarely the strategy that does the damage. Most of the loss comes from the gap between what the rules say to do and what you actually end up doing at the screen at 3am, or in a meeting, or while the market is moving faster than you can think.
You check the chart twenty minutes later and the entry is already gone. Or worse, you're holding a position the market handed you while you were busy with something else. Some weeks this washes out. Other weeks it's the difference between a clean month and one you don't want to talk about.
Closing the winner two ticks early because you got nervous. Hanging onto the loser because surely it'll come back. Skipping the rule today because today, just today, the setup looks different. None of these feels expensive in the moment. Add them up across a year and they're often most of what a working strategy was supposed to make.
Most retail trading bots ship as compiled code or a SaaS dashboard you can't open. When something goes wrong, there's nothing to look inside. You won't know which filter killed the signal, why your position has suddenly doubled, or whether the system is still talking to the exchange at all. You realise something is off from the P&L, and by then the damage is done.
Custody-as-a-service feels convenient on the day everything is fine. The day to think about is the one when it isn't. The broker freezes deposits, the counterparty bank gets bailed in over a weekend, the exchange decides your jurisdiction has become inconvenient. I lost about ten years of work to one such day in April 2013, which is more or less why this framework exists.
Every engagement goes through the same five stages, in this order. The order matters. We don't compress it and we don't skip steps, even when a client wants to be live by Friday.
Your rules written down clearly enough for a junior developer to implement them from your description alone. If they can't be written that clearly, they aren't ready to be automated, and we'll say so before any code is committed.
A short document we both sign before deployment. It lays out your daily loss cap, the maximum position size, how many averaging steps you actually want allowed, and what conditions trigger a hard stop. Nothing in it needs to be exotic. The point is that it exists on paper before any live order is placed.
The parts most retail bots get wrong. Idempotent order submission, so the exchange itself refuses duplicates. State reconciliation after every reconnect, including the awkward case where the bot thinks an order failed but it actually filled. Stale signals discarded. Slippage tracked per fill, and trading pauses if it gets ugly.
Paper first, so we can watch what the system does with nothing at stake. Then shadow mode, where it sees live market data but never sends a real order. Then micro-live at a size that wouldn't hurt to lose entirely. Full size only after micro-live has actually held up, and never on a fixed schedule.
Two screen-shared sessions where you learn to run the system yourself. We go through how to read the logs, what the kill switch actually does, and how to change parameters without breaking anything that's running. The most important part takes the most practice: spotting the early signs of drift before the equity curve has to be the one to tell you.
A lot of what gets sold as "trading automation" comes with promises that age badly. Below is the list of things this framework will not do, in plain language. None of them are negotiable.
Your money stays on the exchange you already use, in an account only you can log into. The framework can see open positions and place new orders against them. That is the entire surface area. No escrow accounts, no shared wallets, nothing that gets called a "managed" anything.
Every API key the framework creates is set up without permission to move funds off the exchange. If the system is ever compromised, whether by an attacker, a leaked secret or a bad library update, the worst case is that positions change. The balance stays where it is.
You won't find an annualised yield, a Sharpe target, or a monthly performance figure in any of our material. Returns are a property of your strategy, which is your responsibility. The framework handles execution, which is what we're actually selling.
The system runs the rules your Risk Mandate authorises. Nothing else. We don't take positions for you on a hunch, and there's no one at our end logging into your account to adjust something manually.
You receive the source code we write for your engagement in plain, readable form. You can inspect it, fork it, or hand the whole thing to a different developer six months from now if you decide to. There's no SaaS dependency to lock you in.
If the first question on the call is "how many percent per month", that's the end of the conversation. FAMBF is for traders who already know what their strategy does and want to run it with more discipline than they can hold by hand.
Fourteen questions to run against your own setup, which takes about half an hour. Each one comes from a specific failure mode I've personally watched destroy retail trading accounts in the past year. The PDF is free and there's no email gate. Read it, apply what's relevant, and decide whether your bot has the gaps it covers.
Each question is followed by what the failure mode looks like in practice, what a properly built system does instead, and a score line for tallying up where your own setup actually stands. The list covers reconnect duplicates, stale signals, withdrawal API leaks, slippage drift, missing kill switches, and ten others I've seen in the wild.
Read the checklist
Start small if you're not sure. Most clients begin with the Diagnostic and move into the Sprint once we both agree the strategy is worth automating in the first place.
A short engagement to find out whether your current setup is actually ready for automation. Strategy, rules, and existing infrastructure are reviewed before any code is committed.
21 days from a written strategy to a system running on your account at micro size. Includes the full launch sequence through paper and shadow modes before any real money is at risk.
For traders who already have automation running. We audit it against the 14-point checklist, then fix the gaps most likely to cost real money in a stress event.
No. FAMBF is fixed-fee work. The fee covers the implementation, the operator training, and whatever months of post-launch support you choose to include. We don't take a percentage of your trading results in any form. If our income depended on your P&L, we'd be under pressure to push for trades that aren't in your Risk Mandate, and that's just not the business I want to run.
Only the API permissions that are needed to place orders for you. That's typically "trade" and "read". Withdrawal is explicitly disabled on every key the framework uses. You stay in control of the key and can revoke it at any time from your exchange settings.
That's exactly what the Diagnostic is for. If we go through your rules and they can't be written down cleanly, or the historical edge falls apart under a closer look, we'll say so before any code is written. Roughly one in three diagnostic engagements ends with us telling the trader the strategy isn't ready to automate yet. That's a fine outcome, and it saves both sides a lot of time.
21 calendar days from a signed Risk Mandate to the first live trade at micro size. Specification and the Risk Mandate sit in week one. The execution engine and paper testing fill week two. Shadow mode and the first micro-live trades happen in week three. Promotion to full size depends on how the micro-live phase actually behaves, and is not on a fixed calendar.
For crypto, anything supported by the CCXT library: Binance, Bybit, OKX, Kraken, Coinbase, and most others. For US equities and options we work with Alpaca and Interactive Brokers. For Forex it's typically MetaTrader 5. If your venue isn't on this list, mention it on the diagnostic call and we'll tell you whether it's workable.
For software implementation and operator training, jurisdiction is not a barrier. FAMBF is positioned as software and education, not as an investment service. What you do with the framework on your own account is governed by your own jurisdiction's rules, and we recommend you understand them before deploying.
There's no obligation to stay on a retainer. Monthly support tiers start at $700 for standard maintenance and run to $3,000+ for private desk arrangements that include SLA, monthly risk reviews, and small improvements as they come up. Plenty of clients also choose to operate the system themselves after the training and only call when something specific needs to be done.
Yes. The code for your engagement is yours. There's no obfuscation and no SaaS dependency that locks you in. The reusable core of FAMBF is licensed to you for use with your own system. Anything we build on top for your specific strategy is yours outright.
Thirty minutes on a call to look at your current setup together and decide whether automation is actually the right next move for you. Bring whatever description of your strategy you already have on paper. There's nothing else to prepare.
Book a 30-minute call