FAQ

The answers below are written assuming you've read at least the welcome page. If a term is unfamiliar, the glossary defines it. For platform errors and bug reports, contact support; for direct help, see support.

#Platform & getting started

#Do I need to know how to code?

No. The Visual Strategy Builder is the strategy. There is no script underneath that you'd ever need to edit. Every strategy is a node graph you build by dragging and connecting nodes.

The platform is opinionated about this: there's no "advanced mode" that drops you into Python. The reasoning is that a typed visual builder catches structural errors at design time, which is faster than discovering them in a backtest log at 2 AM. The handful of things that are awkward to express in nodes are mostly things experienced quants don't want to express anyway.

#What kind of trader is this for?

Three groups, roughly:

• Discretionary traders who already have a sense of their edge and want to systematize it. AlgoLift gives them a way to encode their judgment into rules and stress-test whether it's real.
• Code-first quants tired of plumbing: fill modeling and data infrastructure. AlgoLift handles that layer so they can focus on strategy logic.
• Complete beginners who want to learn algorithmic trading from scratch. The Algotrading-101 Academy is a structured curriculum that teaches the underlying math and market mechanics, with labs in AlgoLift at every step.

If you're none of those — you're looking for a platform that hands you a pre-built winning strategy — AlgoLift is not the right place. There aren't any.

#How long does it take to build my first strategy?

About 20 minutes from a blank canvas to a complete, backtested EMA-crossover strategy. The your first strategy walkthrough is the hands-on tutorial.

#What markets and instruments are supported?

AlgoLift holds every print and the order book for these markets:

Micro contracts, other futures, ETFs and crypto aren't available yet.

#Can I trade options?

Limited support today. The platform's primary focus is futures, equities, and crypto. Options strategies that depend on greeks or volatility surfaces are not currently first-class.

#What brokers can I connect to?

None today. AlgoLift doesn't connect to brokers or place trades. Live execution is on the roadmap as AlgoLift Desk, a desktop app designed to connect to your own broker or prop account from your own computer. See going live.

#Backtesting

#Why does my backtest look so much better than my forward test?

Three common causes, in roughly the order they should be investigated:

1. Look-ahead bias. Your strategy is using information that wouldn't have been available at the moment of the trade — most commonly, reading the closed-bar value of the current bar before the bar has closed.
2. Optimistic fill settings. If you backtested with no slippage, no commission, or fills at the trigger price exactly, the live execution is going to be worse — sometimes much worse. Re-run with the Realistic slippage model and your broker's actual commission rate.
3. Overfitting. Your strategy memorized the past instead of finding an edge. Run a parameter sweep — if the strategy only works at one specific parameter value, it's not a strategy, it's a coincidence. The overfitting in trading page covers detection in depth.

#How many years of data should I backtest on?

For most futures and equities, 5–10 years is the right range. You want enough data to cover multiple regimes (the 2018 Q4 selloff, 2020 COVID, 2022 rate hikes, the 2023 yen-carry unwind) so the strategy has been stress-tested on real market state changes — not just the latest bull run.

For very new instruments (newer crypto pairs, recently-listed contracts), you may have less data available. Less than 2 years of data is rarely enough to call a strategy validated.

#What does "tick-accurate" actually mean?

It means the backtest engine processes every individual price print in the historical tape, not just the open/high/low/close of each bar. This matters because inside a 1-minute bar, the order in which the high and low formed determines whether your stop or your target was hit first. Bar-only engines guess; AlgoLift's backtest engine knows. See tick vs. bar data for the full mechanics.

#How fast does a backtest run?

A 5-year MES 5m backtest runs in 30–90 seconds. A 10-year multi-instrument test runs in 2–4 minutes. The speed comes from parallel cloud execution — the engine splits the date range into monthly chunks that run on separate workers simultaneously.

#Why is my strategy losing money in backtest even though my logic seems right?

Usually one of three things:

• The signal is real but the exits are wrong. Most "logic seems right" strategies have entries that work and exits that don't — the strategy enters profitably but exits at the wrong time. Look at the MAE/MFE distribution in Analytics.
• The signal is real but the position sizing is wrong. A strategy with 0.05R per-trade risk can have a real edge and still lose to commissions and slippage. The risk management page covers sizing math.
• The signal isn't actually an edge. EMA crossovers without filters lose money on most major instruments because the markets are mostly choppy. Strategies typically need at least one filter (regime, volatility, session) on top of a base signal to be profitable.

#Strategy design

#Should I start with mean reversion or momentum?

Most beginners are better served starting with momentum strategies. Momentum signals are visually intuitive (you can see them on a chart), tend to have larger per-trade edges, and degrade more slowly than mean reversion. Mean reversion has its place — see mean reversion strategies — but it's harder to backtest honestly because it often requires more nuanced execution.

#How many indicators should my strategy use?

Fewer than you think. Every parameter is a knob the optimizer can use to chase noise. A 2-indicator, 4-parameter strategy is much harder to overfit than a 5-indicator, 12-parameter one. The strategies that survive live deployment almost always have 2–4 main components, not 8.

The overfitting in trading page covers the math behind this.

#Can I import strategies from TradingView, NinjaTrader, or Pine Script?

No automated importer today. You'd rebuild the strategy as a node graph. In practice this is usually faster than expected — even complex TradingView scripts often translate to 10–20 nodes on the canvas.

#How do I combine multiple strategies into a portfolio?

AlgoLift doesn't combine strategies into one portfolio: each project tests one strategy. To judge whether two strategies would smooth each other's drawdowns, which is the only structural reason to run a portfolio at all, compare their runs: the comparison overlays their equity and drawdown curves.

#Optimization

#When should I optimize parameters?

After you've confirmed the strategy has a plausible shape at default parameters, not before. If a strategy doesn't make money with reasonable defaults, optimization rarely fixes it — it just finds the parameter set that overfits the past most. See parameter optimization.

#What's the difference between optimization and walk-forward?

A one-shot optimization finds the parameter set that performed best on a specific historical window. The same parameter set may or may not perform well on future data — usually not.

Walk-forward optimization addresses this by repeatedly fitting on a training window and evaluating on a held-out window, rolling forward in time. You see whether "best parameters from each window" continue to work on the next window. If yes, the strategy probably has a real edge. If no, the optimization was fitting noise.

#My strategy works at exactly RSI period 14 but fails at 13 or 15. Is that bad?

Yes. A strategy that lives on a single parameter value is fit to a coincidence the optimizer happened to find. Real edges produce a plateau of parameter values that all work reasonably well — small changes shouldn't crater performance. This is the most important diagnostic from a parameter sweep.

#Forward testing & deployment

#How long should I forward-test before going live?

Until it has enough trades to judge: the number of trades matters more than the calendar. In AlgoLift a forward stream's verdict reads from 30 trades, and the Forward Test stage says roughly how many weeks that takes at your strategy's in-sample pace. A strategy that trades a few times a week needs a couple of months; one that trades once a week needs more than half a year. See the holdout, then new days.

#Can AlgoLift trade live for me?

Not yet. Live execution is on AlgoLift's roadmap and not available yet. Today the furthest a strategy goes on AlgoLift is a forward test on each new day of real market data. See going live for what is planned.

#Will live trading run on AlgoLift's servers?

No. AlgoLift Desk is being designed to run on your own computer and send orders straight to your broker or prop platform. Your broker credentials would never reach AlgoLift, and the only record of live trades AlgoLift would see is the one you choose to sync.

#What happens if AlgoLift's web app is down?

Nothing live is affected, because AlgoLift does not place trades today. When AlgoLift Desk ships, it is designed to keep running, and to keep open positions protected, even if AlgoLift's servers are unreachable.

#Can I run the same strategy on multiple accounts?

Not through AlgoLift today, since there is no live execution yet. AlgoLift Desk is being designed to drive several accounts from one strategy, each checked against its own firm's rules. Whatever you use, check your firm's rules on copy trading first: several firms restrict it, and some close accounts that trade identically to other traders.

#Pricing & account

#Is there a free tier?

The welcome page covers what's free. In short: research is free. You can build strategies, run backtests, optimize, and forward-test without paying. Live execution isn't available yet.

#How much capital do I need to start trading live?

Depends entirely on the strategy. The platform doesn't impose a minimum. The risk management page covers the math of what's mathematically survivable at different account sizes — short version, $5,000 is the rough floor for trading MES with reasonable sizing, less than that and ordinary losing streaks will eat your account.

#Do you offer prop firm support?

Yes. The prop firm traders use case covers how the platform simulates prop firm rules (daily loss, trailing drawdown, consistency and more) against your strategy before you pay for an evaluation.

#Common platform questions

#Where do I find my old strategies?

In the Strategy Manager: the save icon at the top right of the canvas's toolbar in the Visual Builder opens it, with every strategy you've saved, your folders and your tags. A project also keeps the exact strategy behind every run it made.

#Can I share strategies with someone else?

Strategies are private by default. To share one, create a Proof from a run's Share button and choose Open recipe: anyone with the link sees the full strategy and can copy it into their own account. Sealed and Outline share the results and keep the rules private.

#Can I edit a strategy that's currently in forward testing?

Editing the strategy in the Visual Builder won't affect the running forward test — the forward test runs a snapshot of the strategy at the moment it was deployed. To deploy your edits, stop the existing forward test and deploy a new one.

#How do I get help?

Three options, in roughly the order of speed:

1. Search this guide. The search bar (Shift+S anywhere on the docs) usually surfaces the answer.
2. The glossary and support pages.
3. The support page — for direct help from the team.