A 90-Day Plan to Build Consistency
Consistency doesn't arrive by accident. It's built through a deliberate plan. Here is a simple 90-day structure you can adapt.
Days 1 to 30: Foundation.
- Write your trading plan and checklist
- Choose one pair, one setup, one session
- Start your journal
- Backtest 50 setups
- Practice on demo with tiny stakes of attention: one trade a day maximum
Days 31 to 60: Execution.
- Continue demo or very small live risk
- Follow the checklist on every trade
- Complete weekly reviews
- Track rule violations and reduce them week by week
- Continue backtesting up to 100 setups
Days 61 to 90: Refinement.
- Analyze your journal data: best sessions, best setups, worst habits
- Remove low-quality behaviors, such as trading after 11 a.m. or on Fridays
- Calculate expectancy and drawdown
- Decide whether your data supports going live or increasing size slightly
- Write your next 90-day plan
Daily anchors:
- Morning routine
- Pre-trade checklist
- Post-session journal
- Ten minutes of reflection at night
Success metrics for the 90 days aren't profit. They are: rules followed, journal completeness, backtest size, and reduced impulsive trades.
At day 90, you'll have something most traders never build: evidence about yourself.
Consistency is a habit, and habits are built one day at a time.
A 90-Day Plan to Build Consistency
A 90-Day Plan to Build Consistency
It’s Fairman 
-
PropScalpDesk
- Posts: 521
- Joined: Sat Sep 19, 2026 7:50 pm
Re: A 90-Day Plan to Build Consistency
The weak spot in most 90-day plans is the decision at the end, and this one has it too. "Decide whether your data supports going live" is easy to answer yes to when you're keen. I'd write the pass criteria on day 1, before any trades exist. For example: at least 90% of trades taken per checklist in days 31 to 90, positive expectancy after spread and commission over the combined backtest and forward sample, and a worst drawdown you can actually sit through.
Also be realistic about sample size. One demo trade a day for the first month gives you around 20 trades, since there are only about 21 trading days in 30. That's enough to practise execution and nowhere near enough to calculate expectancy. The 100 backtested setups carry most of the statistical weight here, and the forward trades mainly tell you whether you can execute what you tested.
The success metrics you listed are the right ones. Profit in the first 90 days is mostly luck anyway.
Also be realistic about sample size. One demo trade a day for the first month gives you around 20 trades, since there are only about 21 trading days in 30. That's enough to practise execution and nowhere near enough to calculate expectancy. The 100 backtested setups carry most of the statistical weight here, and the forward trades mainly tell you whether you can execute what you tested.
The success metrics you listed are the right ones. Profit in the first 90 days is mostly luck anyway.
-
LondonNewsTrader
- Posts: 421
- Joined: Mon Sep 21, 2026 9:30 am
Re: A 90-Day Plan to Build Consistency
Ninety days is a useful length because it covers a full round of the main scheduled events: three US payrolls reports, three CPI prints, and at least one meeting for each of the big central banks. I'd mark those dates on the plan at the start, so the journal analysis in the last month can separate how the setup behaved on those days from the rest.
One caution for the demo phase: demo accounts often fill cleanly through releases when a live account wouldn't, with fixed or narrower spreads. If the plan includes trading near news, the move to small live risk in days 31 to 60 matters more than it looks, because that's the first time the costs are real.
One caution for the demo phase: demo accounts often fill cleanly through releases when a live account wouldn't, with fixed or narrower spreads. If the plan includes trading near news, the move to small live risk in days 31 to 60 matters more than it looks, because that's the first time the costs are real.
-
Shadow Trader
- Posts: 5594
- Joined: Wed Sep 09, 2026 5:19 pm
Re: A 90-Day Plan to Build Consistency
I would push back a little on the one trade a day cap in the first month. It protects attention, but as PropScalpDesk worked out it leaves you with about 20 trades, and a scalping setup that fires four times a morning will get skipped for reasons unrelated to its quality.
What worked for me was taking every valid setup on demo during a fixed 90 minute window, and grading each one A or B before entry. That gave me around 150 forward trades by day 60 without stretching my hours.
On the pass criteria, I also wrote down what happens on a fail. If day 90 shows positive expectancy but checklist compliance under 90%, the next 90 days repeat at the same size rather than going live, and I do not get to redefine compliance.
LondonNewsTrader's calendar idea is worth one extra column. I flag every trade inside 30 minutes of a red release, and in my first plan those flagged trades held most of the drawdown while being under 15% of the count.
How are others counting a rule violation? I score it per trade, so one trade with a late entry and a moved stop counts as two.
What worked for me was taking every valid setup on demo during a fixed 90 minute window, and grading each one A or B before entry. That gave me around 150 forward trades by day 60 without stretching my hours.
On the pass criteria, I also wrote down what happens on a fail. If day 90 shows positive expectancy but checklist compliance under 90%, the next 90 days repeat at the same size rather than going live, and I do not get to redefine compliance.
LondonNewsTrader's calendar idea is worth one extra column. I flag every trade inside 30 minutes of a red release, and in my first plan those flagged trades held most of the drawdown while being under 15% of the count.
How are others counting a rule violation? I score it per trade, so one trade with a late entry and a moved stop counts as two.
-
Shadow Trader
- Posts: 5594
- Joined: Wed Sep 09, 2026 5:19 pm
Re: A 90-Day Plan to Build Consistency
The worst drawdown you can sit through needs a number before day 1, and streak maths gives a rough one. At a 45% win rate the longest losing run expected in about 100 trades is around eight in a row, so at 1% risk that is roughly an 8% dip from a perfectly normal sample.
Writing that down early stops the day 61 review from treating an ordinary streak as proof the setup broke. A run of six is then information about variance, not a reason to rewrite the rules.
I would also be careful with the step that removes Fridays or late hours. Ninety days holds only about thirteen Fridays, and if each produced two or three trades, a bad Friday result rests on around thirty trades, which is close to noise.
A cleaner route is to name the filters as hypotheses on day 1, then test them only on the backtest sample and check the forward days against them. A filter invented from the journal in month three is in sample by definition.
LondonNewsTrader, would you exclude the payrolls and CPI days from the expectancy figure, or keep them in and report them alongside? Taking them out flatters the number, but leaving them in can hide a setup that only works on quiet days.
Writing that down early stops the day 61 review from treating an ordinary streak as proof the setup broke. A run of six is then information about variance, not a reason to rewrite the rules.
I would also be careful with the step that removes Fridays or late hours. Ninety days holds only about thirteen Fridays, and if each produced two or three trades, a bad Friday result rests on around thirty trades, which is close to noise.
A cleaner route is to name the filters as hypotheses on day 1, then test them only on the backtest sample and check the forward days against them. A filter invented from the journal in month three is in sample by definition.
LondonNewsTrader, would you exclude the payrolls and CPI days from the expectancy figure, or keep them in and report them alongside? Taking them out flatters the number, but leaving them in can hide a setup that only works on quiet days.
-
FXS Prop Desk
- Posts: 1689
- Joined: Wed Oct 07, 2026 9:52 am
Re: A 90-Day Plan to Build Consistency
The 90-day structure is solid. One suggestion for days 1 to 30: track the time spent on each activity. Many traders find that most of their time goes to chart watching and very little to backtesting and reviewing. Seeing the numbers makes it simpler to rebalance the time towards activities that build skill.
The second and third phases benefit from the same tracking.
The second and third phases benefit from the same tracking.
FXS Official Team | The Idea Lab & Strategy
-
Shadow Trader
- Posts: 5594
- Joined: Wed Sep 09, 2026 5:19 pm
Re: A 90-Day Plan to Build Consistency
Plenty of people want trading as a second income without quitting the day job. It's a more sensible goal than going full-time on a small account, and it still needs a plan.
A workable outline has stages. Stage one is education and demo, long enough to build a tested approach, with no pressure on money. Stage two is small live trading, with funds that can be lost without hardship, focused on execution quality and consistency. Stage three is gradual scaling, if results hold up, with strict risk limits.
Time management matters. A defined window, clear boundaries with work and family, and a routine for review. Chasing trades at lunchtime or during meetings is a recipe for mistakes.
Income expectations should be modest and flexible. Months of zero or negative returns happen, and the plan needs to survive them without panic.
The side-income model lets the skill develop without financial pressure, which is a real edge.
What does the side-income setup look like in practice, and what turned out harder than expected?
A workable outline has stages. Stage one is education and demo, long enough to build a tested approach, with no pressure on money. Stage two is small live trading, with funds that can be lost without hardship, focused on execution quality and consistency. Stage three is gradual scaling, if results hold up, with strict risk limits.
Time management matters. A defined window, clear boundaries with work and family, and a routine for review. Chasing trades at lunchtime or during meetings is a recipe for mistakes.
Income expectations should be modest and flexible. Months of zero or negative returns happen, and the plan needs to survive them without panic.
The side-income model lets the skill develop without financial pressure, which is a real edge.
What does the side-income setup look like in practice, and what turned out harder than expected?