The weekly risk budget
Lesson 24 · about 9 min
Per-trade risk, heat, and correlation limits all constrain the book at a moment in time. The weekly risk budget constrains it over time. It answers a question the others do not: how much may the account lose this week before the process itself is suspended? Having that number written down is what stops one bad week from becoming a bad month.
The three budgets
| Budget | Default value | What happens when it is hit |
|---|---|---|
| Weekly loss | 3% of the account | No new entries for the rest of the week; manage existing positions only |
| Monthly loss | 6% of the account | No new entries for the rest of the month; base risk drops one step next month |
| Drawdown | 10% from equity peak | Base risk halves until equity makes a new high |
The weekly figure is measured on closed trades plus any change in open risk, from the Friday close to the following Friday close. If Wednesday's close shows the account down 3% on the week, the rest of the week is management only.
The numbers assume 1% base risk. At 0.5% base, halve them; the budgets should be roughly three, six and ten times the per-trade risk.
Why a weekly limit and not just a daily one
Day traders use daily loss limits because their damage happens within a day. A swing trader's damage happens across several days as a cluster of positions fails together, usually because the regime turned and the book was built for the old regime. A daily limit would rarely trigger; a weekly one catches the cluster.
Week's equity (R)
0 |‾‾\
-1 | \__ Mon: one stop
-2 | \__ Tue: two more, sector gap
-3 | \_ Wed: weekly limit hit -> no new entries
-4 | ‾‾‾‾‾‾‾‾‾‾ Thu, Fri: manage only, regime rechecked Sat
A trader without the limit, on Wednesday of that week, typically adds new positions to "make it back". The setups available on a Wednesday after a regime turn are the worst of the year. The limit exists for that Wednesday.
Allocating the budget across the week
A 3% weekly limit with 1% per trade is three full losses. In practice the budget is spent something like this:
| Regime | Fresh entries planned | Max weekly loss committed | Room for adds |
|---|---|---|---|
| Green | 3 to 4 at 1% | 3% to 4% (heat cap binds at 5%) | Yes, if heat allows |
| Amber | 2 to 3 at 0.5% | 1% to 1.5% | Rarely |
| Red | 1 to 2 at 0.25% | Under 0.5% | No |
Notice that in amber and red the weekly limit is almost impossible to hit through new entries; it is the existing positions from a previous green week that can breach it. That is the intended behaviour: the budget protects the book during the transition from one regime to the next.
Key idea: A weekly loss limit of about three times the per-trade risk, and a monthly limit of about six, catch the clustered failures that swing trading produces when a regime turns. When hit, stop entering, keep managing.
What "manage only" means
When the weekly or monthly limit is hit:
- Existing stops stay in place. Do not tighten them in a panic; the plan was the plan.
- Trails continue to operate on winners.
- Time stops continue to apply.
- Partials continue at 2R.
- No new entries, no adds, no re-entries after a stop-out, regardless of how good the setup looks.
It is management, not liquidation. The positions that survive the bad week are often the ones that lead the next good one.
The drawdown rule
The 10% drawdown rule is the circuit breaker for a longer decline. At 10% below the equity peak, base risk halves (1% becomes 0.5%). It stays halved until the account makes a new equity high, however long that takes. This does two things: it slows the descent so that a further 10% takes twice as many losses, and it forces the trader to earn back size rather than recover it in one aggressive week.
If the drawdown reaches 15%, stop trading for two weeks, rerun the regime and setup statistics from the journal (Module 7), and restart at half size with no more than three positions. This is rarely needed if the weekly and monthly limits are respected; it is the backstop for the trader who did not respect them.
Writing it down
The budgets go in the one-page risk plan from the risk management course, alongside base risk, the position cap, heat limits and the correlated-group cap. The complete set for a $25,000 account at 1% base:
| Rule | Value |
|---|---|
| Base risk | $250 (1%) |
| Position cap | $5,000 (20%); $2,500 through events |
| Max heat | 5% / 3% / 1.5% by regime |
| Correlated group | 2% / 1% by regime |
| Weekly loss | $750 (3%) |
| Monthly loss | $1,500 (6%) |
| Drawdown halving | 10% below peak |
| Full stop | 15% below peak |
Eight lines. The rest of the process fits inside them.
Try it: Write the eight-line table for your own account. Then pull up your last bad week from the journal. Mark the point at which the weekly limit would have triggered and list every trade you entered after that point. Add up their result.
Recap
- Weekly loss limit about 3x per-trade risk, monthly about 6x, measured Friday close to Friday close.
- When hit: no new entries, no adds; existing stops, trails, partials and time stops continue.
- The weekly limit catches the clustered failures that come with a regime change.
- At 10% drawdown from peak, halve base risk until a new equity high; at 15%, stop for two weeks and restart at half size.
- The full sizing rule set is eight lines; write it and keep it visible.
See it drawn
Original diagrams for the ideas on this page. Illustrative, not real market data.