Durable reference map
A three-stage method to reuse whenever conditions change
Keep the sequence of inputs, calculation and exception testing stable instead of relying on a market forecast.
- Align inputs and unitsGo to equations and definitionsEquity at the calculation time・Budget from current capital・Theoretical size after the capital refresh
- Reconcile the worked exampleGo to table and calculation stepsJPY 100,000 cash with a JPY 8,000 unrealized loss
- Test exceptions and next checksGo to rules and counterexampleAttach the field name, timestamp and time zone to every capital value.
A capital denominator has a definition and a time
Cash balance does not necessarily include the mark-to-market result of open positions. Depending on the provider’s specification, equity or NAV incorporates current unrealized P&L and other defined adjustments.
One percent of JPY 100,000 is not one percent of JPY 92,000. A recorded percentage is incomplete without the capital value and timestamp to which it was applied.
- Capture cash balance, unrealized P&L and equity in separate fields.
- Confirm when fees, deposits and withdrawals enter each field.
- Attach a time zone to the pre-order snapshot.
How a stale balance changes the effective risk rate
Continuing to use JPY 1,000, which is one percent of the earlier JPY 100,000 balance, makes the planned loss about 1.087% of current JPY 92,000 equity. Under a current-equity policy, the budget becomes JPY 920.
This is not a question about trade direction or the number of trades taken today. It is a timestamp mismatch in the denominator.
Recompute two snapshots under one product specification
The table contains hypothetical educational inputs. Stop distance is fixed at 50 points, point value at JPY 100 per point per lot and volume step at 0.01 lot; only the capital snapshot changes.
The current theoretical result of 0.184 lot is rounded down to 0.18, producing JPY 900 planned loss.
Do not confuse the denominator with a daily remaining budget
This article chooses the capital base to which a percentage policy is applied. Article 20 separately subtracts realized loss, open stop risk and reserved risk from a daily cap over time.
Correct current equity does not override a tighter remaining daily limit. Apply the capital policy and the daily budget as separate constraints.
Control an intentionally fixed capital base
A strategy may deliberately use month-start capital. That is not automatically a stale-data error if the reference value, refresh date, treatment of cash flows, maximum divergence and review trigger were defined before the result was known. Do not switch between fixed and current values opportunistically.
- Display fixed base and current equity side by side.
- Link the selected base to the order record.
- Require review when the divergence limit is reached.
Calculation framework
Equity at the calculation time
Read the role of each equation first, then follow the numerical example to check the decision path.
Equity at the calculation time
E_t = C_t + U_t- E_t: equity or relevant NAV at time t
- C_t: cash balance at the same time
- U_t: net unrealized P&L at the same time
In plain language: Build the capital measure from fields defined by the provider at one timestamp.
When this conclusion does not apply: Treatment of credits, unsettled fees and pending transfers varies; do not infer it from a generic label. Return no quantity when E_t is missing or non-positive.
Budget from current capital
B_t = r × E_t- B_t: account-currency loss budget at time t
- r: predetermined budget rate
- E_t: selected capital denominator
In plain language: The dollar or yen budget changes when the denominator changes, even if the percentage is constant.
When this conclusion does not apply: The example rate is an input, not a recommended level. Do not derive a budget or quantity unless E_t > 0 and r ≥ 0.
Theoretical size after the capital refresh
Q_t* = B_t / (D × V)- Q_t*: lots before rounding
- D: stop distance in points
- V: JPY value per point per lot
In plain language: Apply the refreshed budget to the unchanged product loss per lot.
When this conclusion does not apply: Return no quantity unless B_t ≥ 0, D > 0 and V > 0. Daily risk limits and margin capacity remain separate constraints.
Name the capital measure before applying a percentage
A percentage risk rule is incomplete without a denominator. Cash balance, equity, NAV, available funds, and margin figures can differ because they include different combinations of unrealized P&L, settled cash, credits, fees, or provider-specific adjustments. A field showing JPY 100,000 cannot be used safely merely because it looks like account value.
The calculation policy should identify the exact provider field or a documented reconstruction. If current equity is defined as cash balance plus unrealized P&L for this example, JPY 100,000 cash and a negative JPY 8,000 open result produce JPY 92,000. Another provider may define adjacent fields differently, so generic labels should not be treated as universal specifications.
A percentage also needs an effective time. One percent of the 09:00 cash-only snapshot is JPY 1,000, while one percent of the 10:30 current-equity snapshot is JPY 920. Both multiplications are arithmetically correct; only the policy and aligned timestamp determine which denominator belongs to the pending decision.
Understand how stale capital alters the effective rate
Reusing the earlier JPY 1,000 amount after equity falls to JPY 92,000 does not preserve a one-percent policy. It makes planned loss approximately 1.087 percent of the current denominator. The quantity error may appear small, but its direction is systematic: losses can make an old fixed money amount consume a larger current-equity share.
The reverse can happen after gains or deposits. An old budget can become a smaller share of current equity, but that does not make stale data prudent by definition. A process that refreshes after losses and ignores gains, or does the opposite, embeds outcome-dependent discretion. The update rule must be fixed before knowing which denominator produces the preferred lot.
Dividing the selected money budget back by the current denominator is an effective-rate check. It reveals policy drift even when the front-end still displays the configured percentage. This backward ratio should be stored alongside the original capital field and time so a reviewer can distinguish deliberate fixed-base behavior from an accidental stale snapshot.
Align balance, open P&L, and cash movements on one clock
A reconstructed equity value is valid only when its components refer to the same state. Adding a 10:30 unrealized result to a 09:00 cash balance can double-count or omit a trade closure, fee, deposit, or withdrawal that occurred between them. Each component should carry its own source timestamp, followed by a snapshot time and timezone for the assembled denominator.
Cash movements need explicit treatment. A withdrawal requested but not settled may appear in one interface and not another; a deposit can be credited before it is withdrawable; fees can post after an execution. The policy should follow provider definitions rather than infer economic availability from a single screen label. Ambiguity is a reason to withhold a new quantity.
The time zone is not administrative decoration. A daily statement boundary, platform server time, and local clock can place the same transaction on different dates. For reproducibility, the record should retain the original timestamp and offset, the conversion applied for display, and the exact moment at which the quantity calculation read the capital measure.
Replay the hypothetical capital refresh through the lot calculation
The example holds stop distance at 50 points, monetary value at JPY 100 per point per lot, and volume step at 0.01. Loss per lot is therefore JPY 5,000 in both rows. By fixing the product side of the calculation, the comparison isolates the effect of changing only the capital snapshot used for the one-percent budget.
At 09:00, JPY 100,000 multiplied by one percent gives JPY 1,000, which supports 0.20 lot and replays to JPY 1,000 planned price-distance loss. At 10:30, equity of JPY 92,000 gives a JPY 920 budget. Dividing by JPY 5,000 produces 0.184 lot, rounded down to 0.18.
The final backward check is 0.18 times JPY 5,000, or JPY 900. The JPY 20 difference is a rounding residue, not unused permission to round up. These are hypothetical educational values and exclude separate fees, execution overrun, daily-cap consumption, and margin constraints that could make the executable ceiling smaller.
Distinguish a deliberate fixed base from an accidental stale value
Some policies intentionally use month-start or session-start capital to reduce frequent quantity changes. That design is not automatically stale-data misuse. It becomes defensible only when the reference time, refresh schedule, cash-flow treatment, maximum allowed divergence, and exception process were specified in advance and applied to gains and losses consistently.
A fixed base should be labeled as such rather than displayed as current equity. The record can show both the policy denominator and current provider value, then calculate their divergence and the effective current-equity risk share. This preserves operational stability without hiding how far the fixed-base decision has moved from the account’s present state.
The boundary must define what happens when current equity becomes nonpositive or falls beyond the permitted divergence. Continuing to divide by an old positive base can produce a quantity even when the account no longer has valid current capacity. A stop condition in the policy is therefore more important than the convenience of a stable lot.
Keep current-equity budgeting separate from daily remaining capacity
Refreshing equity answers how much a configured percentage represents now. It does not answer how much of a daily loss cap remains after realized losses, open-stop exposure, and pending-order reservations. The two controls can produce different ceilings, and the smaller one should bind without changing the definition of either denominator.
For example, a current-equity policy might yield JPY 920 for a new trade while a separate daily ledger leaves only JPY 600. Using JPY 920 would satisfy the percentage arithmetic yet breach the tighter temporal budget. Conversely, unused daily capacity does not authorize more than the current-equity rule permits. The controls should be displayed side by side.
Margin capacity is a third question. An enabled order or sufficient free margin does not establish that price-distance loss fits either budget. Combining equity, daily capacity, and margin into one opaque availability number removes the ability to see which constraint changed and can encourage incorrect claims that leverage expanded the loss allowance.
Design refresh events around economic state changes
A current-equity workflow should refresh immediately before a new order and after events that materially change the denominator: realized closures, large mark-to-market moves, deposits, withdrawals, fees, or corrections. A periodic refresh alone can miss a state transition between intervals, while refreshing every screen tick without versioning can make later reconstruction difficult.
The implementation should read an atomic snapshot where possible. If fields arrive separately, it should detect inconsistent timestamps and retry rather than combine them silently. The saved calculation needs the provider field names, values, timestamps, account currency, and any conversion performed. A successful API response is not enough if the economic components are temporally misaligned.
Failure behavior should be explicit. A missing capital field, nonpositive current equity, unknown timestamp, or unresolved cash movement should return no quantity and a diagnostic reason. Falling back to the last good value can be appropriate for display continuity, but not for creating a fresh risk decision unless a predeclared fixed-base policy expressly permits it.
Test the policy against selective updating and label confusion
An adversarial test can feed identical economic states under different field labels to see whether the system assumes balance equals equity. Another can introduce an open loss after the snapshot and verify that the old budget is rejected. A third can post a withdrawal between component timestamps and test whether the reconstruction detects the mismatch.
Selective refresh is a behavioral failure worth testing directly. Run one path with a gain and another with an equal loss, then confirm that the same update rule operates in both. If the system adopts current equity only when it raises the lot, the configured percentage is merely a narrative; actual sizing is being driven by the preferred outcome.
Rounding tests should retain the difference between 0.184 theoretical lot and 0.18 executable lot. A nearest-value rule that produces 0.18 happens to remain below budget here, but another input may round upward. The control should always round according to the declared protective rule and multiply the executable result back into account-currency loss.
Preserve enough data to explain a changed or unchanged lot
The audit record should include capital policy name, provider field or reconstruction, cash balance, unrealized P&L, relevant cash movements, equity, percentage, monetary budget, timestamp, timezone, stop distance, point value, raw lot, step, rounded lot, and checked loss. Each input needs a source and version where definitions can change.
An unchanged quantity can still reflect a meaningful refresh. Step rounding may map two nearby budgets to the same executable lot even though theoretical size changed. Saving only the final lot would make the refresh invisible. Retaining both raw results reveals the execution-grid plateau and prevents an analyst from falsely concluding that current equity was ignored.
A no-order result also needs durable evidence. If equity is unavailable or below a policy boundary, storing zero without the reason can look like a software fault. Capturing the failed prerequisite shows that the calculation stopped before producing a potentially misleading quantity and allows the data-quality issue to be reviewed separately.
Interpret provider definitions as scoped evidence
ESMA Decision 2018/796 is no longer in force and appears here only as a historical terminology example from its specific CFD margin-close-out context. OANDA documentation defines statement NAV or equity fields with timestamps, and Interactive Brokers distinguishes unrealized from realized P&L. These sources demonstrate why field semantics matter; they do not create current law, current product terms, or one universal account formula.
A provider’s current schema should govern its own calculation. The educational equation cash plus unrealized P&L is useful for exposing the example, but credits, unsettled fees, financing, and other adjustments can require different treatment. The page should therefore state the assumed construction and direct production logic to verified provider definitions rather than generalizing the equation.
The evidence-bounded conclusion is that capital denominators are stateful and time-specific. Refreshing a defined measure can change the monetary budget and rounded quantity even when the percentage, stop, and point value are unchanged. It cannot establish that one percent is suitable, that current equity is always the intended policy base, or that the planned loss is guaranteed.
Propagate cost and currency assumptions without corrupting the denominator
The JPY 900 replay covers the stated stop distance and point value. If the product’s P&L is first expressed in another currency, the loss-side conversion factor and its timestamp belong in loss per lot. Fees and adverse execution can also consume budget. Those numerator additions should not be hidden by changing the capital denominator after the lot is calculated.
A conversion update and an equity update are different time-sensitive events. The first changes how product loss becomes JPY; the second changes how much JPY a percentage policy permits. Reading a current equity value with an old conversion factor, or the reverse, can still produce a seemingly coherent ratio. The journal needs both timestamps and an allowed-age policy for each.
Known fixed reserves can be subtracted from the monetary budget, while quantity-proportional costs belong in per-lot loss or an iterative replay after rounding. This accounting choice must be documented so a fee is neither omitted nor counted twice. The remaining account-currency allowance is then compared with the current-equity rule and any tighter daily capacity.
Review outcomes without rewriting the sizing-time snapshot
After the trade closes, realized P&L, fills, fees, and final conversion can be compared with the estimate. The original JPY 92,000 denominator and 10:30 timestamp must remain intact even if a later statement revises account values. Reconciliation can append corrections, but replacing the decision-time record would destroy evidence of what the system actually knew.
A loss larger than JPY 900 can arise from execution or cost components rather than from stale equity, while a smaller loss does not prove the denominator was current. Post-trade diagnosis should attribute the variance to price distance, quantity, value, conversion, fees, and fill behavior before judging the capital refresh. Outcome magnitude alone cannot identify the faulty input.
Aggregated review can compare calculations made under current-equity and deliberately fixed-base policies, but it must preserve selection and timing. Choosing the policy that would have produced the best historical lot creates hindsight bias. Any change should be specified prospectively, versioned, and evaluated on later decisions rather than backfilled into earlier records.
State the final decision in terms of the active policy and its limits
Under the hypothetical current-equity policy, JPY 92,000 is the active denominator at 10:30, one percent is JPY 920, and the 0.01-step quantity is 0.18 lot. The checked price-distance loss is JPY 900. That statement is reproducible because it names the policy, state, units, rounding, and amount left outside the order by the execution grid.
Under a prospectively documented month-start-base policy, JPY 100,000 might remain the denominator despite the open loss, subject to its divergence and exception rules. That would be a different policy result, not proof that the 10:30 value is irrelevant. Showing both scenarios avoids labeling every fixed base stale while still exposing its effective current-equity rate.
Neither policy makes the stop loss certain or determines whether an order should be placed. The calculation can enforce consistency between a selected capital measure and quantity, and it can stop when data are missing or constraints conflict. Market outcome, execution quality, and the suitability of the percentage remain outside that arithmetic conclusion.
Decision and control rules
- Attach the field name, timestamp and time zone to every capital value.
- Under a current-equity policy, refresh after open P&L and cash movements are reflected.
- Define refresh and divergence rules for any fixed capital base.
- Divide the selected budget back by current equity to reveal its effective rate.
- Check daily remaining risk and margin capacity separately.
Common failure modes
- Assuming balance and equity are the same without checking provider definitions.
- Reusing a snapshot taken before a withdrawal.
- Selecting fixed capital after losses but current equity after profits.
Evidence and specifications
- ESMA Decision (EU) 2018/796 on CFDs
What this source supports: The no-longer-in-force 2018 decision defined account value for its CFD margin-close-out context as funds in the account together with unrealized net profits from connected open CFDs. This article uses it only as a historical terminology example for distinguishing cash balance from current account value, not as current law or current product terms.
- OANDA — Account statement definitions
What this source supports: The provider’s official statement documentation defines opening and closing NAV/equity as including realized and unrealized P&L and records exact period timestamps and time zones.
- Interactive Brokers — Unrealized P&L glossary
What this source supports: The platform glossary defines unrealized P&L as gain or loss on open positions and distinguishes it from realized P&L reflected after closure.
Questions to resolve
Should I use balance or equity?
That depends on the documented policy. This example uses current equity to measure the risk rate at the order time; confirm the provider’s field definition.
Should unrealized profit increase the denominator?
A current-equity policy normally follows the platform definition, but a separate rule may decline to increase the budget from open gains. State that rule explicitly.
How often must equity be refreshed?
There is no universal interval. Refresh when fills, fees or cash movements change the relevant value between decision and order.
Is the stale JPY 1,000 loss still acceptable because it is the same cash amount?
Check the current percentage as well. In this example it is about 1.087% of JPY 92,000, not 1%.
Recalculate from current inputs
Read the balance, unrealized P&L and equity at the order timestamp, then recalculate size from the selected capital base.
Important: This is educational material about capital snapshots and sizing, not a recommendation of a budget rate or quantity. Equity, fees and cash-flow timing are provider-specific.