$ git diff --no-index --stat=100 \
>   examples/appD/fixtures/proof-notes/model-a/budget-vs-actual/bva-note.md \
>   examples/appD/fixtures/proof-notes/model-b/budget-vs-actual/bva-note.md
 .../fixtures/proof-notes/{model-a => model-b}/budget-vs-actual/bva-note.md    | 15 +++++++++------
 1 file changed, 9 insertions(+), 6 deletions(-)
$ git diff --no-index --word-diff=plain \
>   examples/appD/fixtures/proof-notes/model-a/budget-vs-actual/bva-note.md \
>   examples/appD/fixtures/proof-notes/model-b/budget-vs-actual/bva-note.md
diff --git a/examples/appD/fixtures/proof-notes/model-a/budget-vs-actual/bva-note.md b/examples/appD/fixtures/proof-notes/model-b/budget-vs-actual/bva-note.md
index de7d3bf..ffcdb10 100644
--- a/examples/appD/fixtures/proof-notes/model-a/budget-vs-actual/bva-note.md
+++ b/examples/appD/fixtures/proof-notes/model-b/budget-vs-actual/bva-note.md
@@ -1,7 +1,10 @@
{+# FY2025 budget vs actual+}

## Headline
[-FY2025-]{+YTD 2025+} revenue [-finished at-]{+was+} $11.2M against a budget of $11.6M, [-unfavourable by-]{+a variance of+} ($362K), or (3.1%), and operating income was $388K against $583K, [-unfavourable by-]{+a variance of+} ($195K), or (33.5%).

## Results

| Measure | Actual | Budget | Variance |
| --- | --- | --- | --- |
| Revenue | $11.2M | $11.6M | ($362K) |
@@ -12,13 +15,13 @@ FY2025 revenue finished at $11.2M against a budget of $11.6M, unfavourable by ($
| Operating income | $388K | $583K | ($195K) |

## Month
December 2025 revenue was $827K against a budget of $824K, favourable by $3K, or 0.38%. Operating income was $8K against $13K, [-unfavourable by-]{+a variance of+} ($5K), or (38.0%). {+Gross margin fell to 26.1% against 26.8% budget, a variance of (0.7) pts, while operating expenses were favourable by $438.+}

## Drivers
- [-Revenue missed-]{+**Revenue miss sat+} in three [-months:-]{+months.**+} March [-at-]{+missed by+} ($123K), July [-at-]{+by+} ($132K) and August [-at ($105K). Every-]{+by ($105K); the+} other [-month was within $7K-]{+nine months ranged from ($6K) to $7K. Sales should explain each of those three months by customer and product.+}
{+- **Cost+} of [-budget.-]{+goods sold was favourable by $196K, or 2.4%, but freight out was unfavourable by ($29K), or (8.3%).**+} The [-data does not show why; the sales owners-]{+COGS saving peaked at $215K in August. Operations+} should confirm whether [-these were lost-]{+the saving tracks lower+} volume or [-deferred orders.-]{+lower unit cost.+}
- [-Operating-]{+**Operating+} expenses landed on budget at {+a variance of+} ($649), [-but only because savings offset overruns.-]{+with offsetting lines.**+} Marketing was favourable by $38K, or 28.5%, and [-Technology-]{+technology+} by $25K, or 15.0%. Salaries and wages were unfavourable by ($38K), or (2.5%), and [-Rent-]{+rent+} and occupancy by ($23K), or (5.6%).
- [-Freight out was unfavourable by ($29K), or (8.3%), while COGS was favourable by $196K, or 2.4%. Gross margin still fell short by (0.9) pts. Freight owners should explain the overrun against a lower revenue base.-]
[-- Professional-]{+**Professional+} fees [-reversed.-]{+reversed.**+} The variance peaked at ($31K) in [-2025-03-]{+March+} and [-ended-]{+closed+} the year at ($1K), so the [-March-]{+early+} overrun was [-not repeated.-]{+timing.+}

## Data notes
pnl_fy2025.csv, 120 rows, sha256 `eaa52ac89e97`
$ python3 examples/appD/fixtures/figure_sets.py \
>   examples/appD/fixtures/proof-notes/model-a/budget-vs-actual/bva-note.md \
>   examples/appD/fixtures/proof-notes/model-b/budget-vs-actual/bva-note.md
A = examples/appD/fixtures/proof-notes/model-a/budget-vs-actual/bva-note.md: 55 figures, 46 distinct; every figure is a string of bva.json's display block
B = examples/appD/fixtures/proof-notes/model-b/budget-vs-actual/bva-note.md: 60 figures, 51 distinct; every figure is a string of bva.json's display block
whole note: differ; only in A: -; only in B: $215K $438 ($6K) (0.7) 26.1%
## headline: same 8 figures
## results: same 17 figures
## month: differ; only in A: -; only in B: $438 (0.7) 26.1% 26.8%
## drivers: differ; only in A: (0.9); only in B: $215K ($6K)
## data notes: same 1 figure
