Home / Articles / Practical notes: Never Average Two Balance Sheets

This article is published in English.

Practical notes: Never Average Two Balance Sheets

Operable walkthrough of Practical notes: Never Average Two Balance Sheets: contracts, checks, and drop-in code slots for teams shipping this pattern.

1705 words

Use this as an operator-facing rebuild of the ideas in “Never Average Two Balance Sheets”: clear stages, ordered code slots, and recovery notes that survive a handoff. The Overview stage works best when treated as a measurable surface. Capture one golden transcript, one failure case, and the rollback note before expanding scope. Record timings and token or query cost next to functional results. Cost visibility early prevents surprise bills when the path moves from demo to shared environments.

Drawing the lines

For the Drawing the lines stage, define the inputs, the owner of the step, and the exit criteria before changing code. Operators should be able to re-run the step from a known checkpoint without guessing hidden state. Keep configuration outside application code. Environment files, secret stores, and feature flags belong in one place operators can audit without reading the whole graph. Cite the passages that actually grounded the answer. Without citations, operators cannot tell hallucination from an indexing gap.

One request, end to end

For the One request end to stage, define the inputs, the owner of the step, and the exit criteria before changing code. Operators should be able to re-run the step from a known checkpoint without guessing hidden state. Document the happy path and the recovery path together. Retries, human gates, and dead-letter handling are part of the product, not later polish. Cite the passages that actually grounded the answer. Without citations, operators cannot tell hallucination from an indexing gap.

@app.get("/api/analyze")
def analyze(q: str, market: str = "US", refresh: bool = False):
    snap = _up("marketdata", "/snapshot",
               params={"q": q, "market": market, "refresh": refresh})
    rates = _up("marketdata", f"/rates/{market}")
    result = _up("valuation", "/analyze", method="POST", json={
        "symbol": snap["symbol"], "market": market,
        "snapshot": snap["data"], "rates": rates, "persist": True,
    })
    # Provenance travels with the numbers, so the UI can show who said what.
    result["sources"] = snap.get("sources", [])
    result["disagreements"] = snap.get("disagreements", [])
    return result

Three sources, and the rule that nothing is averaged

For the Three sources and the stage, define the inputs, the owner of the step, and the exit criteria before changing code. Operators should be able to re-run the step from a known checkpoint without guessing hidden state. Prefer small, testable units over sprawling scripts. When a step fails, the failure should point at a single responsibility rather than a tangled pipeline. Cite the passages that actually grounded the answer. Without citations, operators cannot tell hallucination from an indexing gap. For the Three sources and the stage, define the inputs, the owner of the step, and the exit criteria before changing code. Operators should be able to re-run the step from a known checkpoint without guessing hidden state. Record timings and token or query cost next to functional results. Cost visibility early prevents surprise bills when the path moves from demo to shared environments.

# SEC is as-filed, so it outranks everything in the US.
PRIORITY_US    = ("sec", "fmp", "yahoo")
PRIORITY_OTHER = ("fmp", "yahoo")
def _dispersion(values: list[float]) -> float | None:
    """Relative spread across sources. 0.0 means they agree exactly."""
    clean = [v for v in values if v is not None]
    if len(clean) < 2:
        return None
    scale = abs(statistics.median(clean))
    if scale < 1e-9:
        return None if max(map(abs, clean)) < 1e-9 else 1.0
    return (max(clean) - min(clean)) / scale

The currency gate

When working through the The currency gate stage, write down the contract first: required inputs, success signal, and what happens on partial failure. That checklist keeps later code changes honest. Keep configuration outside application code. Environment files, secret stores, and feature flags belong in one place operators can audit without reading the whole graph. Measure recall on a fixed question set before tuning prompts. Prompt churn rarely fixes a weak retrieval surface.

for name, data in sources.items():
    rc = (data.get("reporting_currency")
          or data.get("currency") or base_currency or "").upper()
    allowed_statements[name] = (not base_currency) or rc == base_currency.upper()

Stop hardcoding the risk-free rate

When working through the Stop hardcoding the risk-free stage, write down the contract first: required inputs, success signal, and what happens on partial failure. That checklist keeps later code changes honest. Document the happy path and the recovery path together. Retries, human gates, and dead-letter handling are part of the product, not later polish. Measure recall on a fixed question set before tuning prompts. Prompt churn rarely fixes a weak retrieval surface.

"US": {"rf": 0.042, "erp": 0.045, "tax": 0.21},
def risk_free(market: str, fallback: float) -> dict:
    """{rate, source, observed_on, series}. Never raises."""
    series_id = SERIES.get(market)
    static = {"rate": fallback, "source": "static",
              "observed_on": None, "series": None}
    if not series_id or not enabled():
        return static
    ...

Holdings are derived, never stored

When working through the Holdings are derived never stage, write down the contract first: required inputs, success signal, and what happens on partial failure. That checklist keeps later code changes honest. Prefer small, testable units over sprawling scripts. When a step fails, the failure should point at a single responsibility rather than a tangled pipeline. Measure recall on a fixed question set before tuning prompts. Prompt churn rarely fixes a weak retrieval surface. When working through the Holdings are derived never stage, write down the contract first: required inputs, success signal, and what happens on partial failure. That checklist keeps later code changes honest. Record timings and token or query cost next to functional results. Cost visibility early prevents surprise bills when the path moves from demo to shared environments.

An analyst that cannot overrule the analysis

The An analyst that cannot stage works best when treated as a measurable surface. Capture one golden transcript, one failure case, and the rollback note before expanding scope. Keep configuration outside application code. Environment files, secret stores, and feature flags belong in one place operators can audit without reading the whole graph. Separate chunking policy from retrieval policy. Changing one should not force a rewrite of the other when quality metrics move.

VALUATION
  blended fair value: 1402.11 INR (confidence medium, model spread 0.43)
  per model: dcf=1610.22, epv=1104.50, graham=1288.31, gordon=1377.90
  margin of safety: -8.4% - Trading above fair value
  growth assumed: 11.0% (basis: revenue YoY 11.0%)   discount rate: 11.2%
SOURCE DISAGREEMENTS (treat these figures as uncertain):
  net_income: fmp=261.0B, yahoo=248.5B (spread 4.9%, using fmp)
def _clamp_to_engine(model_verdict, engine_verdict):
    gap = SCALE.index(model_verdict) - SCALE.index(engine_verdict)
    if abs(gap) <= 1:
        return model_verdict, None
    capped = SCALE[SCALE.index(engine_verdict) + (1 if gap > 0 else -1)]
    return capped, (f"model said '{model_verdict}', more than one rung from "
                    f"the engine's '{engine_verdict}' - capped at '{capped}'")

What the manifests actually encode

The What the manifests actually stage works best when treated as a measurable surface. Capture one golden transcript, one failure case, and the rollback note before expanding scope. Document the happy path and the recovery path together. Retries, human gates, and dead-letter handling are part of the product, not later polish. Separate chunking policy from retrieval policy. Changing one should not force a rewrite of the other when quality metrics move.

startupProbe:    { httpGet: { path: /health/live,  port: http },
                   failureThreshold: 30, periodSeconds: 5 }
readinessProbe:  { httpGet: { path: /health/ready, port: http }, periodSeconds: 15 }
livenessProbe:   { httpGet: { path: /health/live,  port: http }, periodSeconds: 30 }

Four things that broke

The Four things that broke stage works best when treated as a measurable surface. Capture one golden transcript, one failure case, and the rollback note before expanding scope. Prefer small, testable units over sprawling scripts. When a step fails, the failure should point at a single responsibility rather than a tangled pipeline. Separate chunking policy from retrieval policy. Changing one should not force a rewrite of the other when quality metrics move. The Four things that broke stage works best when treated as a measurable surface. Capture one golden transcript, one failure case, and the rollback note before expanding scope. Record timings and token or query cost next to functional results. Cost visibility early prevents surprise bills when the path moves from demo to shared environments.

What you’d change

For the What you d change stage, define the inputs, the owner of the step, and the exit criteria before changing code. Operators should be able to re-run the step from a known checkpoint without guessing hidden state. Keep configuration outside application code. Environment files, secret stores, and feature flags belong in one place operators can audit without reading the whole graph. Cite the passages that actually grounded the answer. Without citations, operators cannot tell hallucination from an indexing gap.

Operational checklist

When working through the Operational checklist stage, write down the contract first: required inputs, success signal, and what happens on partial failure. That checklist keeps later code changes honest.

Treat this stage as a contract between inputs and validated outputs. Name the artifacts, define success checks, and refuse silent partial completion.

Measure recall on a fixed question set before tuning prompts. Prompt churn rarely fixes a weak retrieval surface.

Pin dependency versions and record the image digest that ran the demo. Reproducibility beats tribal knowledge.

Record timings and token or query cost next to functional results. Cost visibility early prevents surprise bills when the path moves from demo to shared environments.

Measure recall on a fixed question set before tuning prompts. Prompt churn rarely fixes a weak retrieval surface.

Before promoting the stack, freeze versions, capture a golden transcript for the critical path, and confirm rollback steps. Shared environments need rate limits, tenancy checks, and a clear owner for secret rotation. Prefer boring reliability over clever one-off demos.

Batch note for 6fc55994b561: keep provider keys out of the repo, set a per-session token ceiling, and store transcripts next to the eval fixtures so later model swaps stay comparable.