Finance Principles
What guides every finance decision before the spreadsheet opens?
The Loop
Finance creates value by closing the gap between cash truth and reported truth in real time, then routing every dollar against an explicit risk-and-return discipline so the business stays solvent today and compounds tomorrow.
That sentence is the gauge. Every workflow, instrument, and decision in this section either tightens that gap and that discipline — or it does not earn its place.
First Principles
Five truths that shape every other choice. Hold these and the rest follows.
- Cash is the only fact. Accruals tell a story; cash settles the score. The CFO defends the cash-truth view first.
- Reported truth must match cash truth in days, not weeks. Drift between the two is where fraud, error, and bad decisions hide.
- Compounding rewards the patient allocator. Time beats timing. The best way to win is not to lose.
- Crypto rails are CFO instruments, not a side venture. Stablecoins, on-chain reconciliation, chain-as-audit-trail, and DeFi yield expand what the CFO can measure, decide, and act on.
- AI agents are the new line staff. Close, AP, AR, FP&A, treasury, and audit each have an agent layer. The CFO retains the decision; the agent does the work.
Essential Data
Without these the function breaks. Each entry names the decision it drives.
- Cash position by entity, currency, and rail — drives every payable, payroll, and capital-allocation decision. Missing this, the CFO is blind.
- Stablecoin balances, custody location, redemption windows — drives treasury rebalancing and runway-in-stables forecasting. Native to the crypto-rails era.
- AR aging by customer, segment, and rail (fiat / stables / token) — drives collections priority, working-capital forecast, and AR-agent escalation.
- AP aging with payment-rail tag (ACH / wire / stables / on-chain) — drives payment sequencing, FX exposure, and AP-agent automation rules.
- Forecast vs actual variance per line, per period — drives FP&A re-baselining and CFO board narrative.
- Capital allocation ledger — 70/20/10 by tier with returns — drives reinvestment, kill-or-double decisions.
- Compliance and audit-trail completeness — drives tax filings, regulator response, investor diligence.
- Token-economics state if cap table includes tokens — vesting schedules, liquid vs locked, taxable events — drives comp accounting and tax timing.
Glossary
Terms used across this playbook. Each carries a specific meaning. Treat them as canonical.
- CFO — Chief Financial Officer. Apex decision maker for every finance workflow. Owns the gauge.
- Controller — Owns the books. Reports up to CFO. Runs the close.
- Treasurer — Owns the cash. Manages working capital, banking, custody, and stablecoin rails.
- FP&A — Financial Planning and Analysis. Forecasts, variance analysis, board reporting.
- Cash truth — What the bank, custody, and wallet balances actually show right now.
- Reported truth — What the books say, after accruals, accounting policy, and timing adjustments.
- Close — The process of locking a period's books. Monthly, quarterly, annual.
- Runway — Months of operations the current cash and stables position covers at current burn.
- Runway-in-stables — Runway measured in stablecoin reserves alone, separate from fiat runway.
- Stablecoin — Token pegged to a fiat currency (USDC, USDT, USDP, EURC). The default crypto rail for treasury.
- DeFi yield — Yield earned by deploying stablecoins to on-chain protocols. Subject to smart-contract and rate risk.
- On-chain reconciliation — Matching the ledger to wallet activity in real time, using the chain as the source of truth.
- Chain-as-audit-trail — Treating immutable on-chain history as primary audit evidence.
- 70/20/10 — Capital allocation discipline: 70% core, 20% adjacent, 10% experiments. Tiers are protected — experiments cannot cannibalize core.
- Agent — A named AI worker assigned to a specific finance function. Examples: close-agent, FP&A-agent, AP-agent, AR-agent, treasury-agent, audit-agent.
- Instrument — A measurable signal or a tool that produces one. Crypto and AI both deliver new instruments to the CFO toolkit.
Decisions, Data, Quality
Each row shows the decision, the data required, the current quality bar, and what breaks if the data is bad.
- Pay this invoice now / hold / route on-chain — needs AP aging + cash position by rail + counter-party history. Quality bar: complete and reconciled within 24 hours. Bad data → late fees, missed discounts, supplier escalation.
- Collect this AR or escalate — needs AR aging by customer + counter-party signal + payment-rail tag. Quality bar: refreshed daily. Bad data → working-capital strain, customer relationship damage.
- Rebalance stables to fiat / hold / deploy to yield — needs stables balance + redemption window + DeFi rate sheet + runway target. Quality bar: real-time. Bad data → liquidity gap or unforced yield drag.
- Fund this experiment / hold / kill — needs 70/20/10 ledger + experiment-tier returns + CFO conviction score. Quality bar: monthly. Bad data → tier-discipline breaks, urgent kills important.
- Raise capital now / wait — needs runway in months + stables runway + burn trajectory + market signal. Quality bar: weekly during raise window. Bad data → raising under duress.
- Recognize this revenue / defer — needs contract terms + delivery proof + on-chain settlement record where applicable. Quality bar: real-time on close. Bad data → restatement risk, audit qualification.
Loop Picture
Cash truth ──┐
├──► Reconciliation ──► Reported truth ──► CFO decision ──► Action
Reported truth ──┘ │
▼
70 / 20 / 10 allocation
│
┌────────┴────────┐
▼ ▼
Core ops Bets
│ │
└──── feeds ──────┘
cash
The CFO sits where the arrows meet. Every other Position feeds this loop. Every Process page describes one of the arrows. Every Platform tool is an instrument that measures one of the boxes. Every Performance gauge alerts when an arrow stalls.
Context
- Performance — Is the loop working? Numeric gauges and decision flows.
- Platform — Tools and instruments the loop runs on.
- Process — Workflows that carry the loop.
- Positions — Who runs each part of the loop.
- Tight Five — The five-questions pattern this playbook uses.
- Naming Standards — Where this glossary sources its discipline.