Contents
- What AET is (and what it is not)
- The statutory architecture: §§531–537
- The accumulated earnings credit (and why it’s “shrunk” in real terms)
- “Reasonable needs”: the specific–definite–feasible standard
- Bardahl formula: working capital math (peak vs average)
- IRS audit approach: analytics, ATGs, and litigation hazards
- Dividend timing tool: §563 (2.5-month “top-up” dividend)
- Modern context: Moore (2024), OBBBA (2025), CAMT interactions
- A 2025 defense playbook (minutes + models + hygiene)
1) What AET is (and what it is not)
AET is not a replacement for corporate income tax—it’s a penalty tax. It applies when a corporation is “formed or availed of” to avoid shareholder tax by letting earnings accumulate instead of paying dividends. The tax is computed annually on “accumulated taxable income,” not GAAP retained earnings and not E&P. :contentReference[oaicite:2]{index=2}
Why the IRS cares
Retention can defer shareholder-level tax and, in some estate scenarios, eliminate it through basis step-up. AET is the statutory backstop meant to neutralize that arbitrage by adding a 20% penalty. :contentReference[oaicite:3]{index=3}
2) The statutory architecture: §§531–537 in plain English
The core structure is: §531 imposes the tax, §532 defines who is subject, §533 creates a presumption of avoidance when accumulations exceed reasonable needs, §534 provides a mechanism to shift the burden to the IRS with a detailed statement, and §535 provides an accumulated earnings credit (a limited “safe harbor”). §537 defines “reasonable needs” and demands specific, definite, feasible plans for anticipated needs. :contentReference[oaicite:4]{index=4}
3) The accumulated earnings credit: the “safe harbor” that hasn’t kept up
The minimum credit is $250,000 for most corporations (reduced to $150,000 for certain personal service corporations). The report emphasizes the strategic reality: the credit isn’t indexed for inflation, so more businesses hit “AET territory” earlier in their lifecycle than Congress likely intended—making the reasonable-needs defense routine, not exceptional. :contentReference[oaicite:5]{index=5}
4) “Reasonable needs” defense: specific, definite, and feasible plans
The key qualitative defense is “reasonably anticipated needs,” but it requires more than a vibe. The standard is specific, definite, and feasible— shown through budgets, contracts, minutes, permitting steps, active negotiations, and a real timeline. The report’s case discussion highlights how “contingency” narratives can fail when they aren’t quantified and documented contemporaneously. :contentReference[oaicite:6]{index=6}
Strong reasons
Expansion, acquisitions, retiring bona fide business debt, and working capital buffers supported by math and records. :contentReference[oaicite:7]{index=7}
Weak reasons
Indefinite “rainy day” reserves, unrelated investment portfolios, and shareholder loans/personal spending through the company. :contentReference[oaicite:8]{index=8}
5) Bardahl formula: the quantitative working-capital defense
Bardahl working-capital analysis estimates how much cash a business needs to fund one operating cycle: inventory days + receivable days − payable days, applied to annual cash operating costs. The report emphasizes the most litigated choice: “average cycle” (IRS-favored) vs “peak cycle” (taxpayer-favored), especially for seasonal businesses where survivability is determined by the busiest month, not the average month. :contentReference[oaicite:9]{index=9}
Peak vs average is often the whole case
Using peak month inventory/AR can justify dramatically higher retention than an annual average. Defending “peak” typically requires monthly balance sheet detail, not just year-end financials. :contentReference[oaicite:10]{index=10}
6) IRS scrutiny: what triggers attention in 2024–2025
The report describes a modern enforcement posture: analytics-driven selection, examiners trained to “vertical analyze” Form 1120 balance sheets, and recurring red flags like growing cash, zero dividends, and shareholder loans. AET often appears as a collateral issue in broader audits involving related-party transactions or compensation disputes. :contentReference[oaicite:11]{index=11}
7) Dividend timing tool: §563 “top-up” dividends by March 15
AET has a unique timing release valve: dividends paid within 2 months and 15 days after year-end can be treated as paid during the prior tax year for AET purposes. Practically, corporations can run their Bardahl and “reasonable needs” work in February, then pay a targeted dividend by March 15 to eliminate excess accumulations for the prior year. :contentReference[oaicite:12]{index=12}
8) Modern context: Moore (2024), OBBBA (2025), and CAMT interactions
The report highlights three “why now” drivers: (1) Moore v. United States (2024) reinforced Congress’s power to tax undistributed earnings in certain structures, strengthening the conceptual footing of anti-deferral regimes; (2) OBBBA (July 4, 2025) changed taxable income dynamics (including expensing); and (3) CAMT creates a separate tax floor for certain large corporations, which can indirectly reduce AET base via higher federal taxes—but only where CAMT applies. :contentReference[oaicite:13]{index=13}
9) A 2025 defense playbook: minutes + models + hygiene
What to put in the permanent file
- • Annual Bardahl model (average + peak), with supporting monthly schedules
- • Board minutes mapping retained amounts to specific projects and timelines
- • Budgets, contracts, term sheets, permits, quotes (evidence of feasibility)
- • A formal policy against shareholder loans + evidence of enforcement
- • Dividend policy (even modest) + §563 “top-up” dividend procedures
- • A draft-ready §534 statement template for rapid response to IRS notices
The report’s core theme: contemporaneous documentation turns “intent” from a presumption into an evidentiary record. :contentReference[oaicite:14]{index=14}
Bottom line
AET risk is manageable if you treat retained earnings like a governed capital allocation program. Run the Bardahl math, document specific plans, avoid shareholder-loan optics, and use the §563 timing window to “true up” distributions after year-end. The best defense isn’t clever—it's organized. :contentReference[oaicite:15]{index=15}