US Tax Design — Leveraged Investing

Date: 2026-03-25 Status: Draft — for sd-math M1-Task-2 (US interest-only projections) Scope: Taxation of equity investment returns for US investors in taxable accounts

---

1. Current Tax Treatment (2024–2026)

Capital Gains

Holding period	Rate	Notes
Short-term (< 1 year)	Ordinary income (10–37%)	Taxed identically to wages
Long-term (≥ 1 year)	0% / 15% / 20%	See income thresholds below

2024 Long-term CG thresholds (single filer):

• 0%: taxable income ≤ $47,025
• 15%: $47,026 – $518,900
• 20%: > $518,900

Net Investment Income Tax (NIIT): Additional 3.8% on net investment income for high earners — single filer income > $200,000 / married > $250,000. Applies to CG, dividends, interest, passive income. Effective max LT CG rate = 23.8% (20% + 3.8% NIIT).

Dividends

Type	Tax treatment
Qualified dividends	Same preferential rates as LT CG (0%/15%/20%) + potential 3.8% NIIT
Ordinary (non-qualified) dividends	Taxed as ordinary income

Qualified dividends require: paid by domestic US corporation (or qualified foreign), investor holds stock > 60 days in the 121-day window around the ex-dividend date. Most S&P 500 ETF distributions (SPY, IVV, VOO) qualify.

No gross-up mechanism — unlike Canada’s eligible dividend gross-up (1.25×) and dividend tax credit, US dividends are simply taxed at the preferential rate with no grossing-up.

Investment Interest Deductibility — KEY DIFFERENCE FROM CANADA

US rule: Investment interest expense (margin interest, investment loan interest) is deductible on Schedule A (Form 4952), but limited to net investment income for the year. Excess carries forward indefinitely.

Net investment income (for this purpose) includes: taxable interest, ordinary dividends, short-term CG, and other passive income. It excludes qualified dividends and LT CG — unless the taxpayer makes an irrevocable election to treat them as ordinary income (giving up the preferential rate in exchange for a larger interest deduction).

Implication for sd-math US model:

• Canadian analysis: interest_deduction = interest_paid × marginal_rate (full deduction)
• US analysis: interest_deduction is limited to net_investment_income × marginal_rate, which in a typical leveraged equity scenario (dividends + minimal realized CG annually) may be very small
• In practice, for a long-term buy-and-hold leveraged equity investor with mostly deferred CG: the US interest deduction may be near zero each year, with a large carryforward that only unlocks on sale
• This makes leveraged investing materially less tax-advantaged in the US than Canada

For initial sd-math implementation: model as interest_deduction = min(interest_paid, net_investment_income) × marginal_rate. Net investment income = annual taxable distributions (ordinary dividends + short-term CG distributed).

---

2. Historical Tax Rate Changes

Capital Gains — US History

Period	Max effective LT CG rate	Notes
Pre-1913	0%	No federal income tax
1913–1921	Same as ordinary income	16th Amendment; max ~7% initially, rising
1921	12.5%	Revenue Act introduced preferential rate (>2 yr holding)
1934–1941	Graduated exclusions	Complex; effective max varied ~15–30%
1942–1968	~25% effective max	Simplified; 50% exclusion, max ordinary 50%
1969	~35% effective max	Tax Reform Act; altered exclusion percentages
1976–1977	~35–39%	Increased via Tax Reform Act
1978	~28% effective max	Reduction via Revenue Act
1981	20%	ERTA (Economic Recovery Tax Act)
1987	28%	Tax Reform Act of 1986 eliminated preferential treatment — all CG taxed as ordinary income, effectively capped at 28%
1991–1996	28% cap	Explicitly capped at 28% even as ordinary rates rose
1997	20% (10% low bracket)	Taxpayer Relief Act; holding period ≥18 months
1998–2002	20% / 10%	Simplified to 12-month threshold
2003–2007	15% / 5%	JGTRRA — major reduction
2008–2012	15% / 0% low bracket	0% for 10/15% bracket taxpayers
2013–present	0% / 15% / 20% + 3.8% NIIT	ATRA + ACA; three-bracket structure

Key historical breakpoints for sd-math:

1. 1986: Preferential LT rate eliminated — CG taxed as ordinary income. Leveraged investing became significantly less efficient during this period.
2. 2003: LT rate dropped to 15%; qualified dividends introduced at same rate — most investor-favorable change in US history.
3. 2013: 20% rate added for high earners; NIIT (3.8%) introduced.

Dividends — US History

Period	Dividend treatment
Pre-2003	Taxed as ordinary income (max 39.6%)
2003–2012	Qualified dividends: 15%/5%; ordinary: ordinary income
2013–present	Qualified: 0%/15%/20% + NIIT; ordinary: ordinary income

Pre-2003 implication for historical analysis: For any historical simulation before 2003, dividends must be taxed at the investor’s marginal ordinary income rate, not at the preferential qualified dividend rate. This materially reduces the after-tax distribution in historical simulations.

---

3. sd-math US TaxProfile Design

@dataclass
class TaxProfile:
    country: str                    # 'ca' | 'us'
    marginal_rate: float            # Combined marginal rate (fed + state for US)

    # Canada-specific
    prov_tax_rate: float = 0.0
    fed_tax_rate: float = 0.0
    dividend_tax_rate: float = 0.0  # Effective rate on eligible dividends (incl. gross-up/DTC)

    # US-specific
    lt_cg_rate: float = 0.0         # 0.0, 0.15, or 0.20
    niit_rate: float = 0.0          # 0.0 or 0.038
    qualified_div_rate: float = 0.0  # = lt_cg_rate + niit_rate (effective)
    interest_deductibility: str = 'full'  # 'full' (CA) | 'net_investment_income' (US)

Sample US profiles for golden fixtures:

Profile	Marginal rate	LT CG rate	NIIT	Qualified div	Effective max LT rate
Middle Income	22%	15%	0%	15%	15%
High Income	37%	20%	3.8%	23.8%	23.8%

Note on state taxes: US state CG tax treatment varies widely (0% in TX/FL/WA; up to 13.3% in CA). For initial implementation, use federal rates only and document as a simplification. Add state tax as a future enhancement.

---

4. Key Algorithmic Differences: Canada vs US

Aspect	Canada	US
CG inclusion	50% of gain × marginal rate	Full gain × preferential rate (0/15/20%)
Dividend gross-up	1.25× eligible + dividend tax credit	No gross-up; preferential rate applied directly
Interest deductibility	Full deduction against all income	Limited to net investment income (Form 4952)
NIIT	None	3.8% on NII for high earners
Provincial/state tax	Province-specific (QC unique)	State-specific (omit in Phase 1)
Pre-2003 dividends	Always had preferential treatment	Taxed as ordinary income (major difference for historical)

---

5. Golden Fixture Parameters for US (M1-Task-2)

For consistency with CA fixtures: same loan, rate, horizon, return distribution.

Parameters: $100,000 loan, 9% interest rate, 10-year horizon, 70% deferred CG / 25% taxable CG / 5% dividends.

Two US tax profiles (mirrors CA 35%/50%):

Profile	Marginal rate	LT CG rate	NIIT	Effective div rate	Comparable CA rate
US-Middle	22% (fed 22%)	15%	0%	15%	~35% CA
US-High	37% (fed 37%)	20%	3.8%	23.8%	~50% CA

Fixtures to generate from LevPro (if US mode exists) or calculate from first principles:

• 2 one-page summaries (US-Middle + US-High) with returns: 0%, 3%, 7%, 10%, “Better Than”
• 2 leverage projections (US-Middle at 10%, US-High at 7%)

If LevPro has no US mode: derive expected values analytically using the algorithm in sd-math-design.md with US tax parameters substituted. Validate against manual calculations for one return scenario before writing pytest fixtures.