The liability-driven book capstone — pension / insurer mandate.
You owe long-dated liabilities; the job is to backstop them — earn surplus over the liability while keeping the funded ratio from lurching. This assembles the family's liability ingredients into one book and judges it on funded-ratio terms (surplus vol, funded-ratio drawdown vs a long-duration liability), not raw return:
- LDI hedge (
binnacle) — long duration to track the liability - spread carry (
bedrock) — credit for surplus return - growth leg — plain equity, because
banyanshowed the endowment replica is dominated - convex tail (
contextual_rescue) — long-vol so the surplus survives the spread blowout
Book: SPY 30 · LQD 25 · TLT 25 · GLD 10 · VIXM 10. Liability proxy = TLT. Alpaca SIP daily.
KEEPER (liability mandate). backstop earns a positive surplus over the liability at a far smaller funded-ratio drawdown than a naive 60/40, with positive skew and crisis convexity. It does the mandate's actual job — fund the obligations without betting the funded ratio on rates — and it gives up absolute return to do it, which is the right trade for a plan.
Funded-ratio (the mandate's scorecard, vs a TLT liability):
| book | surplus ret | surplus vol | funded-ratio maxDD |
|---|---|---|---|
| BACKSTOP | +5.9% | +10.9% | −22% |
| naive 60/40 | +9.7% | +17.4% | −37% |
| pure LDI (all TLT) | 0.0% | 0.0% | 0% |
| return-seeking (SPY) | +15.4% | +24.5% | −48% |
Absolute terms:
| book | Sharpe | CAGR | maxDD | skew |
|---|---|---|---|---|
| BACKSTOP | +0.85 | +6.0% | −21% | +0.16 |
| naive 60/40 | +0.93 | +9.7% | −21% | −0.24 |
| SPY | +0.88 | +15.0% | −34% | −0.35 |
2016-01 … 2026-07 (10.5y).
- Positive surplus at controlled funded-ratio risk. backstop beats the liability by +5.9%/yr while cutting the funded-ratio drawdown to −22% (vs −37% for a naive 60/40 and −48% for all-equity). Its surplus efficiency is on par with 60/40, but its funded-ratio drawdown — the thing that makes a plan underfunded — is far shallower, and its skew is positive where 60/40's is negative.
- Crisis convexity: green in COVID (+3.1%) and the 2025 tariff tantrum (+1.7%), −6.0% in the 2022 rate
shock (vs 60/40 −7.4%, pure-LDI −8.2%). The long-vol tail (the rescued null) pays when the surplus would
otherwise crater — the lesson from
contextual_rescue, applied to a liability book. - Pure LDI is flat by construction (0% surplus, 0% funded-ratio drawdown — it is the liability); the point of backstop is to add surplus without reintroducing the funded-ratio swings a return-seeking book carries.
Real actuarial liability cashflows (not a TLT proxy) with key-rate hedging, plus after-tax treatment and realistic rebalancing/tail-layer friction — the honest plan-level build the ETF proxy only approximates.
export ALPACA_KEY_ID=... ALPACA_SECRET_KEY=...
python3 research/backstop_1.pybackstop (liability) joins bulkhead (preservation / FO),
belligerent (aggressive / HF), and
brawl (doctrine) — same keeper ingredients, different mandates.
Blaque Baux — a Carter Warrens research initiative. Research only; not investment advice. The verdict falls where the data puts it — a plan is judged by its funded ratio, not its return.