DBC Global Compass — Seed Completion Addendum (2026-07-12)

SIMULATED — demonstration research. This item was AI-generated as seed content for the DogBone Capital platform. It is not human analyst work and not investment advice.

1. What was rejected, and why

Prior to this remediation pass, seven designed lines were submitted to dbc-global-compass and rejected by the fund's pre-trade IPS check with a 422 response. Verbatim rejection reason (IPS_UNIVERSE check):

"EQUITY is not permitted by fund mandate." (actual: "EQUITY", limit: "ETF,CRYPTO")

Rejected symbols: TLT, IEF, EEM, GLD, UUP, BTCUSD, ETHUSD.

Root cause (hypothesis, now confirmed by this pass): the instrument-classifier reference data is misclassifying these tickers' asset class as EQUITY instead of ETF (or, for BTCUSD/ETHUSD, CRYPTO) — even though the fund mandate's IPS_UNIVERSE check explicitly permits ETF,CRYPTO. This is a classifier/reference-data bug, not a mandate design choice; the policy already allows both fund types. Notably, BTCUSD and ETHUSD were also caught, so the bug isn't confined to ETF-vs-equity confusion — it also misfires on crypto tickers that should hit the CRYPTO bucket the mandate explicitly allows.

2. Hypothesis test and substitute mapping

Hypothesis: benchmark-allowlisted ETFs (the platform's index/benchmark catalog) classify correctly as ETF and clear IPS_UNIVERSE, unlike the rejected names. Confirmed by probing AGG first: IPS_UNIVERSE returned "ETF is permitted by fund mandate." (actual: "ETF") and the order was accepted/resting. All five substitute lines below cleared the same check on every tranche.

Designed line (rejected)	Substitute placed

TLT + IEF (duration/rates barbell)	AGG	$129,773.07 (1,323 sh, 3 tranches)	Core aggregate-bond ETF, ~6yr effective duration; single-instrument proxy for the long+intermediate UST barbell. Adds corporate-credit/MBS exposure the barbell wouldn't have had — imperfect.
(mandate-listed credit sleeve, unseeded)	HYG	$59,934.40 (752 sh, 2 tranches)	Direct implementation of the mandate's explicit credit allowance — not a stand-in for a different asset class.
EEM (emerging-market equity)	ACWI	$99,798.78 (633 sh, 3 tranches)	Global-equity ETF (~10% EM-weighted, ~60% US-weighted). Preserves a global-equity diversification view but materially dilutes the EM-specific tilt the designed line intended.
(treasury/cash carve-out, partial)	BIL	$59,837.73 (654 sh, 2 tranches)	1-3mo T-bill ETF, near-zero duration; dry-powder parking, distinct role from the AGG duration line.
GLD (gold hedge)	XLE	$59,921.60 (1,088 sh, 2 tranches)	Energy-sector equity ETF. Explicitly not a real gold substitute — XLE is cyclical/high-beta and will move with risk-off episodes, not against them. Placed only to give the book some commodity-linked sensitivity while gold stays unreachable.
UUP (dollar)	None	—	No in-mandate ETF proxy identified; theme remains unfilled.
BTCUSD / ETHUSD (crypto sleeve)	None	—	Mandate explicitly permits `CRYPTO`, but both tickers hit the same classifier bug as EQUITY; no workaround attempted this pass.

Symbol	Shares	Est. Notional	Weight
SPY (seed)	170	$128,367.00	12.84%
EFA (seed)	1,149	$119,875.17	11.99%
AGG (substitute)	1,323	$129,773.07	12.98%
ACWI (substitute)	633	$99,798.78	9.98%
HYG (substitute)	752	$59,934.40	5.99%
XLE (substitute)	1,088	$59,921.60	5.99%
BIL (substitute)	654	$59,837.73	5.98%
Total		$657,507.75	65.75%

DBC Global Compass — Seed Completion Addendum (2026-07-12)

Contributors

1. What was rejected, and why

2. Hypothesis test and substitute mapping

Version history

3. Resulting book (post-remediation)

4. Themes still missing