Decision — Skip Alpaca Fully Paid Securities Lending program (for now)
Context
Founder received the Alpaca Master Securities Lending Agreement (FPSL program, V01.2025.03, 49 pages) via iMessage on 2026-05-13 evening. Asked Ray for a read on the agreement and any reasons not to enroll.
The FPSL program lets Alpaca lend out the customer's fully-paid securities (typically to short sellers + DTC custodians for fail-to-deliver coverage), pay the customer a share of the lending fee, and post collateral at BMO Harris Bank N.A. via 17a-4 LLC as independent collateral administrator.
Decision
Skip enrollment for now. Founder iMessage 2026-05-13 20:23 ET: "No need. We can call this decided. I'm skipping it."
Reasoning surfaced during the conversation
Five concerns flagged from the agreement read:
- No indemnification. Schwab, Fidelity, IBKR all indemnify customers against borrower default. Alpaca explicitly does not — collateral at BMO is the customer's "only source of satisfaction." If a borrower defaults during an intra-day price gap, the customer eats the shortfall above the 102% collateral.
- Fee split. Standard tier is 20% to customer (Alpaca keeps 80%) — meaningfully below the 50/50 industry norm. Alpaca Elite tier gets 50% (industry norm).
- Cash-in-lieu kills qualified-dividend tax treatment. Replacement payments on dividend-paying stocks become ordinary income (up to 37% federal) instead of qualified-dividend rate (up to 20%). Alpaca makes no guarantee to recall before ex-div.
- SIPC protection forfeited on loaned shares. Voting rights waived during loan. 5-business-day cure period before customer can claim collateral on default.
- Re-enrollment friction. Annex E §7 contemplates restrictions on re-enrolling after opt-out.
Confirmed during the conversation:
- Customer retains beneficial ownership the entire time
- Sale auto-recalls the loan (T+1 or up to 5 BD settlement)
- Move/transfer (ACATS) auto-recalls
- Recall to vote / use as margin requires manual action
- Disabling FPSL is straightforward via Alpaca's Plans & Features tab
Founder-side reasons skip is the right call right now:
- Account holdings are small enough that lending revenue is rounding error
- Founder is still building the investing skill (per [[01-projects/investing/README]] strategy-vs-execution boundary table — co-define strategies, build skill incrementally)
- Adding lending-program complexity before the base layer (position sizing, thesis discipline, execution mechanics) is solid is premature optimization
- Re-enrollment is possible per the agreement (with possible friction)
- If founder later upgrades to Alpaca Elite ([[https://alpaca.markets/elite]]) the economics flip from below-industry to industry-norm
Ray's calibration error worth flagging
In the second iMessage of this conversation, the founder asked "50% is awfully high! That would mean risky, right?" which was a misread (50% is the fee SPLIT, not the yield rate — totally different concept). The original Ray summary message had presented the 50%/20% as numbers without explicitly framing them as fee-split percentages, which created the misread. Per [[feedback_calibrate_overconfidence]] — Ray walked back cleanly with "Sorry, I should have made that clearer last message." Lesson for future investing-product summaries: always state the units and what the percentage measures BEFORE presenting the number. Fee splits, yield rates, expense ratios, and gross-vs-net returns are easy to confuse.
Revisit conditions
Pull this back up if any of the following:
- Founder upgrades to Alpaca Elite tier (50% split = industry-normal economics, concern #2 dissolves)
- Account grows past the threshold where lending revenue becomes non-trivial (rule of thumb: $50k+ in lendable equity, since typical FPSL yields are 10-200 bps annually — needs meaningful base to matter)
- Founder establishes the base investing discipline (thesis docs, position-sizing rules, decision logs running cleanly) and is looking for marginal yield optimization
- Founder explicitly says "let's revisit Alpaca lending"
The four other concerns (no indemnification, qualified-dividend tax loss on dividend payers, SIPC loss while loaned, re-enrollment friction) still apply regardless of tier or account size — they're structural to Alpaca's program design.
Related
- [[01-projects/investing/README]] — investing project, strategy-vs-execution boundary
- [[01-projects/investing/decisions/2026-05-12-alpaca-account-status]] — parent Alpaca account decision (account opened or exploration-only is the prior question)
- [[06-reference/concepts/2026-05-13-fde-asymmetric-edge-rdco-positioning]] — RDCO's bigger investing question is asymmetric edge, not yield optimization on existing holdings
- [[~/.claude/projects/-Users-ray/memory/feedback_calibrate_overconfidence]] — calibration lesson
- [[~/.claude/projects/-Users-ray/memory/feedback_distinguish_decision_from_action]] — filing this decision log was a Ray-just-execute action, not a founder-input action