NSA-PB closed at $32.99 on April 26, 2026 — a 4.55% current yield. NSA-PA, the sister Series A preferred from the same issuer, in the same merger, with the same 1:1 conversion to PSA preferred, and the same 6.000% coupon, closed at $22.97 — a 6.53% yield. Two preferred shares of National Storage Affiliates, scheduled to convert to PSA preferreds with materially unchanged terms on the same date, trade 200 basis points apart.

The proof is the sister

Per the merger agreement filed March 17, 2026 (8-K, Item 1.01), each NSA-PB share "will automatically be converted into the right to receive one validly issued Parent preferred share... having rights, preferences, privileges and voting powers that are MATERIALLY UNCHANGED, including with respect to dividend rights, liquidation preference, redemption rights, voting rights and ranking." NSA-PA gets identical treatment under the same agreement.

NSA-PA carries one structural difference: it is callable now (issued October 2017, callable since October 2022). NSA-PB is non-callable through September 15, 2043. That difference creates a ceiling on NSA-PA's price — PSA could call NSA-PA at $25 par + accrued at any time post-merger — but NSA-PA is at $22.97, below par, so the call right enhances rather than caps NSA-PA's expected return. It does not explain why NSA-PB trades 200 basis points tighter than its identical sister.

PSA's existing 14 preferred series (PSA-PF through PSA-PS), with coupons spanning 3.9% to 5.6%, all cluster at 6.13-6.31% yield (mean 6.25%, std dev 6 bp). The curve is flat across coupon and duration. NSA-PA at 6.53% sits 28 bp wide of that curve — a reasonable deal-close uncertainty premium for a preferred that hasn't converted yet. NSA-PB at 4.55% sits 170 bp inside the curve, with no PSA-internal precedent for such a yield.

Implied fair value for NSA-PB at PSA curve mean: $25 × 6.00% / 6.25% = $24.00. At NSA-PA's deal-close-adjusted yield (6.53%): $22.97. NSA-PB at $32.99 is 38-44% above either anchor.

What retail misread

The 2043 non-call period is being priced as a duration premium. Buyers appear to treat long call protection as a feature worth paying for — a yield-stable instrument insulated from rate changes. But the non-call is a restriction on the issuer's right to redeem at par; it doesn't insulate the present value of cash flows from market yield repricing on conversion. The protection is against an event that doesn't help current buyers (redemption at par would lose them money from $32.99) and doesn't prevent the event that does (mark-to-market reprice to PSA's curve post-conversion).

The microstructure tells the story plainly. On March 30, 2026, NSA-PB printed an intraday high of $55.08 — momentarily implying a 2.73% yield on a security converting to a 6.25% PSA preferred. That isn't mispricing in any normal sense; it's a price that has lost contact with what the instrument is. The Mar 27 - Apr 6 volume spike (137k-356k shares vs 1-2k normal daily) confirms the flow as retail and momentum-driven rather than informational.

The 10-K/A filed April 22 is unrelated noise — a routine Part III governance amendment per General Instruction G(3), no restatement, KPMG unchanged.

The historical analog

LBRDP (Charter / Liberty Broadband, announced November 13, 2024) had the same 1:1 conversion mechanic with materially unchanged terms and ≈13 years of remaining call protection. Pre-announcement: $23.79. Day of announcement: $24.07 (+1.2%, on 30x volume). One month later: $24.15. Five months later: $24.10. The market correctly priced the announcement as a credit transfer with a 1-2% adjustment. NSA-PB's +57% YTD is unprecedented against this analog. The pattern is rare — under five instances per decade in REIT preferreds, because most modern preferreds are callable within five years.

The mispricing persists because NSA-PB has 5.67M shares outstanding, no options, no published institutional coverage, and 2-3k normal daily volume. There is no arb infrastructure to fade it.

Risks (ranked by impact on action)

1. Borrow scarcity is the binding constraint. A naked NSA-PB short returns ~+18% gross from price drift to fair value over nine months, but borrow on a 5.67M-share retail-held preferred is likely 15-35% all-in. Net EV after borrow and dividend pass-through: approximately -4% on a clean short. A pair (long NSA-PA / short NSA-PB) clears Sharpe 1.0 only with sub-15% borrow — unlikely at retail prime brokerage. The mispricing is real; the short trade is sub-scale.
2. PSA preferred curve compression. A 170 bp tightening would lift fair value to ≈$30 and narrow the gap from -27% to -9%. Stress test: requires Treasury rates falling ≈150 bp or a flight-to-quality bid into PSA paper. P < 5%.
3. Squeeze re-ignites. New retail wave or social-media flow pushes NSA-PB above $40. Tail risk, not base case.
4. Merger renegotiated with cash redemption at premium. The 8-K language ("materially unchanged") is binding and preferred holders don't vote on the merger; renegotiation requires re-vote. Implausible. P < 5%.

Discovery calendar

What would change our mind

• PSA preferred curve compresses below 5.5% across the complex. Lifts fair value to $27+ and narrows the gap to single digits.
• Definitive proxy or S-4 reveals materially modified preferred terms versus the March 8-K. Breaks the conversion-mechanics edge.
• Sustained NSA-PB volume above 10k/day without news, suggesting a forced-flow holder (CEF/ETF accumulation rather than retail churn). Changes the squeeze-unwind hypothesis.
• NSA-PB borrow becomes broadly available at <10% all-in. Doesn't change the thesis but changes whether it is harvestable.

Implication for existing holders

This isn't a new-position recommendation — borrow scarcity makes a short sub-scale, and the math says don't initiate. But existing NSA-PB holders sitting above $25 face a -27% expected drawdown to fair value over six to twelve months, with no offsetting alpha beyond a $1.13 dividend stream. The dividend is safe (14.7x Core FFO coverage, 30% NOI decline still leaves 10x), but a safe dividend on a security 38% above its post-merger anchor is the same trade as buying a 6.25% PSA preferred at $34.

Evidence