Time Horizon

18-24 months. BLA submission target Q4 2026. PDUFA (if Priority Review) Q2-Q4 2027. Buffer for CRL cycle extends to June 2028. Cash runway expires Q2-Q3 2027 — the financial clock runs in parallel with the regulatory clock, and they converge at the same point.

This is not a trade you hold for years. The binary resolves within two years or the company runs out of money. The time horizon shapes everything: no time for catalysts to compound, no time for thesis evolution, no time to be patient. You're right within 24 months or you're wrong.

Base Rate

Reference class: BTD-designated orphan biologics seeking accelerated approval
Base rate: ≈69% approval (FDA 2006-2023 BTD data)
Adjusted for orphan + rare pediatric: ≈75-80%
Prior odds (conditional chain, not base rate): 0.65 × 0.90 × 0.75 = 0.44
Buffered to June 2028: 0.55
Prior odds: 1.22 (55:45)

The base rate for BTD drugs (≈69%) is meaningfully higher than all-comers (≈12% from IND). But SPRB's conditional chain discounts from the base: P(BLA filed) = 65% is the weakest link. Manufacturing PPQ, not science, is what drags this below the class average.

The 55% estimate embeds three sub-probabilities: filing (65%), acceptance (90%), approval (75%). The DNLI precedent (March 2026) shifts the latter two upward. The filing probability is gated by manufacturing — no primary source evidence to update this beyond management guidance ("on track").

B — Business Model

Single-asset, pre-revenue biotech. Binary CF structure.

TA-ERT (tralesinidase alfa) is a recombinant fusion protein — rhNAGLU fused to IGF2 — delivered via weekly intracerebroventricular (ICV) injection to treat MPS IIIB (Sanfilippo Syndrome Type B). This is an ultra-rare fatal pediatric lysosomal storage disease affecting approximately 100-200 treatable patients globally. Zero approved therapies exist. Zero active competitors.

The molecular logic solves three problems sequentially: (1) supply the missing enzyme (rhNAGLU), validated by 30 years of approved ERTs; (2) bypass the blood-brain barrier via ICV delivery, validated by BioMarin's Brineura (2017); (3) get the enzyme INTO neurons via IGF2 fusion tag that hijacks CI-M6PR for internalization — this is the key innovation. Without it, enzyme reaches CSF but cannot enter neurons.

Revenue decomposition if approved:

Unit economics: Weekly ICV infusions at specialty centers. $500-800K/year per patient (Brineura-comparable pricing). Monopoly = no price competition. 7-year orphan drug exclusivity. Fully outsourced manufacturing model.

Patient pool: ≈2 new MPS IIIB diagnoses per year in US, ≈38 living, ≈100-200 globally treatable. MPS IIIB constitutes ≈20% of all MPS III cases. Not currently included in newborn screening panels — post-approval RUSP nomination (3-5 year horizon) could 2-3x the treatable pool.

Phi — Financial Trajectory

The derivative is mixed. R&D declining sharply ($46M FY2024 to $20M FY2025) as tildacerfont program winds down. G&A rising ($15M to $17M) from commercial buildout. Q4 2025 G&A spiked to $7.0M (vs $3.2M Q3) from professional fees and CCO team ramp. Steady-state burn ≈$9-10M/quarter, accelerating toward $12-15M as BLA costs, confirmatory trial, and commercial infrastructure scale.

Cash bridge:

Cash runs out at or before the FDA decision. Equity raise is virtually certain (75% probability). ATM via Jefferies established March 2026 (3% commission) — the mechanism exists. Fully diluted: ≈2.0M shares (46% max dilution from current 1.37M outstanding).

Avenue Capital facility: $50M total, $15M drawn (Tranche 1). Tranche 2 ($10M, Mar-Sep 2026) and Tranche 3 ($15M, Sep 2026-Mar 2027) are conditional on "key regulatory milestones" — likely BLA filing and FDA acceptance, respectively. No minimum cash covenant (favorable). All assets pledged including IP. Interest rate: greater of prime + 5.25% or 12.25% floor.

The acquisition math is remarkable: SPRB bought TA-ERT from distressed Allievex (ABC proceedings) for net $11.1M in November 2024. That's a cheap option on a $500M+ potential asset inside a $96M vehicle.

Capital raise history: $293M aggregate gross proceeds inception through September 2025, including Oct 2025 PIPE ($50M at $68/share). The company raises capital every 6-18 months.

K — Competitive Position

Monopoly. HHI = 10,000. S = 100%. dS/dt = 0.

Competitive landscape: All dead. Abeona ABO-101 (MPS IIIB gene therapy) discontinued due to drug supply issues. Lysogene LYS-SAF302 (MPS IIIA gene therapy) failed Phase 2/3, company liquidated. The closest adjacent programs are Ultragenyx UX111 (MPS IIIA, 2nd CRL for CMC) and Denali DNL126 (MPS IIIA, Phase 1/2). Different disease subtype.

Replication timeline for a new entrant: 7-10 years. Natural history comparator data becomes ethically irreproducible post-approval. The clinical dataset IS the moat.

Structural vulnerability: SPRB does not own TA-ERT patents. The program operates entirely under an exclusive license from BioMarin (originally Oct 2019, acquired via Allievex). BioMarin can terminate if SPRB ceases all material R&D, challenges licensed patents, or becomes insolvent. In practice, BioMarin has no incentive to terminate while SPRB advances the BLA — BioMarin receives $25.5M at first milestone plus royalties for zero development cost. But in insolvency: BioMarin recovers a near-approval asset for free. This is the structural tail risk that the simple binary model understates.

G — Governance

Insider ownership (officers + directors): ≈6.9% pre-PIPE, ≈4-5% post-dilution.

Insider buying signal is the strongest governance evidence. CEO Szwarcberg ($795K), President Gharib ($320K), Director Ways ($73K) — all open market Code P purchases at $58-80 in October-December 2025. Total: $1.19M in personal capital during the company's darkest period (post-delisting, post-55% RIF). This is voluntary money, not grants.

CEO comp: $1.5M total (FY2024). PSUs tied to "financing and strategic objectives" — vested at PIPE closing. Reasonable for micro-cap biotech.

Executive Chairman Grey was a BioMarin director 2005-2021. This is the connective tissue between SPRB and its licensor — a relationship asset, not a conflict.

Post-RIF rebuilding: SVP Regulatory + Quality (Feb 2026), SVP Clinical Development Ops (Feb 2026), CCO Dale Hooks (Mar 2026, $480K base, 30yr rare disease experience). Board added Keli Walbert (Dec 2025, rare disease commercial leader). Pattern: deliberate commercial infrastructure buildout, not crisis staffing.

Shareholder base transition. VCs who backed old SPRB (tildacerfont era) are exiting: Abingworth dropped below 5%, board rep Muralidhar resigned October 2025. HealthCap sold at $206, down to 3.4%. New money: Millennium Management (Izzy Englander) took 9.6% passive position (13G Feb 2026). This is fresh institutional capital validating the TA-ERT pivot.

Going concern: present in all recent filings. Constrains institutional participation regardless of fundamentals.

Beta — Factor Profile

Factor decomposition is inapplicable. With 1392% annualized idiosyncratic volatility on a 1.4M share float, any regression against SPY or XBI produces noise betas and R-squared near zero. Beta 3.82 is an artifact of thin trading, not systematic exposure.

%Idio variance: ≈100%. Not in any ETF, not in any factor index, not in any systematic quant screen. No listed options — $96M market cap is below the threshold for options market makers.

This is pure binary. Returns are driven entirely by: (1) FDA approval probability updates, (2) dilution events, (3) insider/institutional buying, (4) biomarker validation catalysts. No factor model tells you anything useful at the stock level. The Paleologo framework applies only at portfolio level — sizing, diversification, Kelly.

Delta — Expectations Gap

For a pre-revenue binary biotech, the only extractable parameter is P(approval). Everything else is roughly consensus. The price IS a probability.

$69.53 = P* x V_approval + (1-P*) x V_failure

Market-implied P(approval) sensitivity:

Central estimate: P_market ~ 25-35%. Our conditional chain: 55%. Gap: +20 to +30 percentage points.

Four analysts cover SPRB, all bullish (mean target $198, range $140-283). Stock at $69 trades at a 65% discount to consensus. In a large-cap, this would mean analysts are wrong about terminal value. In a micro-cap binary biotech with going concern, the discount rate is the variable — not the terminal value.

Why the gap persists:

1. Structural access barrier. Going concern language blocks compliance screens. 1:75 reverse split flags risk systems. $96M cap below institutional minimums. 1.4M share float prevents position-building. No listed options prevents synthetic exposure. This is not a disagreement about probability — it's a constraint on capital deployment.

2. Inference chain. DNLI AVLAYAH approved March 24, 2026, validating CSF HS as surrogate for MPS accelerated approval. The read-through to SPRB requires connecting: DNLI approval -> CSF HS biomarker class validated -> SPRB's CSF HS-NRE is a related assay -> FDA confirmed acceptance in Feb 2026 Type B meeting post-RGNX CRL. Four analysts get it. The marginal price-setter does not. SPRB up only 28% on a catalyst worth materially more.

3. Reflexive dilution loop. Going concern -> no institutions -> depressed price -> worse dilution terms -> lower value per share -> more going concern fear. Self-reinforcing.

4. Analyst target cuts. Citizens: $259 -> $170 (March 2026). Directionally inconsistent with DNLI de-risking. Why? BLA timeline slip (Q1 -> Q4 2026) extends the cash bridge problem. Analysts are saying: drug works, but the company might not survive long enough. That's a financial structure call, not a science call.

Dimension-level gap ranking (by |Delta| x q):

Alpha vs Beta

Expected return (18mo):  +51%
  Market beta:           ≈0% (beta is noise at 1392% vol)
  Sector beta:           ≈0% (not in XBI, too small for sector factor)
  Idiosyncratic:         +51% (100% of the return is stock-specific)

This is the rare case where alpha vs beta decomposition is trivially simple: it's all alpha. There is no beta to extract. The 3.82 reported beta is a statistical artifact of thin trading, not systematic exposure. At $96M cap with 100K daily volume, no factor model has explanatory power.

The alpha itself decomposes into:

• Regulatory edge (+30-40pp): DNLI validation not fully priced, P_us > P_market
• Financial structure discount (-10-15pp): Market correctly prices going concern, dilution risk
• Net idiosyncratic alpha: +20-25pp

Steelman Bear Case

The strongest argument against the thesis is not that TA-ERT fails. It's that SPRB the company cannot survive long enough to realize TA-ERT's value.

The drug works. Four concordant endpoints (CSF HS-NRE p<0.0001, BSID-C p=0.005 and p<0.0001, CGMV stabilization, VABS-II stabilization). 7.3 years of safety data. FDA confirmed the surrogate. DNLI validated the biomarker class. The science risk is real but modest.

The company risk is severe:

• Cash-zero arrives at or before PDUFA. Equity raise is virtually certain.
• ATM at current price ($70) means ≈30% dilution. At distressed price ($30-40) means ≈50%.
• The reflexive loop amplifies: going concern -> no institutions -> low price -> worse dilution -> more going concern.
• BioMarin can terminate the license on insolvency. Avenue Capital has a lien on all IP. In bankruptcy: Avenue seizes IP, BioMarin terminates license, shareholders get zero.
• The BLA already slipped 9 months (Q1 -> Q4 2026). Another slip pushes cash-zero before filing, not just before PDUFA.

The honest engagement: this argument is partially correct. The financial structure risk is real, and the market is not wrong to discount for it. The bull thesis requires believing that the ATM + Avenue conditional tranches can bridge the gap, that PPQ completes on time, and that BLA files Q4 2026. If any of these break, the reflexive loop accelerates.

Where the bear case weakens: SPRB has the ATM mechanism (Jefferies, March 2026). They can drip-feed shares at market prices rather than a distressed secondary. Avenue's no-minimum-cash covenant means no forced acceleration. The PRV alone ($200M+) resolves all financial problems — if they get to approval. And the 55% probability estimate already discounts the 35% chance that BLA never files. The bear case is in the price at 25-35% implied P.

Kill Criteria

Thesis invalidated if:
- BLA submission delayed beyond Q2 2027 — cash-zero before filing
- FDA issues guidance distinguishing CSF HS-NRE from DNLI's assay — regulatory path collapses
- BioMarin terminates license (any trigger) — asset gone
- PPQ manufacturing failure disclosed — BLA timeline collapses
- Avenue Capital accelerates loan (despite no cash covenant, other triggers exist)

Thesis weakened if:
- ATM usage at prices below $50 — severe dilution erodes per-share value
- Going concern language intensifies in FY2025 10-K — institutional access further constrained
- Confirmatory trial enrollment stalls — post-marketing commitment at risk

Thesis strengthened if:
- Avenue Tranche 2 drawn ($10M, requires regulatory milestone) — confirms BLA progress
- BLA files on time (Q4 2026) — resolves PPQ uncertainty, confirms timeline
- Analyst updates post-DNLI with raised targets — consensus catching up
- NBS developments for MPS IIIB — long-term TAM expansion

Key Risks

Manufacturing (PPQ). The gating factor for BLA. CMOs unnamed in filings (competitively sensitive but also prevents independent assessment). BLA already slipped 9 months because PPQ wasn't ready. Management says "on track" — but they said Q1 2026 was on track until it wasn't. No primary source evidence to verify PPQ status independently.

Dilution. Virtually certain (75% probability). Severity depends on timing and price. ATM at $70 is manageable (≈30% dilution). Distressed raise at $30-40 is destructive (≈50%). The reflexive loop means dilution timing matters: before BLA filing (weak negotiating position) vs after (stronger, higher price expected).

Biomarker assay distinction. SPRB uses CSF HS-NRE. DNLI's AVLAYAH used CSF HS. RGNX's rejected RGX-121 used CSF HS D2S6. FDA distinguishes between assay types, not the biomarker class. FDA confirmed SPRB's HS-NRE in Feb 2026 Type B meeting post-RGNX CRL. But formal BLA review could revisit this. Low probability, high impact.

Small dataset. N=22 treated, up to 7.3 years. For an ultra-rare disease with ≈2 new US patients per year, this is about as good as it gets. The confirmatory trial (14 patients, 5 years, placebo-controlled) must be initiated before approval — enrollment in an ultra-rare population adds execution risk.

BioMarin license dependency. SPRB does not own the patents. Insolvency triggers termination. This creates a binary tail: either SPRB survives to approval (BioMarin happy with milestones + royalties) or SPRB fails and BioMarin recovers a near-approval asset for free. For SPRB shareholders, the downside in the failure scenario is genuinely zero, not "residual cash value."

What to Watch

LR Signal

LR = 1.8 (mild-to-strong bullish divergence)

The DNLI approval is Tier 1 evidence (FDA accelerated approval, irreversible) with LR 2.8 for the regulatory pathway. Four concordant clinical endpoints at p<0.0001 and p=0.005. $1.19M insider buying at $58-80. Monopoly competitive position.

Offset by: going concern (Tier 1, bearish), BLA slip (Tier 1, 9 months), certain dilution, BioMarin license dependency. Analysts cutting targets not raising them.

Net: the bullish evidence is higher quality (FDA actions, SEC filings) and more numerous than the bearish evidence (which is primarily financial structure). But the financial structure risk is not hypothetical — it's in the cash flow statement. LR reflects genuine bullish divergence from market pricing, constrained by a real path-dependency problem.

Evidence