Prism API scores public companies on 57 validated psychological dimensions — extracted from SEC filings and earnings transcripts using proprietary psychometric methodology.
| What Prism measures | What competitors measure |
|---|---|
| Organizational psychopathy | Word frequency |
| Ethical leadership | ESG checklist compliance |
| Psychological safety | Glassdoor star ratings |
| Innovation orientation | Patent counts |
| Moral disengagement | Nothing comparable |
100% public data · No PII · No MNPI · No web scraping
Aug 2025 — Mar 2026 · 50L/50S · Semi-annual rebalance · Win rate: 87.5% · OOS Sharpe 1.65
| Test | Result |
|---|---|
| OOS Sharpe (2018-2025) | 1.65 |
| Fama-French 5-factor | All signals survive |
| Post-2021 persistence | No fade |
| Multi-architecture | 3 models consistent |
| 600+ ML configs tested | None beat composite |
| DOJ prosecution (LOO) | AUC 0.956 |
Under review at Management Science (INFORMS)
| Signal | Annual Alpha | Sharpe | t-stat |
|---|---|---|---|
| Uniqueness Premium | 14.5% | 1.26 | 4.55 |
| Innovativeness Premium | 16.2% | 0.99 | 5.57 |
| Risk Premium | 18.2% | 1.26 | 5.27 |
| Tier | Includes | History | Price |
|---|---|---|---|
| Tier 1 — Index Suite | 4 indices (returns + constituents) + prosecution risk | 5 years | $75K/yr |
| Tier 2 — Index + Raw Scores | Tier 1 + all 57 psychometric dimensions + derived signals | Full (2001–present) | $150K/yr |
| Tier 3 — Exclusive | Tier 2 + custom signals, BYOD scoring, white-label | Full + custom | Ask |
April 2026 —
Auction for exclusive rights to Prism Flagship Signal. Starting bid $2.5M. 24-month exclusive access. One winner.
May 2026+ —
Indices and raw scores available on Neudata. Tier 1 & 2 open access.
| Allocation | Net Alpha | 1yr Alpha | 3yr Alpha | Fair Exclusive Price |
|---|---|---|---|---|
| $100M | 37.0% | $37M | $148M | $15M |
| $500M | 34.8% | $175M | $679M | $68M |
| $1B | 33.1% | $331M | $1.27B | $127M |
| $2B | 30.8% | $609M | $2.3B | $231M |
| Year | Clients | Revenue |
|---|---|---|
| 1 | 3 | $300K |
| 2 | 8 | $900K |
| 3 | 14 | $1.8M |
| 5 | 28 | $3.6M |
| Revenue TBD — dynamic pricing linked to signal performance | ||
| Seed size | Y5 AUM | Annual fees |
|---|---|---|
| $100M | $200M | $8M |
| $500M | $1B | $40M |
| $1B | $2B | $80M |
| Performance deteriorates beyond $2B — requires universe expansion | ||
Standard 2% mgmt + 20% perf fees
600+ ML configurations tested. None beat the simple psychometric composite. The methodology cannot be replicated by brute-force computation.
7,377 tickers, 257K documents, 25 years of quarterly history. Years of compute and data licensing to replicate.
Management Science submission. Morris Lab at Columbia. No competitor has peer-reviewed validation.
Domain-adapted psychometric items are proprietary. Scoring pipeline undisclosed. Construction methodology proprietary.
+52.6% live since founding (August 2025).
Net alpha after round-trip execution costs at semi-annual rebalance.
| AUM | Position Size | Net Alpha | Execution |
|---|---|---|---|
| $100M | $1.9M | 37.0% | Clean |
| $500M | $9.6M | 34.8% | Clean |
| $1B | $19.2M | 33.1% | Manageable |
| $2B | $38.5M | 30.8% | TWAP/VWAP |
| $5B+ | Requires universe expansion | International, large-cap |
Square-root market impact model (Almgren-Chriss) applied to current holdings with $1M ADV liquidity floor.
| Gen Tsudaka, Ph.D. | Founder & CEO | Theory and methods |
| Birgitte Sylta | Management | Deloitte, NYU Mgmt |
| Brian Gully | Communications | Project Coordinator, Brown |
| Patrick White | Development | PMF, institutional outreach |
| Michael Cronin | Strategy | Partnerships |
| Jonas Muthoni | Public relations |
| Ying Hong | External security consultant |
| Margaux Wienk | External scientific advisory |
| 1 engineer | Pipeline automation, API, data delivery |
info@stronglight.io
info@stronglight.io · prism.stronglight.io
This document is for informational purposes only and does not constitute an offer to sell securities. Stronglight Inc. is not a registered investment adviser. Past performance is not indicative of future results. All data derived from public sources.