EQ
Eloquentix
STRA Signal
Strategic Education
weekly enrollment signal.
Capella + Strayer enrollment read from panel-normalized public signals:
reported USHE numbers, Google Trends, SEC filings, FAFSA pace, GI Bill,
NCLEX pass rates. Simple forecast. Honest backtest. One page.
Backtest-implied forecasts
One panel per target series. Each fits leading signals via intercept-free OLS,
aggregates by r², and shows the naïve baselines a model must beat.
Analyst workpapers
Specialist agent reads, before synthesis. Open any card to see the raw analysis.
Backtesting
Walk-forward fold-by-fold predictions for each target. At each held-out
quarter the model is refit on prior history only, then asked to predict
the held-out quarter. Compare the model's predictions to the naïve
baselines that anchor credibility — if the model isn't beating both,
say so honestly.
Data sources & methodology
Every signal in the app, with source, scrape method, coverage, and how
we know there's new data. Live = pulled on every request.
Cron-refreshed = scraper runs Mon 04:00 UTC.
Static-curated = hand-reviewed panel; refresh date stamped per signal.
Known limitations
Things this tool doesn't do, or can't do honestly with the data it has.
Read these before acting on the brief.
-
USHE total only — no Capella vs Strayer split per quarter.
We confirmed by reading every 10-Q since 2017: STRA does NOT publish
quarterly Capella vs Strayer enrollment. USHE is reported as a
single segment in MD&A; per-university splits appear only in the
annual 10-K (Item 1) and in occasional press-release commentary.
The combined view masks divergent dynamics (Capella healthcare-led
growth vs Strayer post-marketing-cut flat).
-
10-Q signal series start in 2021–2022. Employer-share %
(16 quarters from 2021Q1), student persistence (16 quarters from
2020Q4 with Q3 gaps from disclosure-lag), USHE/ETS segment revenue
(11 quarters from 2022Q3) — STRA changed disclosure templates in
2022, so pre-2021 narrative doesn't match the regex. Older filings
had different MD&A structure that we don't yet parse.
-
Strayer marketing-cut structural break (Q3 2022) not modeled separately.
Strayer cut marketing >50% from Q3 2022, but since we don't have
quarterly Strayer-only enrollment, we can't fit a regime dummy. The
break gets absorbed into residuals on the combined USHE series.
-
Google Trends excluded from regressions. pytrends
normalizes the historical series against today's window, introducing
look-ahead bias. We use it as a direction-only sentiment prior at
the latest reading (β=0.5, σ=1.0), not as a fitted feature. Honest
fix requires point-in-time Trends snapshots over multiple weeks
(not done yet).
-
FAFSA + GI Bill + NCLEX still snapshot, not live.
FSA's Data Center is JS-rendered (Playwright required); VA's
Comparison Tool API doesn't expose quarterly beneficiary counts
publicly; NCLEX state-board reports are PDF-heavy. Each chart card
shows its snapshot date in the footer.
-
2019Q3–2020Q1 partly missing. The 8-K extractor
dropped values below 50K as pre-integration segment-definition
anomalies (STRA briefly reported USHE differently during the COVID
quarter). Three quarters in early 2020 are absent.
-
Persistence series has Q3 gaps by design. Trailing-4Q
persistence is reported with a 1-quarter lag in 10-Q narrative
(Q1 10-Q discusses Q4-prior persistence, Q3 10-Q discusses Q2 not
Q3). Series therefore has gaps every Q3. We treat this as the
natural reporting cadence rather than try to fabricate Q3 values.