Consistent
Beat Magnitude as Systematic Guide Discount Factor
When a company beats its own guidance by a consistent magnitude
across 4+ consecutive quarters, that pattern should be treated as a
systematic adjustment factor for forward estimates — not dismissed as a
one-off. The guide becomes a floor, not a central estimate. Analysts who
model to the guide without applying a beat-factor underestimate the
business. This extends beyond revenue: companies that sandbag revenue
also tend to sandbag margin guidance, meaning EBITDA beats are often
correlated with and additive to revenue beats.
Variant: Widening beat magnitude — when the absolute
dollar beat is not just stable but expanding each quarter, this signals
that demand is outrunning even management's own internal models. This is
a stronger signal than stable sandbagging: the company itself can't
fully see what's coming. Treat widening magnitude as a
demand-acceleration indicator, not merely a conservative-guidance
indicator.
Variant: Compressing beat magnitude — when a company
with a documented multi-quarter beat pattern shows sequential
compression in beat % (e.g., 21.6% → 18.6% → 12.3% → 8.8% → 6.8% over 5
quarters), this is an early warning that the "guide is a floor" thesis
is normalizing. The guide discount factor should be reduced
proportionally. A useful invalidation threshold: if beat magnitude drops
below 5% for two consecutive quarters, treat the "conservative guidance"
assumption as broken and model to the guide rather than applying a beat
adjustment. Compressing magnitude does not immediately invalidate the
thesis — management may still be conservative — but it reduces the
asymmetric upside embedded in forward estimates.
Evidence
- APP Q4 FY25: Beat pattern of 4–8% over four consecutive quarters
(Q1–Q4 FY25). Q4 actual $1.658B vs guide $1.585B (+4.6%). Q1 FY26 guide
$1.745–1.775B — applying the same 4–6% beat factor implies actual closer
to $1.85B.
- RDDT Q4 FY25: Beat pattern averaging +11.2% over four consecutive
FY25 quarters (+7.5%, +19%, +8.3%, +10%). Q4 actual 726mvsguidemidpoint 660m.
Management also under-guides on EBITDA margin — Q4 EBITDA was 45.1% vs
implied ~37-38% guide. Dual sandbagging (revenue + margin) compounds the
magnitude of the earnings surprise.
- Pattern is distinct from the beat-and-raise masking deceleration
dynamic (existing insight) — here, the beat magnitude is stable
and the raise cadence is intact. The guide is structurally conservative,
not shrinking.
- RBRK Q4 FY26: Beat every quarter since IPO. Revenue beat +10.4%
($377.7M vs 341 − 343Mguide).FCFbeat + 18238M
actual vs $202M guide). FY27 FCF guide $270M expected to print $310-330M
(+15-22%). Beat magnitude has been consistent across all metrics, not
just revenue — dual sandbagging (revenue + FCF) compounds the
signal.
- CRDO Q3 FY26 (stock-analysis): Five consecutive beats with
widening absolute dollar magnitude: $15M → $20M → $33M → $33M →
67M.Dualsandbaggingonrevenue + grossmargin—guided64–66800M
→ >$1.3B). Beat cadence also confirms conservative guidance culture
for a company growing 200% YoY where internal demand visibility is
structurally limited by short-cycle PO dynamics.
- NVDA Q4 FY26 (stock-analysis): Five consecutive beat-and-raise
cycles at 2.5–5.6% magnitude (stable, not widening). Management
explicitly described guidance as "deliberately undemanding."
Named-exclusion variant: Q1 FY27 guide of $78B
explicitly excludes China DC compute revenue by name — this is a harder
form of conservative guidance where a specific excluded stream is
bounded and disclosed. If export restrictions ease, the beat will arrive
against an explicit named baseline, creating asymmetric rather than
diffuse upside optionality. This variant (named exclusion of a
regulatory-constrained revenue stream) is distinct from general
sandbagging: the excluded revenue is identifiable in advance, the upside
is bounded, and the mechanism is public.
- PAY Q4 FY25: Compressing magnitude variant.
Seven-quarter beat history: +9.4%, +21.6%, +18.6%, +12.3%, +8.8%,
+11.0%, +6.8%. The last beat (6.8%) was the narrowest in the series;
revenue landed at the low end of PAY's own guide range for the first
time. Three of the last four quarters came in below 12%. Beat magnitude
trending down while underlying fundamentals (adj EBITDA margin 37.3%
record, FCF $125M, +362% YoY) remained strong — confirms the compression
is guidance calibration normalizing, not business deterioration. Q1 FY26
(~May 2026) is the key datapoint: beat >10% = conservative culture
intact; beat <5% = thesis shifts to modeling the guide.
- TTD Q4 FY25: Compressing magnitude variant — below
invalidation threshold. Four-quarter beat history: +7.1% (Q1),
+1.8% (Q2), +3.1% (Q3), +0.8% (Q4). The final beat of +0.8% is the
narrowest in TTD's public history. Two of the last four quarters (Q2 and
Q4) came in below the 5% invalidation threshold; the Q3 recovery to 3.1%
created false hope before the final collapse. Non-monotonic compression
(brief recovery in Q3) is a known variant — the correct signal is the
overall trend and the sub-5% readings, not any single quarter.
Notably, TTD's Q1 FY26 guide of 678M(+10630M guide to beat to
$700M+. Q1 FY26 actuals (May 8) are the thesis test: if beat reverts
above 5%, conservative culture may be intact; if the guide prints
near-flat again, model to guide with no adjustment.
Implication
When reviewing guidance for companies with an established 4+ quarter
beat cadence, distinguish between: (a) stable magnitude —
conservative guidance culture, apply mean beat % as adjustment; (b)
widening magnitude — demand acceleration signal, not just
sandbagging; apply higher upside multiplier and increase position
conviction; (c) compressing magnitude — guidance conservatism
normalizing; reduce the beat adjustment factor, watch for the
inflection. Calculate: (a) mean and range of historical beat %; (b) the
trend of that magnitude over the last 3-4 quarters; (c) for widening
cases, model the upper end of the historical range; (d) for compressing
cases, model the lower end of the historical range and set a
5%-for-two-quarters invalidation threshold. For companies that also
sandbag margins, model the EBITDA/GM beat separately. A predictable
beater with widening beats deserves higher conviction on forward
estimates than a merely consistent one; a compressing beater deserves
reduced certainty on the beat-adjusted forward estimate.
PROMOTION CANDIDATE: This pattern has been confirmed 5 times (APP,
RDDT, RBRK, CRDO, NVDA). Consider updating
bert/references/philosophy.md.