$AGNC · AGNC Investment Corp

Outside coverage
EPS 0–1
DIR 0–0
MAE +1.0%

Latest call · 2026-04-19

✓ Scored · earnings 2026-04-20 AMC

The call

EPS
$0.32
MISS· -11.1% vs street
Direction
🔴 DOWN
1d -2.5% · 3d -4.0%
Confidence
MEDIUM
Positioning: hype_neutral
Spot at call
$10.90
as of 2026-04-19

Head-to-head · Claude vs the Street vs reality

Claude Street Actual
EPS $0.32 $0.36 $0.42
Revenue
Direction (1d) 🔴 DOWN ⚪ — 🔴 DOWN
1d move -2.5% -1.5%
3d move -4.0%
EPS error $0.100 $0.060 Street closer
EPS Street closer C: $0.100 · S: $0.060 Direction

Thesis

Q1 had wider MBS spreads + rate vol — TBV took a hit. Stock is pricing a Fed-cut rescue that has not come. NLY prints Tuesday; bad AGNC TBV pre-positions the sector.

What would flip it

Dovish Fed signal pre-earnings tightens spreads, flips TBV narrative.

💡 Book-value miss > EPS miss. Short AGNC, watch NLY read-through.

The market's narrative

Rate cuts coming, spreads normalizing, dividend safe at current level.

Where the Street may be wrong

  • Q1 MBS spread widening + rate vol pressured book value meaningfully — likely -3-5% TBV decline
  • Pay-downs slowing with mortgage rates elevated = earnings-available-for-distribution under pressure

Peer read: Annaly (NLY) reports Tuesday — if AGNC misses on TBV, NLY pre-positioning suggests sector weakness

Reasoning

  • Consensus EPS already down 18% YoY ($0.36 vs $0.44) — further miss would confirm deterioration
  • Stock at 30-day high $10.90 on rectangle breakout — priced for good news that likely won't come
  • TBV declines are the real story for mREITs; MBS spread widening in Q1 = headwind
  • Dividend held at $0.12/mo provides floor but not a catalyst
  • Above EMA 50/200 at $10.41 — chart has modest support if print is in-line

Risks to the call

  • If Fed signal on June cuts accelerates pre-earnings, MBS spreads tighten = better TBV
  • Management commentary on Q2 book value recovery could neutralize the Q1 miss

Lesson from the post-mortem

Mixed outcome. Direction call DOWN was correct (stock did pull back after the print on TBV erosion) but EPS estimate $0.32 vs actual $0.42 was worse than Street's $0.36 — we modeled too aggressive a spread-widening drag on earnings-available-for-distribution. TBV thesis (–5.6% to $8.38) was exactly right in magnitude and was the real market-moving datapoint, even though EPS beat headline. Lesson: for mREITs, keep two books — headline EPS track and TBV track — because the market weights TBV heavier on rate-vol quarters, and our edge is better on TBV than on dividend-coverage EPS.