Dynex Capital's Q3 book value declines, reports loss

Dynex Capital

19/10/2026 13:13

Overview
U.S. mortgage REIT reported a sequential drop in Q3 book value per share and a net loss
Book value fell as wider agency MBS spreads pressured asset valuations
Company raised $384 mln capital and expanded agency MBS portfolio by $3.2 bln

Outlook
Company says current Agency mortgage market offers attractive forward return opportunities
Company sees stable financing and strong technical conditions in the mortgage market

Result Drivers
MBS SPREAD WIDENING - Co said wider agency MBS spreads drove decline in book value and asset valuations, partially offset by gains on interest rate hedging portfolio
PORTFOLIO GROWTH - Higher interest income attributed to increased capital deployment into agency MBS and 10% growth in average interest-earning assets
EXPENSE REDUCTION - Operating expenses fell due to lower stock-based compensation

Company press release: ID:nBw9ZVwvMa

Key Details
Metric
Beat/Miss
Actual
Consensus Estimate
Q3 Loss Per Share
 
$1.34
 
Q3 Net Interest Income
 
$101.56 mln
 

Analyst Coverage
The current average analyst rating on the shares is "buy" and the breakdown of recommendations is 4 "strong buy" or "buy", 2 "hold" and no "sell" or "strong sell"
The average consensus recommendation for the specialized REITs peer group is "buy"
Wall Street's median 12-month price target for Dynex Capital, Inc. is $14.50, about 32.7% above its October 16 closing price of $10.93
The stock recently traded at 7 times the next 12-month earnings vs. a P/E of 8 three months ago

Reuters Recommended Reads
Oct 16 - Ames National Q3 profit jumps as wider interest margin lifts income
Oct 16 - US regional banks join bigger peers in signaling resilient economy as lending holds firm
Oct 18 - UK housing market showing signs of mortgage rates impact, Rightmove says 

For questions concerning the data in this report, contact Estimates.Support@lseg.com. For any other questions or feedback, contact reuters.support@thomsonreuters.com.

(This story was created using Reuters automation and AI based on LSEG and company data. It was checked and edited by a Reuters journalist prior to publication.)