SAFT vs SPNT

Safety Insurance Group, Inc. vs SiriusPoint Ltd. — Valuation Comparison 2026

SAFT

Fire, Marine & Casualty Insurance
Safety Insurance Group, Inc.
Quality
8.1
out of 10
Value Trap
6
SAFE
Price
$70.17
Last close
Models
12/13
Active
VS

SPNT

Fire, Marine & Casualty Insurance
SiriusPoint Ltd.
Quality
8.6
out of 10
Value Trap
18
SAFE
Price
$21.35
Last close
Models
12/13
Active

Model-by-Model Comparison

ModelType SAFT Fair ValueSAFT Upside SPNT Fair ValueSPNT Upside
Bayesian DCF Intrinsic $86.11 +22.7% $13.79 -35.4%
Earnings Power Value Intrinsic $30.37 -56.7% $26.33 +23.3%
EROIC Spread Intrinsic $•••.•• ••.•% $•••.•• ••.•%
First Chicago Scenario $•••.•• ••.•% $•••.•• ••.•%
Markov DDM Intrinsic $•••.•• ••.•% $•••.•• ••.•%
ML-RIV Intrinsic $•••.•• ••.•% $•••.•• ••.•%
PWERM Option-Based $•••.•• ••.•% $•••.•• ••.•%
Regime Cross-Sectional Relative $•••.•• ••.•% $•••.•• ••.•%
Sentiment SOTP Hybrid $•••.•• ••.•% $•••.•• ••.•%
CUCE Ensemble Ensemble $•••.•• ••.•% $•••.•• ••.•%
FTNN Topology Relative $•••.•• ••.•% $•••.•• ••.•%
RCMH-DCF Intrinsic $•••.•• ••.•% $•••.•• ••.•%
🔒

Unlock Full 13-Model Comparison

Access all valuation models for SAFT vs SPNT — including EROIC Spread, First Chicago, Markov DDM, PWERM, and 7 more.

Access Full Analysis — From $27/mo →

SAFT vs SPNT — Which Stock Is More Undervalued?

SPNT scores higher with a 8.6/10 quality rating vs SAFT's 8.1/10. Both stocks are analyzed daily using SEC EDGAR filings across 13 independent models.

Comparing Safety Insurance Group, Inc. (SAFT) and SiriusPoint Ltd. (SPNT) across 13 institutional-grade valuation models reveals how each company's intrinsic value stacks up against its market price. CirclFi's engine processes SEC EDGAR 10-K and 10-Q filings, FRED macroeconomic data, and GDELT news sentiment to generate independent fair value estimates daily.

SAFT currently trades at $70.17 with a QOC of 8.1/10, while SPNT trades at $21.35 with a QOC of 8.6/10.

Both companies are analyzed with models spanning intrinsic (Bayesian DCF, EPV), scenario-based (First Chicago), regime-switching (Markov DDM, RCMH-DCF), machine learning (ML-RIV, FTNN Topology), and ensemble methods (CUCE).