Assured Guaranty AGO
Assured Guaranty is a melting-ice-cube bond insurer aggressively returning capital it hasn't earned; Puerto Rico losses alone (2-4x reserves) will force a capital raise or strip the AA rating it needs to write new business.
Thesis
David Einhorn argues Assured Guaranty (AGO), one of the last surviving municipal bond insurers, is a 'melting ice cube' aggressively returning capital it hasn't earned. The post-2008 muni rating recalibration killed the rating-arbitrage business, and amortization of AGO's existing book outpaces new and acquired business by roughly $40-65M of net par each year. Levered 32-54x against $7B of statutory capital, AGO's $239B insured book includes $12.2B of below-investment-grade exposure dominated by Puerto Rico ($5B par, $8B with interest). Greenlight calculates Puerto Rico losses of $2.7-4.5B versus AGO's $1.3B total expected-loss reserve, claiming AGO masks the gap by using internal credit ratings systematically higher than Moody's/S&P. Recognizing realistic losses would breach S&P's $1.8B remaining loss-absorption capacity, force a capital raise, or strip the AA rating AGO needs to write new business.
SCQA
Assured Guaranty is one of the few bond insurers to survive the 2008 crisis, guaranteeing $239 billion of municipal and structured-finance debt against just $7 billion of statutory capital — leverage of 32 to 54 times.
The post-crisis muni rating recalibration ended AGO's rating-arbitrage business, so the book amortizes faster than new writings. Worse, $12 billion of insured exposure — anchored by $5 billion of defaulted Puerto Rico debt — is below investment grade and severely under-reserved.
Greenlight is short AGO and urges regulators and rating agencies to scrutinize the company's internal-rating-driven reserves and curtail aggressive dividends and buybacks before AGO's loss-absorption capacity is depleted.
Three independent analyses peg Puerto Rico losses at $2.7-4.5 billion (midpoint $3.6B) versus AGO's $1.3B total reserves, implying a forced capital raise, downgrade from AA, or runoff that would devastate the equity.
The three reasons
- 1
AGO's muni-wrap business is structurally dead post-rating-recalibration; book amortizes faster than new business
- 2
Puerto Rico exposure of $5B par ($8B with interest) implies $2.7-4.5B losses vs only $1.3B total reserves
- 3
Opaque internal credit ratings consistently higher than Moody's/S&P enable systemic under-reserving
Primary demands
- Regulators should scrutinize AGO's loss reserves, which appear to lean on internal credit ratings systematically more generous than third-party agencies
- Regulators should carefully study whether to approve additional dividends and buybacks given depleting loss-absorption capacity
- Investors should short AGO equity (and optionally hedge by going long Puerto Rico bonds at ~80c)
KPIs cited
Pattern membership
Where this document fits across the library's 12 rhetorical / structural patterns.
Precedents cited
- Allied Capital (Einhorn's 2002 Sohn short — explicit 'history repeats' framing)
- MBIA and AMBAC (post-2008 bond-insurer collapses into runoff after ratings downgrades)
Notable slides (6)
Notes
Classic Einhorn Sohn presentation — stage-set with explicit 'history repeats' callback to his 2002 Allied Capital short (slides 3-4) and recurring 'melting ice cube' visual metaphor (slides 5, 21, 32). Speaker-notes-below-slide format suggests this is the post-presentation distribution version (transcript + slides). The CEO-quote-contradiction at the close is unusually pointed: Frederico's 2001 quote about the 'fun you can engineer into an insurance company's balance sheet' is used as the punch line. AGO management issued a same-night press release calling Greenlight's analysis a 'fundamental lack of understanding'; slide 1 of this version preemptively rebuts that claim. No discrete sum-of-parts but the deck does triangulate three independent loss methodologies (Greenlight, market-implied, Moody's-implied) which together yield a converging $2.7-4.5B PR loss estimate. Stake size not disclosed.