Tl;dr. During COVID we built positions in mortgage insurers when they traded at significant discounts to book value. As prices recovered and crossed back above book, we sold. That round-trip — buy when discounted, sell when fair — is the discipline. The sector itself isn’t the point; the framework is.
What mortgage insurers actually do
Mortgage insurance companies (MIs) protect lenders from losses on home loans where the borrower puts down less than 20%. The borrower pays the premium; the lender gets the protection. If the borrower defaults, the insurer covers a piece of the lender’s loss.
It’s a useful business in normal times. Borrowers with smaller down payments can still buy houses. Lenders take less risk per loan. Insurers earn a stream of premium income for taking the tail risk on a diversified pool of mortgages.
It’s a cyclical business. Housing market booms generate strong premium volume and minimal defaults. Housing market downturns mean both lower new-business volume and higher losses on the back book.
What makes MIs interesting from a value perspective is that the book value of the business is largely statutory capital — regulatory reserves they hold against future losses, calibrated against their underwriting standards. That book value is reasonably defensible. The market price, on the other hand, can swing dramatically based on housing sentiment.
That gap is where the opportunity lives.
The COVID setup
In March 2020, mortgage insurers’ stock prices got crushed. The fear was straightforward: a global pandemic would cause widespread job losses, borrowers would default, MI losses would explode, and the equity might be wiped out.
We looked at it differently. Three things mattered:
1. Capital position. The MIs had spent the years since 2008 rebuilding regulatory capital under tighter standards. Their balance sheets going into COVID were materially stronger than going into the 2008 housing crash. They had the cushion to absorb meaningful losses without diluting equity.
2. Underwriting quality. Post-2008, MIs had been underwriting much higher-quality loans — borrowers with better credit scores, lower DTI ratios, more documentation. The book of business at risk was qualitatively different than the 2007-2008 vintage that nearly killed the industry.
3. Forbearance bought time. Federal mortgage-forbearance programs absorbed the immediate shock of unemployment. Borrowers weren’t being forced into default at the speed the market price implied. That bought time for housing prices and employment to stabilize.
At deep discounts to book value — some MIs were trading at 0.6–0.7x book — the math worked. Even if losses ran meaningfully above normal, the stocks were priced as if the equity was at risk of being wiped out, which the capital and underwriting facts didn’t support.
We built positions.
The underlying earning power made the asymmetry even sharper. Many of these MIs were generating normalized return on equity above 10%. Buying a 10%+ ROE business at 0.6–0.7x book is effectively buying an earnings yield of roughly 15–17% on your cost basis — a spread over almost any other risk-adjusted opportunity in financials at the time. The capital was defensible, the underwriting was clean, and the earnings power was real. The discount was the dislocation.
The names where the math compounded most cleanly: MGIC Investment (MTG), NMI Holdings (NMIH), and Radian Group (RDN). Our contemporaneous research on each is linked above — the analytical framework is the same one driving the broader piece, with the company-specific math worked out.
What actually happened
The pandemic-default tsunami didn’t materialize. Housing prices, far from collapsing, accelerated meaningfully through 2020–2022 as low interest rates and household formation drove demand. Forbearance programs gave delinquent borrowers room to recover or sell at a profit. Actual MI losses came in well below early-pandemic-scenario projections.
The market noticed. The stocks re-rated from deep discount toward book value, then above. The capital cushion that had been hidden inside a low multiple became visible. Premium income kept compounding through the recovery as origination volumes stayed elevated.
That’s the “buy discount” half of the trade working.
The harder half: selling as prices crossed book
The discipline that’s harder to talk about is selling. As MI stock prices rose past book value and into modest premiums to book, we sold. Not all at once — in tranches as the margin compressed.
This is the half most investors get wrong. A position that has worked is psychologically easier to hold than a position that hasn’t. The story is intact, the business is doing well, the dividend is good. Why sell?
Because the margin of safety we bought is gone. At 0.6x book, we owned a defensible business at a discount with a wide margin. At 1.0x book, we owned a defensible business at fair value with no margin. At 1.2x book, we owned a defensible business at a premium with the margin going against us.
Selling when the margin compresses is the same discipline as buying when it widens. Both halves matter. The half that’s harder is the second one.
That’s not a market call. That’s a math call. The cash that comes out of an above-book MI position can be redeployed into the next discount we find. The position that stays in place at above-book has no margin and is fully exposed to the cyclical downturn whenever it comes.
What this framework is good for — and what it isn’t
The “buy below book, sell above book” framing works for businesses where book value is defensible. That tends to be regulated financial companies: insurers, certain banks, some REITs where book is grounded in real estate that has a separate valuation framework.
It works less well — or doesn’t work at all — for:
- Asset-light businesses where book value isn’t meaningful (most software companies, services, brands)
- Companies with goodwill-heavy book value from past acquisitions (book overstates real economic capital)
- Businesses with structural quality decay (where book may be intact but the earning power isn’t)
For the right kind of business, though, the framework is durable. It’s been running since Graham and through every subsequent generation of value investors who paid attention to balance sheets first.
The honest limit
The buy side of this trade requires patience. Stocks at 0.6x book can go to 0.4x book before they re-rate. We owned some of these positions through interim drawdowns of 20–30% from our cost basis before the math reasserted. If you can’t hold through that, the discipline doesn’t work — and the original margin was just buying yourself the right to panic.
The sell side requires equal patience in the other direction. When you sell a winner at fair value, the stock usually keeps going for a while. We’ve left meaningful upside on the table by selling above book on names that subsequently traded to 1.5x or 1.8x book. That’s the cost of the discipline. We don’t try to catch the top — we try to get paid for our margin of safety while we have it, and then move on to the next one.
What we won’t do is hold a position at a premium to defensible value just because the story is intact. The story being intact is what justified the buy. The math being intact is what justifies the hold. When those decouple, we trim.
How this shows up in client portfolios
For clients who were with us in 2020–2024, you owned mortgage insurers as a meaningful position. You experienced the discount-to-book buying, the recovery, and the sales above book. The cash from those sales went into other positions where we saw similar margin.
You don’t currently own mortgage insurers in size. The thesis worked, the margin closed, the cash got redeployed. If MI stocks drop materially below book again — which they will eventually, because they’re cyclical — we’ll likely revisit. The framework doesn’t change. Only the price does.
That round-trip is what the discipline looks like in practice. Buy when the math works. Sell when it doesn’t. Sit on cash when neither side has an edge. Repeat.
Disclosure
T&T Capital Management LLC (TTCM) is an SEC-registered investment adviser. This is research and commentary, not personalized investment advice. Forms ADV Part 2A and CRS are available on request or via the SEC IAPD website. CRD #158407.
This piece discusses general investment concepts and TTCM’s historical positioning in the mortgage-insurance sector. TTCM and its principals may hold positions in securities discussed within the broader theme. Position-level holdings as of the most recent quarter-end are disclosed in TTCM’s Form ADV.
Past performance and historical positioning are not indicative of future results. Specific entry and exit prices and multiples referenced are illustrative of approach rather than precise records.
Want to see what working with us looks like in practice? Plain-language walkthrough — discovery call, ADV review, IPS, ACAT transfer, ongoing reviews. See our process →
