Open in interactive viewer → charts, metric popovers & call review

Two Banks and a Pharmacy: Fairfax Spends Nearly $8 Billion While Its Bond Book Blinks

Book value rose 8.4% and underwriting stayed clean — but the interest-income engine stalled, the premium growth wasn't what it looked like, and the October shopping list dwarfs the holdco chequebook.
FFH.TO · Earnings Call · 2026-11-04

Two banks and a pharmacy

The quarter itself was a decent Fairfax quarter: $930 million of net earnings, $3 billion for the first nine months, book value per share of $1,349.15 against $1,260.19 at year-end — up 8.4% adjusted for the $15 dividend. Underwriting income of $524 million on a 92.5% combined ratio. Nothing to sneeze at. But the thing that actually changed at this company in the last six weeks wasn't in the income statement. It was in the commitment book. In October, Fairfax was selected by the government of India as strategic buyer of a 60.7% stake in IDBI Bank for roughly $5.6 billion. Then came the Boots Group partnership with Wittington — an enterprise value of about $8.9 billion including assumed debt, with Fairfax committing up to approximately $2.3 billion to own 50% of the equity. Both will be equity-accounted. That is nearly $8 billion of new, illiquid, equity-method exposure announced in a single month, against $1.9 billion of cash and investments at the holding company, a $2 billion undrawn revolver and about $2.4 billion of market-traded investments and associates. The Eurolife life operations sale — $906 million, roughly $200 million of pre-tax gain landing in Q4 — helps at the margin, but it doesn't change the arithmetic. And on the single largest commitment, management barely engaged. Asked about union opposition to IDBI, Peter Clarke deferred entirely: “as the transaction is not yet closed and we are still waiting for regulatory approval, there is really not much more that we can say” — Peter Clarke, Host · 2026-11-04. Fair enough legally. But when the biggest cheque in company history is being written, a referral to a press release is itself information.

The rate paradox, and a stalled engine

The reported investment line was ugly: net losses on investments of $859 million versus a $426 million gain a year earlier. “In the quarter, our net losses on investments were $859 million, driven by mark-to-market unrealized losses on our bond portfolio, principally due to increasing interest rates.” — Peter Clarke, Host · 2026-11-04 Mechanically: $828 million of bond markdowns, $679 million of offsetting IFRS 17 discount uplift, netting to $148 million. Also worth flagging — a $258 million unrealized loss on the Digit Insurance preferred shares, a reversal on what had been one of Fairfax's best positions, partly offset by $177 million of equity gains at Metlen, Eldorado, IIFL Finance and Equinox. Higher rates cutting book value while improving reinvestment is the oldest Fairfax story there is, and the macro backdrop remains squarely in the "impact of high rates" camp globally. The problem is the reinvestment half of that trade looked wobbly this quarter. Interest and dividend income came in at $669 million against $655 million — up 2%. Three months ago the same line stepped up roughly 11% and management described a run rate near 5%. Now: “I think our interest income increased more than the absolute number, with dividends down. The third quarter is usually light on dividends.” — Peter Clarke, Host · 2026-11-04 Perhaps. But the $2.7 billion annual run rate Wade Burton invoked last quarter implies a higher quarterly print than this. The portfolio is bigger ($78.8 billion total, $53.1 billion fixed income), still short duration at just over two years, still over 75% governments, still holding $8.3 billion in cash and short-term bonds. That's a wonderfully defensive posture — and, this quarter, a flat-earning one. The offset came from associates: $378 million versus $305 million, led by Eurobank's $158 million, $53 million from the Waterous Energy Fund, and $35 million each from Poseidon and EXCO. Nine-month associate profit of $792 million annualises to roughly $1.06 billion — quietly meeting the long-standing $1 billion marker that analysts had been probing all year.

Growth that isn't, and the soft market's fourth autumn

Gross premiums of $8.9 billion, up 9.3%, look like an acceleration. They aren't. International operations were up 27.1% gross but only 13.2% net, and the gap has a name: “Driving the increase was Gulf Insurance, primarily due to the significant rate increases on marine business as a result of the conflict with Iran. The majority of this business was not retained by the company, so it only affected gross premiums and not net.” — Peter Clarke, Host · 2026-11-04 Strip it out and international grew 13%. North America grew 5.1%, Crum & Forster 6.6%, Northbridge was flat in Canadian dollars, Odyssey fell 1%. Jaeme Gloyn asked directly whether something had shifted in the market. The answer is worth dwelling on:

Really not. In the third quarter, we continue to see similar trends that we have reported the last number of quarters, softening across many of our companies, especially in North America, and particularly in property and our reinsurance business, and property cat as well.

Peter Clarke, Host · 2026-11-04
That is the fourth consecutive call saying the same thing — near-identical language appeared in July: “we continued to see similar trends that we have reported in the last number of years, softening across many of our companies, especially in North America.” — Peter Clarke, Host/Moderator · 2026-07-31 The softening market is no longer news; it is the weather. And the tell is in the company's own vocabulary: hard market sat in the top ten of Fairfax's keyword profile in 2024; it has now fully inverted. Peer corroboration is easy to find — IGIC used the same "soft market" framing this week, and reinsurers with Ukraine conflict exposure are telling a similar story. The underwriting still holds: favourable development of $156 million, 2.2 points of combined ratio benefit, and Northbridge absorbing over 10 points of Canadian catastrophe business losses in an above-average season. The fourth-quarter actuarial review is the real test.

The capital-return promise, again

Bart Dziarski asked when excess capital gets upstreamed. Clarke's answer was constructive — and familiar: “That has slowed down now, and so a lot of our companies are producing excess capital with premiums slowing down. That provides more access to dividends for us. More to come in the fourth quarter.” — Peter Clarke, Host · 2026-11-04 Except he said a version of this in February: “with the premiums flattening off, it does produce -- it can produce excess capital at our insurance operations that would produce more dividends up to the holding company.” — Peter Clarke, Host, likely Investor Relations or similar · 2026-02-20 That was nine months ago. Excess capital has been a Fairfax talking point since at least 2024, and it keeps being promised for "the second half" or "year-end." Meanwhile leverage is drifting up — total debt to total capital excluding non-insurance companies at 27.3%, from 26.2% at December 31 — and the company repurchased 1,086,435 shares for $1.8 billion at an average of $1,630.87, well above today's $1,349 book value. The genuine comfort is the $5.1 billion ($256 per share pre-tax) excess of fair value over carrying value in non-insurance associates and market-traded subsidiaries, up from $151 per share at year-end, with $3.6 billion of it Eurobank and $0.8 billion Poseidon. It just never lands in book value — and the phrase "intrinsic value," which dominated Fairfax Q&A for a year, has quietly slipped out of this quarter's keyword profile entirely. That is the shape of the story now: a company that has stopped being a hard-market compounder and become a patient, cheaply-funded balance sheet converting a soft market's spare capital into two enormous, illiquid positions — at precisely the moment its bond book is marking down and its interest income has stopped compounding. Watch Q4 for two things: whether the upstreamed dividends finally arrive, and whether the funding of Boots and IDBI comes from cash or from more paper.