Jefferies Financial Group Inc. JEF

The business

Jefferies is the largest independent full-service investment bank, the biggest one not attached to a giant commercial bank. Two engines drive it. One advises companies on selling themselves and raises capital for them by underwriting their stock and bonds. The other buys and sells securities for clients across bonds, stocks, and currencies. A smaller asset-management arm and the leftover investments from its old Leucadia parent round out the rest. The firm collects a fee each time a deal closes or a block of securities changes hands, which makes the revenue lumpy and tied to how busy Wall Street is. The moat is reputation and a two-decade roster of senior bankers, because the client hires the banker, and Jefferies has spent twenty years recruiting them from the shrinking bulge-bracket firms. When a mid-size company decides to sell itself, a Jefferies banker is often in the room precisely because the firm is large enough to underwrite the deal and small enough to still care about it.

The numbers

The sequence shows a business climbing off a trough, with the top line running faster than the bottom line. Read the quarters:

Revenue in the spring quarter jumped 37% against the year before, but operating income rose only 14%, so costs, mostly banker compensation, ate most of the top-line gain. Reported earnings landed under the analyst consensus in both of the last two quarters. That is the tell of this business: a good market lifts revenue faster than it lifts profit, because the people who generate the revenue get paid more when it comes in.

Here is the compounding math, and it is not flattering on its face. Revenue grew 15% a year over three years, but diluted earnings per share went from $3.06 in 2022 to $2.83 in 2025, a slight decline, because 2022 was a peak and 2023 an air-pocket. The one clean win is the share count, down 13% over that stretch as the firm bought back stock. Book value sits at $49.27 a share against a $55 price, so you pay 1.1 times book for a business earning an 8.5% return on that equity. That multiple is the market saying Jefferies does not clear its cost of capital through a full cycle.

What this memo believes that the price does not: the first-half pace, $431 million of net income, is the start of a walk back toward the low-teens returns Jefferies earned in 2021, not a cyclical head-fake, and 1.1 times book hands that recovery over for free. The single print that settles it is the return on equity in the November full-year report.

Management

Richard Handler has run this firm for two decades, and the record reads better in the buyback than in the pay stub. Management repurchased $860 million of stock at an average price of $32 in 2022 and another $169 million near $35 in 2023, then cut buybacks to a trickle as the price climbed past $50, which is the pattern you want, heavy when cheap and quiet when dear. Executive compensation runs high in absolute dollars, as it does across Wall Street, but the disclosed figures for named officers place the cost in line with the seat. The loudest line in the record is not insider activity in the usual sense: Sumitomo Mitsui bought $1.3 billion of stock in the open market over the past year, close to 12% of the whole company's market value, every dollar of it discretionary rather than pre-scheduled. A strategic partner spending its own cash at these prices says something no proxy statement can.

How it fails or surprises you

The deal window slams shut. The business is a bet on activity, and activity is a bet on rates and confidence. In 2023 advisory and underwriting dried up and earnings fell from $6.13 to $1.10 in two years. This is the fact this memo's bullish read explains least well, that revenue can grow across a decade while per-share earnings go nowhere. The print that reveals it first is two consecutive quarters of falling advisory and underwriting revenue.

The old marks bite. The merchant-banking leftovers and asset-management performance fees are lumpy and mark-sensitive, and a bad quarter there can swamp a good banking quarter without warning. The tell is a large non-operating write-down landing in a single period against otherwise clean results.

Sumitomo goes further (right tail). The partner already owns a growing double-digit slice bought at today's prices. A larger stake, a deeper balance-sheet alliance, or an outright bid is not priced into a stock at 1.1 times book. The print is a disclosed stake increase past a new threshold or a formally expanded alliance, either of which reprices the whole thing.

Closing thoughts

The shape of the payoff is an uncertainty a named print resolves, not an exposure you simply survive. Downside is cushioned by tangible book at $41 a share, roughly a quarter below the current price, the level the stock would drift toward if the deal cycle turned cold. Upside is a return on equity that walks into the low teens and drags the book multiple up with it, plus the free option on the partner doing more. The fatter tail here is the upside, because you are buying a recovering franchise near book value rather than a peak one at a premium, but the left tail is real and named, and the November full-year report is the referendum the desk has been waiting on. That print is still ahead of us, and nothing in the weeks since has settled which way banking, the asset-management drag, or the Sumitomo stake breaks.

The bet is still that companies keep hiring Jefferies to advise on mergers and underwrite securities offerings, and that the firm keeps more of what it earns than it pays out to earn it. What breaks it is a dealmaking freeze that takes return on equity back toward the mid-single digits of 2023. The one pair of numbers that tells you first: advisory-plus-underwriting revenue and the compensation ratio, moving the wrong way together for two quarters running.