Buried near the back of Jefferies Financial Group's latest quarterly report is a sentence that sizes the problem the market has spent most of 2026 pricing. Point Bonita Capital, the trade-finance arm inside Jefferies' asset-management business, ran roughly $3 billion of receivables assets supported by $1.9 billion of invested equity — "of which $113 million, or 5.9%, is owned by LAM," the firm's own Leucadia Asset Management unit . That $113 million is the disclosed extent of Jefferies' equity in the vehicle whose two successive alleged frauds have become the stock's defining story.
Jefferies is a wholesale investment bank: advisory, equity and debt underwriting, and sales and trading in equities and fixed income, all reported as one segment, Investment Banking and Capital Markets, which produced $5,935.8M of the firm's $6,445.5M of nine-month net revenues . The second and only other reportable segment is Asset Management at $493.6M, with a small residual line making up the difference; it holds fund stakes, revenue-share interests in affiliated managers, and the remains of a merchant bank — Tessellis, Stratos, HomeFed, M Science — that the firm has spent years unwinding. Geographically it is still an American house, with the Americas at roughly three-quarters of fiscal 2025 revenue. One of those segments has demonstrably re-rated its own economics over three years. The other has produced two consecutive years of losses nobody was modeling. How much of the second belongs in the multiple applied to the first is the question that decides this stock, and the market's current answer is too severe.
The de-rating runs through a trade-finance fund, not a missed quarter
At $46.12 on October 30 — a session in which the stock fell 6.16% against a 3.58% decline in Goldman Sachs and a 1.8% drop in the S&P 500 — Jefferies sits about 31% below its $66.60 52-week high . There is no guide-down behind that. Jefferies does not guide, and the first nine months of this fiscal year set records in advisory, capital markets and the combined banking segment .
What happened instead is a sequence of credit events inside one fund, compounded by the revenue hole they left. Point Bonita had purchased receivables from First Brands Group since 2019 and been paid on time and in full for almost six years when, in September 2025, First Brands stopped remitting and filed for Chapter 11 amid what the bankruptcy proceedings have since alleged was a massive fraud; the portfolio held roughly $715 million of purported receivables from retailers including Walmart and AutoZone, and Jefferies' exposure to First Brands is now valued at zero . Then the same portfolio produced a second case. Receivables bought from Radiant World and Sapphire Minmetals began defaulting in January 2026, and Jefferies' own filing states that Point Bonita "has subsequently obtained substantial prima facie evidence that Radiant and Sapphire were engaged in a sophisticated and systematic fraud involving extensive falsification of documents," with English High Court proceedings underway and a worldwide freezing order obtained . A lawyer for the fund has told the court that falsified documents cover receivables with a face value of nearly $900 million across more than one lender's financings ; both companies deny wrongdoing. Management's framing at its October 19 investor meeting was that the fund's own outstanding exposure "approaches $500 million," that a relatively modest portion is for Jefferies' account, and that nothing has been reserved because the firm expects to collect .
The no-reserve stance is the weakest part of the disclosure: a firm that has obtained prima facie evidence of systematic fraud and still carries the receivables at full value is asking for trust it has twice failed to earn. But the $500 million headline the market has anchored on is the fund's exposure, not Jefferies'. The firm's own equity in the entire portfolio, First Brands included, is the disclosed $113 million — roughly $0.36 a share after tax if written off in full, against a share price down more than twenty dollars.
What is not bounded is the litigation. Western Alliance Bancorporation sued Jefferies, Leucadia Asset Management and affiliates in New York Supreme Court in March, alleging breach of contract and fraud over a $126.4 million loan balance collateralized by First Brands receivables, and the bank charged that balance off in full . Jefferies' public response was that the loan was non-recourse beyond the special-purpose vehicle's assets, with no guarantee from the parent — a defense that is contested and unadjudicated, and one a fraud claim is designed to get around. A single third-party claim therefore already exceeds the fund equity at stake. Treat that honestly: the most likely outcome is a multiple problem — litigation costs, slower fundraising, a permanent governance discount — but the tail in which recourse attaches and the number runs well past $113 million is real, and it belongs in the bear case rather than being argued away.
The core franchise re-rated its own economics while nobody was looking
Strip the fund noise away and the operating business has done something unusual for a mid-tier broker. Against a 2023 base — the first year of the current segment framework — Investment Banking and Capital Markets grew net revenues 74%, from $4.5 billion to $7.8 billion on a trailing-twelve-month basis, while pre-tax profit rose over 150% from $500 million to about $1.3 billion, lifting the segment's pre-tax margin from 11% to over 16% . Profit has grown at twice the rate of revenue, for an incremental margin of 23% on more than three billion dollars of added revenue — and the firm was hiring throughout.
The August quarter made the point concretely. Investment banking net revenues of $1.33 billion were up 17%, with a record advisory quarter up 25% and equity underwriting up 69%; equities set a record at $626 million, up 29%, on cash and electronic trading and an expanding prime brokerage business . The mechanism behind the margin matters more than the revenue. Management reports an adjusted non-compensation ratio that strips out brokerage and clearing — costs that rise one-for-one with equities volume — and on that basis the ratio has fallen from 30% in 2023 to 24%, with only 15 cents of adjusted non-comp cost attaching to each incremental dollar of revenue . Fixed income is the counterweight. It fell 26% year over year to $176 million , and management frames 2026 and probably 2027 as the period that business needs to rebuild. It is the one core line still going backwards, and the reason the revenue mix is more cyclical than the margin trend alone suggests.
None of this rests on balance-sheet stretch. Tangible gross leverage at the August quarter end was 9.1x on Jefferies' definition against a 16.3x average for its peer group on the same measure, and Level 3 assets — the hard-to-value positions marked off models rather than prices — remain around 3% of financial instruments owned ; cash and other liquidity sources of $19.9 billion equaled 24.5% of total assets . The fear here is about credit judgment in a boutique fund, not about the firm's funding.
The compensation ratio is the lever, and consensus will not underwrite it
The gap between a 16%-plus segment margin and a firm-level pre-tax margin of 13.6% over nine months ($878.8M on $6,445.5M) is partly the legacy portfolio and partly compensation. The comp ratio ran 53.8% over nine months against 52.7% a year earlier, even as non-comp fell to 32.6% from 35.6% . The reason comp went the wrong way is specific and finite: the equity awards granted during the hiring wave amortize on a schedule, and the chief financial officer has described 2026 as the peak point of that amortization, adding that on a like-for-like revenue profile the firm "should be making that turn" .
This matters more than any revenue assumption. On roughly $9.3 billion of net revenues in fiscal 2027, every 100 basis points of comp ratio is about $93 million pre-tax, or close to $0.29 of as-converted earnings per share. My own fiscal 2027 estimate is $4.88 against a $4.68 consensus , on revenue within 1% of the consensus figure — so while tax, share count and non-comp assumptions all contribute, a sub-100-basis-point move in the comp ratio is on its own enough to account for the entire gap. Very little comp leverage is embedded in the published numbers.
The honest objection is that the saving may never reach shareholders. The CFO's turn is explicitly conditional on like-for-like revenue, core headcount is still growing — excluding the Stratos, Tessellis, HomeFed and M Science subsidiaries, employees rose to 6,388 at August 31 from 5,990 at the November 2025 year-end — and management has never committed to a ratio. What tilts it is that the amortization is a disclosed schedule running off against a revenue base that has grown 22% over nine months, while the decision to spend the saving is a choice management has to make affirmatively.
Asset management has to stop subtracting, and Hildene is that bet
The segment that caused the damage also has the clearest path to arithmetic improvement. Asset Management net revenues of $85.6 million in the August quarter were less than half the prior year's $176.9 million as capital was pulled out of legacy strategies, and the nine-month figure fell to $493.6 million from $523.2 million . That decline, not just the fraud headlines, is a real part of why the stock de-rated.
The reconstruction has three parts, all landing in the fiscal first quarter. Tessellis, an Italian telecom holding, is being sold after a goodwill write-down booked this year . The remaining merchant-bank positions are being monetized. And $340 million recycled out of the legacy fund portfolio buys half of Hildene, a structured-credit manager overseeing more than $18 billion, which is separately acquiring the annuity writer SILAC . Management expects that capital to earn a return "more than competitive" with the firm's overall return , while cautioning that meaningful SILAC growth is a couple of years away . Jefferies itself frames the Tessellis sale and the merchant-bank wind-down as the route to "meaningfully higher operating margins" .
The skeptical read deserves stating plainly: a firm whose credibility problem is private credit is answering it by buying more private credit, in a quarter when retail credit vehicles across the industry are capping withdrawals. Hildene substitutes one credit exposure for another, with better fee economics and an insurance balance sheet attached — it is not a de-risking. The test is empirical, and has to be run on underlying revenue, because the first post-close quarter will also carry a one-time markup of the existing Hildene stake. Asset Management net revenues sustainably back above the roughly $165 million quarterly pace of fiscal 2025, excluding that gain, would settle it.
The bear case is that the one-offs are the business
Two years running, Jefferies has booked something it did not expect. Fiscal 2025 ended with the First Brands collapse inside Point Bonita, an exposure now valued at zero ; fiscal 2026 has carried the Tessellis goodwill write-down , a fixed-income mark tied to the UK lender Market Financial Solutions , and the asset-management decline . Items labeled non-recurring that recur annually belong in the multiple, not in an adjustment column, and that is the strongest argument against paying up here. It also cuts at the numerator: a forward estimate built on a clean operating base is worth less if a third consecutive year of charges is the base rate.
Earnings quality in the latest quarter reinforces the caution. Pre-tax earnings of $351.0 million were only 5.8% above the prior year even as net revenues rose 8.5%, and the quarter's effective tax rate came in at 17.9% against 26.9% a year earlier — below the 24.8% quarterly rate shown in the September earnings release , after the 10-Q recognized excess tax benefits on share-based awards . The headline earnings growth was substantially a tax outcome. The firm has also been buying its own stock at prices the market no longer agrees with: nine-month repurchases of 8.3 million shares cost $441 million, an average of $53.25, with the authorization refreshed to $250 million . Over a longer arc the capital return has been real — $6.4 billion since November 2018, equal to 79% of tangible book, with fully diluted shares down 24% — but this year's execution was expensive. And the cycle risk is not hypothetical: earnings at this firm collapsed when the last issuance cycle turned, and flat revenue against a growing headcount is precisely the configuration in which operating leverage disappears.
At 9.5x forward earnings, the credit residue is already paid for
Jefferies trades at 9.45x my fiscal 2027 estimate and 9.85x the $4.68 consensus , against a sell-side 12-month target of $60 . The two estimates are not quite like for like: mine is struck on the as-converted base of roughly 238.5 million diluted shares across the voting and non-voting classes, above the 229.7 million outstanding at the quarter end because it includes unvested awards, net of assumed buybacks . On book, the stock changes hands at 0.99x its $46.55 reported book value per common share , while Goldman Sachs trades at 2.40x its $372.64 book value per common share .
That gap is not free money. Goldman earned a 22.4% return on tangible common equity over nine months against Jefferies' 12.9% return on adjusted tangible equity — a measure whose denominator excludes the deferred tax asset and is the more flattering of the two, so the true return gap is wider still. Most of Jefferies' discount is earned by lower returns rather than awarded by fear, and anyone expecting a re-rating toward bulge-bracket multiples is expecting the wrong thing. Capitalize a sustainable return a shade above the current 12.9% at an 11% cost of equity with 3% long-run growth and you support roughly 1.3x tangible book, which is where the $35.21 adjusted tangible book per fully diluted share already puts the shares.
The upside is one turn of multiple, not three. At 10.5x the $4.88 estimate the shares are worth $51 twelve months out, about 11% in price before the dividend. Eleven and a half times a bull case near $5.65 — comp leverage, a clean Hildene close and a leveraged-finance recovery arriving together — gets to $65; 8.5x a bear case near $3.92, where the deal pipeline cools and a Radiant reserve lands, gets to $33, and that bear is also where an adverse Western Alliance outcome would sit. Positive but narrow, and resting on three things already visible rather than hoped for: $113 million of disclosed own-account fund equity, a compensation schedule that rolls off on its own, and a core segment already earning over 16% pre-tax.
The year-end print, due in early January, tests the second of those. A full-year compensation ratio at or below 53.0% with no Radiant reserve confirms it; a first-half 2027 ratio stuck at 53.5% or higher kills it. The market is charging Jefferies for an open-ended credit liability when the disclosed own-account mark is a third of a dollar a share, and the genuine residual — third-party litigation and fundraising damage — warrants a lower multiple, not a permanently impaired franchise. At 9.5x forward earnings the shares are worth owning on that distinction.
This article is for informational purposes only and does not constitute investment advice or a recommendation to buy or sell any security. It reflects the analysis and opinions of the author as of the date of publication and may not be updated. Do your own research and consult a licensed financial advisor before making investment decisions.