In short: Referenced only — held, no new view. Weight ~5.95%. Sheet: NPATA 9 → 20, 30.50%/yr, the book's highest expected return by a wide margin, and 122% total growth in three years. That input is an outlier, and the sheet does not discuss it.
In short: STRONG BUY (portfolio). Fair value A$8.4 vs A$4.89 — 42.1% undervalued; ER 15.8%; fwd PE 18.1 vs a 29.3 five-year average; the price implies −9.5% growth against 12.0% expected (+21.5pp), fifth on the universe's reverse-DCF list. Also the seventh-worst performer YTD at −41.8%.
In short: ~5.5% of the book, and the highest expected return in the entire portfolio at 30.50% a year on NPATA rising from 9 to 20 by 2028 — one of the three names growing above 20%. 14x NTM P/E on a 20% EPS CAGR; 29,900 shares (the largest share count) yielding $5,202.47. Down 30% over twelve months. The governance flag raised in April — Brett Kelly margin-called on pledged shares — is not revisited here.
Kelly Partners buys Australian accountancy practices and runs them in partnership with the accountants who founded them, a model deliberately copied from Constellation Software's approach to small software firms.
It carries the highest expected return of any holding here: profits (measured on a cash basis that adds back acquisition accounting) are forecast to rise from 9 to 20 by 2028, implying about 30% a year. It is also one of only three names in the book expected to grow earnings above 20%, at 14 times next year's earnings — an unusually cheap price for that rate of growth.
One caution the letter does not repeat. In April this position was demoted a conviction tier for a governance reason: the founder had been margin-called on $64m of shares he had pledged. That risk has not been revisited, and it is precisely the kind of thing "integer management with skin in the game" is supposed to screen for.
In short: STRONG BUY (portfolio). Fair value A$7.6 against A$4.4 — 42.1% undervalued; fwd PE 18.1 vs a 29.3 five-year average (38.2% under); RDCF needs 7.0% against 12.0% expected. Also the universe's third-worst performer YTD at −47.6%, on a 13.7% ten-year CAGR — the widest gap in the list between price action and rating.
In short: BUY. The most extreme numbers in the table, in both directions. Bought 13 Nov 2023; $2.6 against a $5.3 fair value — +103.5% — while being the worst performer in the book at -55.1% YTD. EPS 9 (2025) → 20 (2028) is +122.22% in total, a 30.50% annual rate, by far the highest expected return of any holding. Forward PE 18.1 against a 29.3 five-year average. Results due 6 August. The governance concern documented earlier in this archive is not revisited here.
Kelly Partners buys Australian accountancy practices and runs them under one roof. It is the most extreme position in the book in both directions: down 55% this year — the worst performer by a wide margin — while carrying the highest expected return of any holding, at 30.5% a year.
That figure comes from earnings per share forecast to rise from 9 to 20 by 2028, a more than doubling in three years. The fair-value calculation puts the shares at roughly double today's price, and the current multiple is 18 against a five-year average of 29.
The size of both numbers is the warning. A projection of 30% annual growth from a company whose shares have more than halved is either a genuine mispricing or a forecast that has not yet caught up with reality — and results due on 6 August will settle a good deal of it. Note also that this update does not revisit the governance problem documented earlier in this archive, where the founder was forced to surrender a third of his stake after a margin call. The rating is BUY; the risk is not discussed here.
In short: STRONG BUY, Strong(+) conviction — and still falling: −55.1% YTD, now second-worst in the universe. FV A$7.0 vs A$4.0 = 42.1% under; fwd PE 18.1 vs 29.3 (38.2% under); RDCF 7.5% vs 12.0% expected. The rating has not moved through a fifty-per-cent decline, which is either conviction or the position that most needs re-underwriting.
In short: STRONG BUY, Strong(+) conviction — and the universe's worst performer YTD at −52.3%. FV A$6.9 vs A$4.0 = 42.1% undervalued; fwd PE 18.1 against a 29.3 five-year average (38.2% under); RDCF 10.0% required vs 12.0% expected. The gap between a −52.3% year and a 12.8% ten-year CAGR is the widest price-versus-record spread on the list.
In short: BOUGHT — 7,700 shares, limit 4 AUD, half the Judges proceeds. "Kelly Partners Group is a serial acquirer providing services like bookkeeping, tax planning, and advice on how to grow a business or handle money wisely." Three reasons: (1) "Plenty of runway — Kelly Partners Group should keep growing at tremendous rates going forward." (2) The forced-selling explanation: "In the past, Brett Kelly received quite some margin calls… I heard from a great source that the margin calls should be over now. The margin calls definitely had a negative impact on the stock price. As a result, the stock is now oversold." (3) Valuation on the accounting-sector metric: "P/NPATA 2027: 14.3x; 2028: 11.5x; 2029: 9.2x. That's cheap for a company that has doubled its revenue on average once every 3 years." Context: the 7 May list had it as the worst YTD performer on the whole watchlist at -48.8%, Strong+ conviction and STRONG BUY on only 2.08% modelled undervaluation. Brett Kelly shared the Omaha stage with the author.
Kelly Partners buys accounting firms in Australia — bookkeeping, tax and business advice — and runs them under a shared structure. Half the money from the Judges sale goes here: 7,700 shares at a limit of 4 Australian dollars.
The reason the shares are cheap is unusual and has nothing to do with the business. Brett Kelly, who runs the company, had borrowed against his own shareholding and received margin calls — demands to put up more money or sell. Forced selling by a founder pushes a share price down regardless of how the company is performing. The information acted on here is that those margin calls are now finished, which means the selling pressure should stop.
On value, the shares are measured against a metric used in the accounting sector — profit after tax but before writing off the cost of the firms acquired, which is the closest thing to real cash earnings for this kind of business. On that basis the shares cost 14.3 times expected 2027 profit, 11.5 times 2028, and 9.2 times 2029. For a company that has roughly doubled its revenue every three years, that is cheap.
The caveat is stated in the article and should be kept: those multiples only hold if the company hits ambitious growth targets.
In short: STRONG BUY, Strong+ conviction — and simultaneously the worst YTD performer on the whole watchlist at -48.8% (5-yr CAGR +10.3%, 10-yr +13.9%). Model: EPS growth 12.0%, FWD PE 30.6 against a fair exit 25, expected return 10.17%, fair value 4.4 against 4.3 = only 2.08% undervalued — so the Strong Buy is a conviction rating, not a valuation one.
In short: Fifth-worst performer at -7.2%. Performance table only; no fresh view. The worst YTD name on the 7 May watchlist at -48.8%.
In short: Listed Strong+ on the conviction slide; covered in Part I. No new view here.
In short: Strong+ conviction, with a governance flag. An accounting serial acquirer "down 39.8% since the beginning of the year and 60% from its peak. In the meantime, the underlying performance was (very) good." Valued on NPATA (profit before non-cash amortisation — "very similar to Warren Buffett's idea of 'Owner Earnings'"): $11m in 2026 = 20.8x forward, falling to 14.3x / 11.5x / 9.2x on management's 2027-29 targets — "not expensive given the long runrate." But the tier cap is the founder: after the AFR reported Brett Kelly was margin-called on $64m of pledged shares and had to hand over 34% of his stake, "This is not good governance if you ask me. It's something Warren Buffett would never ever do." A call with Kelly is booked for the Monday, with Omaha later in the month.
Kelly Partners rolls up small Australian accounting firms — bookkeeping, tax, business advice — buying a controlling stake while leaving the local partners with skin in the game. The shares have been cut roughly in half this year, down 60% from the peak, while the underlying business kept performing well, which is the setup Slegers looks for.
He values it on NPATA rather than reported profit. When you buy an accounting firm, accounting rules make you write off the price of the client relationships over the following years, which reduces stated profit even though no cash goes out the door. Stripping that charge back out gives a number much closer to the cash the owner actually gets — the same idea as Buffett's "owner earnings." On that basis the shares are at about 21 times this year's figure, dropping to roughly 14, 11 and 9 times if management's 2027, 2028 and 2029 targets are met.
The reason this sits one tier below his very best ideas is not the business but the man running it. Founder Brett Kelly borrowed personally against his own shareholding; when the price fell, the lender demanded more collateral and he ended up surrendering 34% of his stake — over seven million shares — to an undisclosed lender, which itself pushed the share price down further. Slegers is blunt: "This is not good governance if you ask me. It's something Warren Buffett would never ever do," and he openly wonders whether Kelly is "a great Owner-Operator, or a 'marketeer' for investors like us." He is not selling; he has booked a call with the CEO and will meet him in Omaha, and promises to report back. Read the position as owned and cheap, with the founder question genuinely open.
In short: STRONG BUY — and the watchlist's worst performer at −40.0% year to date (against a 21.4% five-year CAGR). Note the tension the sheet itself prints: at 30.6x forward against a 29.3x five-year average it is 4.4% above its own history, so the Strong Buy rests on growth (15.0% expected, +5.0pp of reverse-DCF margin) rather than on the multiple. Added to four days earlier at AUD 5.50.
In short: BOUGHT — $10,000, Q 2,700, limit AUD 5.50. "Essentially a holding company for accounting firms" that buys 51% stakes "and let[s] the original accountant keep running the business (keeping them motivated with their own stake)." Three-part model: "Accounting clients rarely switch, so revenue is predictable year after year. Taxes always need filing (it's recession-proof). Growth is simple: just acquire more firms and repeat." The setup: "down 36.7% since the beginning of the year and 60% from its peak. In the meantime, the underlying performance was (very) good." Valued on NPATA — amortisation added back because it is "not a real cash expense… very similar to Warren Buffett's idea of 'Owner Earnings'" — at 21.8x expected 2026, then 15.0x (2027), 12.0x (2028) and 9.6x (2029) on management's own $16m/$20m/$25m targets. The stated condition: "very cheap if Brett Kelly can execute its plans."
Kelly Partners buys 51% of small accounting practices and leaves the original accountant running the firm with the other 49%. That structure is the point: the person who keeps the clients also keeps half the profits, so they stay motivated. Clients rarely change accountants, tax returns must be filed in good years and bad, and growth simply means buying the next practice.
The shares have fallen 36.7% this year and 60% from the peak while, on the company's own reporting, the business kept performing. The valuation is done on NPATA — profit with amortisation added back — because amortisation is an accounting entry for past acquisitions rather than money leaving the business; it is the same adjustment Buffett describes as "owner earnings." On that basis the shares are at about 22 times this year's expected profit, falling to roughly 15, 12 and 10 times on management's targets for the next three years.
The condition is stated openly and should be kept: it is "very cheap if Brett Kelly can execute its plans." Those forward multiples are management's targets, not results.
In short: Disclosed holding, category "Where Rules and Humans Still Win." "Taxes and accounting are legally required, and business owners need a trusted advisor to navigate them" — regulation as the demand floor, relationship as the moat. Added to a week later (2026-MAR-15).
In short: A Strong Buy, and one of the three positions carrying the book. "For us, Medpace ($MEDP), Games Workshop ($GAW), and Kelly Partners Group ($KPG) are doing really well right now" — roughly +$40,500 unrealised on a ~7.0% weight, the third-largest gain in the portfolio, despite the ~50% six-month drawdown argued three days earlier as the #2 buy.
In short: Top Buy #2 — expected return 15.3%, "Current Undervaluation: 94.9%." "An amazing Owner-Operator. Brett Kelly owns over 46% of the company," and it "has grown over 19% yearly since it went public… you double your money every 3.75 (!) years." Down nearly 50% in six months: "You would expect something is seriously wrong with such a market reaction. But the company is actually executing well on its expansion to the United States… they're seeing more growth opportunities than they can currently fund. That's why they've raised some debt." The cause is named as AI-disrupts-accounting fear — "I think it's quite funny. All of a sudden, it seems like AI can disrupt any industry. I'm not very concerned about this." Brett Kelly at the AGM: "the company is trading at a share price today, that if we had excess capital, we would certainly be buying our shares back… with a great deal of enthusiasm and at large scale."
Kelly Partners buys Australian accounting firms and now US ones too, leaving the partners in place and taking a share of the profits — bookkeeping, tax and business advice, the least glamorous and most repeat-purchase work in finance. Founder Brett Kelly owns more than 46% of it, and it has compounded at over 19% a year since listing, which doubles money roughly every four years.
The shares have halved in six months. Slegers argues the operating facts point the other way: the US expansion is going well, and management's problem is that it has found more firms to buy than it has money to buy them with — hence the debt raise, and a guide to 20%+ continued growth. The market's worry is that AI will make accountants redundant, which he treats with open impatience: "All of a sudden, it seems like AI can disrupt any industry." The strongest evidence he offers is the founder's own words at the annual meeting — he would be buying back stock "with a great deal of enthusiasm and at large scale" if he had spare capital, and the price has fallen further since. The model puts the shares at 94.9% below what he thinks they are worth. One caveat the reader should carry forward: this issue predates the disclosure, two months later, that Kelly was margin-called on shares he had pledged — a governance problem that eventually limits how much conviction the archive gives this name.
In short: STRONG BUY in the portfolio, despite trading 4.4% above its own five-year multiple (30.6x against 29.3x). The case is growth: 15.0% EPS growth, a 13.2% expected return, and an AUD 11.7 fair value against AUD 7.13 — 38.8% under. Worst YTD performer in the list at −15.1% against a 32.2% five-year CAGR; argued in full a week later as the #2 buy in the portfolio.
In short: STRONG BUY. Weight 7.6%, performance +106.2% — "currently the second-best performing stock in Our Portfolio… up +106% since we first bought it in November 2023." Executing on the US expansion in "a fragmented industry [that] provides them with plenty of acquisition targets." Two takeaways from the annual meeting: "Founder Brett Kelly believes the stock is undervalued" and "they're seeing more growth opportunities than they can currently fund." FY31 targets — revenue $500m, EBITDA $175m, NPATA $40m — imply +24.4% / +31.4% / +28.2% a year: "If the company achieves these growth rates, the stock would be very undervalued in hindsight." Valuation: 31.2x forward against a 29.4x five-year average ❌ — the only Strong Buy in the book that fails its own multiple test — offset by an Earnings Growth Model return of 13.4% ✅ and a reverse DCF requiring 17.1% using NPATA against management's 30% ✅.
Kelly Partners buys small Australian accounting firms and runs them jointly with the local partners, who keep a stake so they still care.
It has more than doubled since it was bought in November 2023 and is still rated the strongest thing in the book — which needs explaining, because on the face of it the shares cost 31 times earnings, slightly more than their own five-year average.
Two things resolve it. First, reported earnings badly understate this business: buying an accounting firm forces years of write-offs against profit even though no cash goes out, so Slegers uses NPATA, which adds that charge back. Second, management has published targets for 2031 — $500 million of revenue, $40 million of NPATA — that imply growth of roughly 25-30% a year, and the founder says publicly he thinks the shares are undervalued and that the only thing limiting the company is capital, not opportunities to buy.
Compare it with Games Workshop in the same issue, at almost exactly the same multiple and rated Hold. The difference is not the price; it is that one has a published growth plan the price can be measured against and the other does not.
In short: #5 pick — the whole case is one slide from the shareholder meeting. FY31 targets: revenue $500m, EBITDA $175m, NPATA $40m, which against today's numbers implies +24.4% revenue, +31.4% EBITDA and +28.2% NPATA a year. "Those figures look very attractive. As a result, Kelly Partners could deliver amazing returns for shareholders." Table: 2.5% net margin (the lowest on the page — minority interests and amortisation), 13.5% ROIC, 35.9x forward.
Kelly Partners buys small Australian accounting firms and runs them in partnership with the local principals, who keep a stake.
The entire case here is one slide from the annual meeting: management's targets for 2031 of $500 million of revenue, $175 million of EBITDA and $40 million of NPATA. Converted into annual rates that is roughly 24%, 31% and 28% a year — very fast for an accounting roll-up.
Two cautions the write-up does not raise. These are management's own numbers, presented without the 30-40% haircut Slegers normally applies to forecasts. And the reported profit margin of 2.5% in the same issue's table shows why NPATA is used at all: reported earnings are almost meaningless here because of the accounting charges from buying firms and the profits owed to partner-shareholders. Both facts are fine — they just mean the case rests on the target being met, not on the current numbers being cheap.
In short: #8 most-picked; −22.9% — and the issue's cleanest value-vs-price divergence: "Kelly Partners Group grew its intrinsic value (Owner's Earnings) by 22.6% last year. Yet the stock moved in the opposite direction… The result? The company became 45% cheaper!" Brett Kelly is quoted explaining the absent buyback as a capital constraint rather than a valuation judgment: "We are currently limited by the capital available… We're overwhelmed with opportunities, so we haven't done any buybacks… If we had extra capital, we would be buying back shares enthusiastically and on a large scale." Slegers' read: "Brett Kelly clearly thinks the stock is undervalued at today's price."
Kelly Partners buys small Australian accounting firms — tax, bookkeeping, business advice — taking a controlling stake while leaving the local partners with a share, so they keep caring about the result.
The number that matters in this write-up is the gap between two things that usually move together. The underlying value of the business, measured as owner's earnings, rose 22.6% in 2025. The share price fell 22.9%. Put those together and you are paying about 45% less for each pound of profit than a year earlier, with nothing having gone wrong operationally.
The founder agrees, and the way he says so is the useful detail. Asked why the company is not buying back its own cheap shares, Brett Kelly answers that every available pound is already committed to buying more accounting firms — "we're overwhelmed with opportunities" — and that with spare capital he would be repurchasing "enthusiastically and on a large scale." A company with more good uses for money than money is a specific and checkable kind of good problem.
In short: STRONG BUY — bought 13 November 2023, 7.6% of the portfolio and the second-largest contributor to the record (about +$56,000 of profit). One of the three names Slegers says "drove the majority of returns."
Nothing matches this filter.
Verbatim excerpts from the public transcripts (auto-pulled at each mention's timestamp, lightly cleaned). Timestamps link into the video; "source page" opens that commentator's full analysis at this row. Click a mention's header line to expand it (one open at a time). For personal study — not investment advice.