The cavalry isn't coming: UK stocks, relistings, and AI analysts | Ben Isaac, Brizo Capital

Summary

• Ben Isaac’s hot take is that AI materially expands the need for good analysts rather than replacing them. Pre-chewed information can become commoditized — expert-network databases are “ultimately a quasi-commodity” — so the premium shifts to differentiated, on-the-ground work (trade shows, cold calls, scuttlebutt) requiring “judgment, finesse, people skills, EQ” that models “will never have access to.” Across his 70–75 Columbia students, median work quality has risen sharply with AI. On whether the absolute best pitches are materially better than 4, 5, or 7 years ago, though, he says, “I don’t know”; there was strong work five years ago. “These are still not a substitute for investment judgment.”
• The episode’s governing thesis on UK stocks: “the cavalry isn’t coming to save us.” Ben says it’s easy to find statistically cheap companies globally, with a meaningful fraction of good businesses at attractive valuations. The US — around 70% of global market cap — and perhaps Australia, India, and the Nordics are the only robust, well-functioning equity markets, in his view. Stewardship issues in disclosure, incentives, and “crucially capital allocation” can cap valuations. Brizo is particularly attracted to the UK: many companies are global businesses, and there’s “a really logical, deep pool of strategic and private-equity buyers.”
• Isaac defends US relistings as more than multiple arbitrage, with data. Of 30-plus “real businesses” relisted over five years, he says it’s hard to find ones that didn’t rerate on announcement and keep that rerating, while liquidity grew even faster. Average UK turnover is a third to a quarter of that of a similarly sized US stock. Suboptimal listings can have operating costs: one UK company pays its head of North America “multiple times what the CEO is getting paid,” and UK dividend policy is “trapped in amber.” Walker’s pushbacks — “one-time sugar high” versus buybacks, and selection bias — draw Isaac’s candid “I can’t disprove the counterfactual.”
• The Walmart-of-Kenya trap: demand a spread over local cost of equity, and decompose the risk premium rather than slap on a country number. A Kenyan retailer at 10x earnings with 10% growth implies roughly a 20% return against a low-to-mid-20s cost of equity — “probably kind of expensive.” In Brazil, with the Selic in the low teens, Isaac says you need 20%+ fundamental returns “just to get out of bed.” A Brazilian ag business he likes has mostly dollar-denominated streams, trades at a single-digit earnings multiple, and has historically grown 20–40% nominally. He contrasts a Turkish business owning the Houston Airport operating concession with Almaty Airport, which he argues does not have a Turkish cost of equity.
• Only 10% of UK PLCs require directors to own equity, which “creates obvious cascading consequences.” Isaac says patient capital and a 1–5% stake can change the contours of engagement, but fixes like buyback mechanics may require “gradual education processes.” A five-year fix for a cash-burning steel mill at 30% of replacement value (sellable at 70% of book) is a different proposition from the same hypothetical mill already generating 7–12% ROE.
• UK private-equity bids can leave shareholders selling before later value is realized, but Isaac sees opportunistic bids as part of a governance transition. In “year 10 of basically just continuous outflows from active management,” local managers may be incentivized to take a win. Isaac sees early bids as a signal that “the water is warm,” helping bring competitive bids and price discovery. He’s fought bids and “gotten yelled at by some very large-cap CFOs.”
• South Korea is “a super attractive market” competing with the UK for Brizo’s capital. Reforms are stacking up and valuations are compelling, but Isaac says family-controlled businesses are a bigger part of the market than in Japan, and key decision-makers may be slow to adapt. Japan has fewer complicated holdco structures.

Deep dive

1. AI in the classroom: pre-chew information, don’t outsource the thinking

• Isaac teaches Columbia’s Applied Security Analysis (with co-professors Evan and Zach; the course is approaching its 20th anniversary) and has worked with interns at Brizo. His core tension: students want professional-looking output fast, but there’s a gap between Claude “helping them do the initial analysis” and “doing their thinking for them.” His gym metaphor: “If you build a machine to do the leg press for you… it doesn’t make you any stronger.”
• The house rules: use the tools for clean information-gathering, alternative sources, Python scripts, and big scrapes — diligence that was “incredibly time-consuming and painful” takes far fewer person-hours — disclose where you used them, and have them “focus on pre-chewing information as opposed to digesting information for you.”
• Across 70–75 students, median work quality is up dramatically. On whether the absolute best pitches are materially better than they were 4, 5, or 7 years ago, Isaac says, “I don’t know”; he recalls strong work five years ago. “These are still not a substitute for investment judgment.”
• Walker’s corroboration from Substack pitches: the writing reads beautifully until you push back and “they’ve got nothing.”

2. Firm adoption is “all over the map” — and hallucination-talk is a tell

• What firms want in hires varies by mandate and compliance posture. Big shops without the IT architecture are “still stuck on things like Copilot, which really doesn’t have the same kind of horsepower as Claude” (Walker: “I vomited in my mouth a little bit”). Pod shops have amazing tools and data cleanliness; a Claude account that can’t touch company files is “radically less powerful.”
• Disconnected from everything but the raw internet, these tools are “only so much better than Google” — their power comes from feeding them incremental context.
• Walker’s view: a year ago hallucinations topped his own worry list; now if someone’s first concern is hallucinations, he thinks they may not have been playing with the tools much.

3. The hot take: AI expands the need for good analysts

• Isaac first hedges — “for now, people don’t know,” and near-term pressure comes more from economic constraints at large organizations, including fee compression and AUM challenges. But his “perhaps hot take” is that AI “actually materially expands the need for good-quality analysts,” because the premium shifts to differentiated information requiring “judgment, finesse, people skills, EQ.”
• Expert-network databases can become “ultimately a quasi-commodity,” particularly as organizations get comfortable feeding that material into a context window. Meanwhile, senior people are tied up in investment committees and administrative work; young, energetic analysts can do on-the-ground scuttlebutt that models “will never have access to.”
• Brizo hires many international interns to bring different cultural contexts and perspectives. Walker’s job-seeker corollary, inspired by recent Pershing Square Challenge winners: “If you’re a CEO, a hard hat is a short indicator. If you’re an analyst or an aspiring intern and you’ve got a hard hat photo… that’s a job-offer signal.”

4. The student superpower: people will take your call

• Isaac’s commercial for MBAs: “While you’re a student, you have a superpower: a lot of people actually want to get on the phone with you.” The hit rate on a polite 15–20-minute research ask is “pretty good,” and learning to take doors slammed in your face “without taking it personally” is a durable career skill.
• Walker’s arithmetic: expert networks run $750–$1,000 an hour, ex-CEOs sometimes charge $5,000, and students can get a few minutes with CEOs for free. “You can get access that is literally priceless.”
• Isaac’s caveat: know “the load-bearing components of your thesis.” Massive write-ups that reveal “I know the name of the CFO’s cat” can be impressive without driving outcomes. Knowing what’s relevant informs whom to contact and what to ask; sounding prepared “unlocks a totally different level of conversation.”

5. The hunting ground: cheap stocks everywhere, functioning markets almost nowhere

• Brizo’s global, fundamentals-driven mandate is “pretty agnostic about where” companies are, assuming they can underwrite them and be treated fairly. Isaac says it is easy to find statistically cheap companies globally, and a meaningful fraction are good businesses at attractive valuations. In his view, the US — perhaps alongside Australia, India, and the Nordics — is the only robust, well-functioning equity market. The US is about 70% of global market cap, while the vast majority of roughly 40,000 publicly traded securities are outside it.
• Why the UK specifically: many companies are large global businesses not tied to the UK domestic economy, with English-language operations and “a really logical, deep pool of strategic and private-equity buyers” if public markets fail to value them. Southeast Asia may offer opportunities, but the rules on the ground and governance evolution are at a different stage.
• The problem is stewardship, “not necessarily coming from malice”: disclosure, compensation, incentives, and “crucially capital allocation.” Operationally, many businesses “aren’t broken”; these are “fixable problems.” Walker names their shared framing: “the cavalry isn’t coming to save us. We need to save ourselves.”

6. Engagement: push on open doors, accept gradual education

• Walker’s frustration: he pitches alignment — give management stock, “stock goes up, we all get rich” — and gets treated as “a foreigner sitting in your little shoebox of a closet.” Isaac says patient capital and being a 1–5% shareholder can change the contours of the discussion. He can think of very few cases where an investor simply dictated terms as a “brilliant American” and saw the stock change six months later.
• The structural root: “only 10% of UK PLCs have a requirement that directors own equity. That creates obvious cascading consequences” — not because the directors are bad people, but because they face different incentives. Blockers can include legacy leverage norms or misunderstandings about buybacks, addressed through “gradual education processes — and that can be a very unsatisfying answer.”
• Selectivity matters. Isaac finds the metaphorical, cash-burning steel mill at 30% of replacement value (sellable at 70% of book) less attractive than other opportunities. If fixing it takes five years, that is a different economic proposition from the same hypothetical mill already generating 7–12% ROE. Walker jokes that “the bottom-feeder in me” hears the discount and says, “Yeah, I’m in.”

7. Relistings: real operational consequences, not just multiple arbitrage

• Walker’s skepticism: relisting a UK/Spain-revenue business in the US “feels like financial engineering for financial engineering’s sake.” Isaac pushes back: one large UK company had to pay its head of North America “multiple times what the CEO is getting paid,” because UK CEO compensation is far below the market for American talent. Without well-rated equity and the ability to pay in equity, it can be harder to attract and retain talent; a more liquid, highly valued equity can also be useful for M&A.
• “The dividend policy in the UK is kind of trapped in amber from a very different set of facts.” One Brizo portfolio company says its view of the optimal margin/growth balance doesn’t suit its UK shareholders’ preferences. A changed shareholder base can affect what the business can do.
• The data: among 30-plus “real businesses” — with real revenue, profitability, and substance — relisted to the US over five years from the UK, Australia, Scandinavia, and other jurisdictions, Isaac says it’s hard to find ones that didn’t meaningfully rerate on announcement and keep that rerating. Liquidity grew even more: the average UK equity trades, as a percentage of shares outstanding, at about a third to a quarter of a similarly sized US company. He says relisting failures were driven by underlying business problems: GAN got into trouble because of its business, and would have gone to zero in another market too.

8. Walker’s two pushbacks: sugar high, and selection bias

• Pushback one: isn’t the relisting pop “a one-time sugar high” versus buying back discounted stock? Isaac agrees buybacks are “absolutely a great alternative”; one portfolio company is doing exactly that, and Brizo is happy to “let them cook.” But he can think of few other examples in the markets they’re looking at. For companies constrained by talent or liquidity, relisting can address more than valuation.
• Pushback two: is the rerating caused by the relisting, or by a biased sample of shareholder-minded managements? Isaac says, “I can’t disprove the counterfactual.” The cohort spans industries, and some companies aren’t meaningfully US businesses. In CRH’s case, Trian pushed for the move when the stock was around 35 or 45, Isaac thinks; before that, it wasn’t meaningfully on the board’s agenda. He’s prepared to believe relistings disproportionately select for companies that care about shareholder outcomes.

9. The Walmart-of-Kenya problem: decompose the premium, demand the spread

• Walker’s chronic struggle: a Kenyan company may screen cheaper, but mixing African towers and UK distributors at equal 5% weights hides very different risks. Isaac’s method is to break the premium into currency (“far and away the biggest risk”), government risk, and operating risk. He contrasts a business in Turkey that owns the Houston Airport operating concession — which may move with Turkish equity flows — with Almaty Airport, which he argues does not have a Turkish cost of equity. Hard-currency, inflation-linked tower contracts with energy-price pass-throughs can also make the apparent risk premium misleading.
• The mechanics: a Kenyan retailer at 10x earnings growing 10% nominally implies a roughly 20% fundamental return — against a low-to-mid-20s local cost of equity, “that’s actually not really cheap. In fact, it’s probably kind of expensive.” In Brazil, with real rates in the low teens, he says you need 20%+ fundamental returns “just to get out of bed.”
• A Brazilian ag-linked business Isaac likes has mostly dollar-denominated streams, trades at a single-digit earnings multiple, and has historically grown organically at 20–40% nominally. “We may wind up getting this one totally wrong,” but its growth plus earnings yield is well above the Selic, which is why he finds it attractive. A US company with a 14–17% fundamental return and a 10–11% cost of equity may beat a Kenyan company returning 30%.

10. “Brokenness is relative”: US dysfunction, UK takeouts, Korea next

• Walker channels Einhorn’s jelly-donut critique: are passive-distorted US markets, especially the dregs of the Russell 2000, really well-functioning? Isaac says people critical of the US should tour other markets. US governance is “getting worse at the margin, but it’s still the best in the world, or at least for now.” He points to a liquid market for corporate control, a large private-equity bid for cheap businesses, and established activists. He acknowledges pockets of dysfunction, particularly among smaller and less-loved stocks: “Brokenness is relative.”
• On UK private-equity bids: the UK is “in year 10 of basically just continuous outflows from active management,” which can incentivize local managers to take a win. Isaac’s ecosystem metaphor: “Maybe we’re the deer and they’re the wolves… I’d like to have something where I’m not being eaten.” Early-stage governance renaissances — the US in the ’70s and ’80s, illustrated by Danny DeVito’s liquidation-value chalkboard in Other People’s Money — can start with opportunistic bids signaling “the water is warm.” Isaac has fought bids and “gotten yelled at by some very large-cap CFOs”; he doesn’t like seeing companies bought for less than they’re worth.
• South Korea: Isaac calls it “a super attractive market,” where reforms are stacking up and valuations are compelling. He says family-controlled businesses make up a larger part of the market than in Japan, and decision-makers may be slow to adapt; Japan also has fewer complicated holdco structures. Walker separately contrasts Japan’s low-ownership managerial class with Korea’s family control. Brizo has owned Korean businesses for years and expects to continue investing there. Korea competes for capital with the UK, whose companies tend to be larger and more international, and where corporate governance is already in place.