Summary

Jingshu Zhang’s core call: Fiserv ($FISV, ~$55) at ~6–7x adjusted EPS is priced as a melting legacy ice cube, but he sees “one of the absolute best” executive teams in fintech running a reset year. Zhang clarified that the $8–8.30 adjusted-EPS figure was for 2025, not 2026; his forward case is $9+ of adjusted EPS in 2027. If the team delivers low-double-digit EPS growth, “at least 12 times earnings, if not more” makes it “a triple-digit stock” from ~6x 2027.
The whole payments sector took “1-2-3-4 punches” — IPAY down 17.1% this year, Visa/Mastercard in bear territory — and Zhang argues the AI-disruption punch was, in his view, “wrongly killed.” His four moats: AML/KYC compliance, decades of proprietary data (“a walled garden” LLMs can’t access), two-sided network effects, and the “very little discussed” sales channel — Sumitomo Mitsui distributing Fiserv in Japan and 10,000 US financial institutions cross-selling Clover to mom-and-pops.
The variant perception is a talent thesis: COO Takis Georgakopoulos built JPMorgan payments from seven people and zero revenue to 25,000 people and a $20B run-rate, and Jamie Dimon threw him a private party when he left. Now ex-JPM and Stripe leaders (Lia Tao, Sanjay Saraf, Robert Clarkson, and others) have followed him — “this is the Avengers assemble moment” — none of it visible in press releases or earnings calls yet.
Zhang’s account of what broke under Frank Bisignano (he cited 12% Glassdoor approval): cost cuts so deep that two Client Technology Advocates served DoorDash, eBay, Walmart and Webster Bank, plus service deterioration. Walker’s framing was that Q3'25’s 16% growth included 10 points of Argentine hyperinflation and roughly 4 points of one-time contract sales, with insufficient disclosure; the old CEO left at $200 for the Social Security job. New CEO Mike Lyons took the medicine within months.
Clover isn’t necessarily losing to Toast, Zhang argues, based on interviews with more than 100 and close to 200 restaurants: they serve different niches, and 70% of US POS is still legacy. Toast makes 5x the money on kitchen restaurants (“innovator’s dilemma”), while Clover is far cheaper for non-kitchen mom-and-pops; Zhang said Clover volume was hovering around 10% growth and expected it to grow by around 10% again this year. He believes Paysafe’s ISO grew Clover by more than 50% last year.
On capital allocation, Zhang calls GPN and Shift4 buybacks at 3.5x net debt/EBITDA “irresponsible,” while Fiserv paused its buyback, cut $900M of debt in Q4, and targets sub-3x by year-end. Walker’s counter-zinger stands: Fiserv bought $5.6B of stock at ~$170 in 2025, and it’s now $55.
Walker’s strongest pushback: insider buying is thin (Lance Fritz ~$650k, the legal officer, and Zhang estimated the new CFO at about $1M), CEO PSU metrics were deferred to the 2026 investor day — “are we six months too early?” Walker also noted that the chief administrator’s $500k purchase followed an earlier $800k sale. Zhang concedes: “this is the weak part of the bull thesis.”
Jana surfaced in February with an activist-style push to sell non-core assets. Lyons said the new team “didn’t create this structure, so they are not wedded to it”; Zhang says the team still sees benefits in the integrated FIG/Merchant Solutions model but would not be bound to it. A Clover sale could offer multiple arbitrage from what Zhang believed was roughly 6x EV/EBITDA, and Walker notes that the only payment moves that have worked lately have been split-offs or sales of a company’s last division.

Deep dive

1. Payments took four punches and the whole space is in the doghouse

Zhang’s opening frame: the IPAY payments ETF is down 17.1% this year and even “the kings and queens like Mastercard and Visa” are in bear-market territory. The sequence: the GENIUS Act’s stablecoin-disruption fears, PayPal’s CEO saying in September that consumers were weakening, Cloud’s new technology enabling easy software development and spooking vertically integrated software-payment names earlier this year, then AI-layoff and recession fears compounded by oil spiking on the Iran situation.
Walker’s setup of the puzzle: Fiserv at ~6x guided 2026 EPS “sounds very cheap and is very cheap” — but so are GPN, Shift4 and others, so is the AI risk real for a company processing billions of payments a day?

2. Four moats against AI — and the sales channel nobody writes about

Zhang grants AI is a real threat to FactSet and Morningstar — “I just unsubscribed my Morningstar subscription” because Gemini and Cloud summarize stocks better than its analysts — but payments differ: heavy AML/KYC regulation, decades of proprietary data that is “by its very nature a walled garden,” and self-reinforcing two-sided network effects (crediting Bow Street Capital’s observation that very few SaaS challengers ever cracked payments).
The dimension “very little discussed by all those posts”: distribution. Technical Substack writers “are really good at technology, but they are not sales people” — mom-and-pops need someone to sell and service the box. Fiserv entered cash-heavy Japan through Sumitomo Mitsui Banking Corporation, a merger of two banks and Japan’s second-largest bank by assets, and in the US runs 600 direct salespeople plus 10,000 financial institutions cross-selling Clover for commission.
Walker’s addition: the “every restaurant will vibe-code its own payment processor” fear ignores hardware and fraud — 99.9% uptime is fine for a Morningstar replacement but would “eat your entire margin” at a diner running a thousand transactions a day.
Zhang’s other top payments position: Euronet Worldwide — $2.5B market cap and $400M free cash flow excluding SBC. He said obtaining the licenses needed across 200 countries for cross-border transfers is difficult; even Remitly uses Euronet’s Dandelion infrastructure.

3. Deleveraging beats buybacks when you’re 3.5x levered

Zhang flagged GPN and Shift4 buying back stock at 3.5x net debt/EBITDA — publicly, when they traded at $80 and $70 — as “irresponsible capital allocation”: in a recession EBITDA falls, “the debt’s going to stay there,” and covenants could break. Executives who buy back “into extinction” miss that deleveraging derisks the business so equity “should theoretically enjoy a higher premium, not a lower.” His side heuristic: “usually what the FinTwits love on Twitter will fail.”
Fiserv is doing the opposite: buyback paused, $900M of debt chipped away in Q4, determined to get below 3x by year-end — “they are doing exactly what they are saying.”
Walker, “a former cable bull,” partly converted: good short-sellers use the moment equity flips below 50% of enterprise value while management keeps buying back as a go-short technical — “I’ve seen it so many times lead to tears.” And the zinger: Fiserv bought $5.6B of stock at ~$170 in 2025; the stock is $55.

4. What Zhang and Walker say broke under Bisignano

Fiserv had $21.2B of revenue last year, with 84% from the US and Canada and 16% international. Its two roughly equal businesses are Financial Institutions Group (FIG), which provides core banking, loan and deposit-account processing, digital payments, card transactions and card issuance, and Merchant Solutions, which includes merchant acquiring, digital commerce, mobile payments, security, fraud protection and Clover.
The business quality makes the mismanagement starker: FIG has 98% customer retention because replacing a bank core is “basically like walking down the street while having a heart surgery.” Yet Frank Bisignano — whom Zhang said had a 12% Glassdoor approval rate and described as the worst he had seen — presided over such severe cuts that employees anticipated a round of layoffs whenever there was a holiday. Zhang’s specimen: the Client Technology Advocate group, the “single point of accountability to a critical client,” was down to two employees covering DoorDash, eBay, Walmart, Webster Bank, WebBank and others. Lyons has grown it past 30 and acquired Darren Smith’s Smith Consulting Group, which had been run by Smith for 17 years; many people had left Fiserv for it after Bisignano’s return-to-office policy.
Culture from Zhang’s primary research: monitoring software called Sapiens on every computer left employees so freaked out “they would actually bring it to the bathroom with them”; nicknames like “Frank the Tank” and “John the Giblets” made Fiserv’s own IR “laugh hysterically.” “If it weren’t for the absolutely phenomenal business model… this thing would have been killed a thousand times over.”
Walker’s numbers framing: of Q3'25’s 16% growth, 10 points were Argentine hyperinflation and roughly 4 points were what he described as one-time contract sales — color the old CEO never gave the Street before leaving at $200 for the government post. From $10 in 2010 to $200 in early 2025, it was “the ultimate compounder stock… it just took it too far.” Former CFO Rob Hall bought back $1B of stock in Q3 after Lyons announced the reset; amid infighting rumors, Hall resigned during the Q3 call.

5. Variant perception: the JPMorgan payments “Avengers assemble”

Zhang’s stated edge: “the market sees this as a legacy play that’s going to melt away,” whereas he believes the new team is “one of the absolute best in the fintech space.” The “true legend” isn’t the CEO but COO Takis Georgakopoulos, who built JPMorgan payments from zero revenue and seven people to 25,000 people and a $20B revenue run-rate — double the size of Fiserv’s payments and Merchant Solutions business — and “always thinks in binary trees.”
The proof is who followed him: Ed Gaglio (JPM e-commerce/tech sales → global business development), Lia Tao (JPM embedded finance and solutions head, MIT PhD → enterprise platforms CRO), Sanjay Saraf (JPM chief product and customer officer → merchant services CPO), Robert Clarkson (Stripe Americas CRO → SMB/Clover CRO), Adam Hyde (JPMorgan Chase HR operations), and ex-Nuveen portfolio manager Johar Sohi leaving the buy side to be chief performance officer. “You won’t see the effect immediately” — but talent attracting talent is a positive reinforcing loop.
Lyons’s own pedigree: ex-Maverick Capital fintech analyst, involved in pruning Bank of America’s assets during the financial crisis before Buffett invested through warrants, and second-in-command at PNC — a major Fiserv client where he interacted with 100 Fiserv products.

6. Clover vs Toast: Zhang’s restaurant interviews say different niches, not a knife fight

Zhang interviewed more than 100 and “close to 200” restaurants — the boots-on-ground work Walker singles out as the episode’s differentiator. First finding: 70% of US POS is still legacy, and owners won’t switch while the box works — the Flushing restaurateur’s answer to his Clover pitch: “this thing still functioning, so we are not going to switch.” The market isn’t saturated enough for head-on competition.
The niche split: a kitchen restaurant is worth 5x as much to Toast as a non-kitchen mom-and-pop — “sort of the innovator’s dilemma. Why would I go there if I could make so much money here?” Toast’s in-house integration is smoother, while Clover uses more third-party apps but is cheaper for crepe- and burrito-type shops. Around Union City and Weehawken, Zhang said “it’s all Clover”; he also said Toast “kicks everyone’s ass” in Asian restaurants but “can’t really get into Mexican,” based on his conversations and his acquaintance Wenling Shi, Toast’s CEO.
Cross-validation: Zhang said Clover volume was hovering around 10% growth even through the quarter the stock fell 44% in a day — that quarter was attributed to Argentina and overcharging bank clients, not Clover volume — and he believes Paysafe’s ISO division grew Clover by more than 50% last year, calling it a best fit for US SMBs.
Disclosure delivered live: Zhang is collaborating with Fiserv’s top Clover salesperson, Austin, to refer POS sales for recurring revenue. Walker: “you’re on the Clover referral side… that was the disclosure.”

7. The math: clean adjustments, sub-3x leverage, a triple-digit path at 12x

Zhang likes “one-foot hurdles”: he cited $8–8.30 adjusted EPS for 2025, explicitly correcting the earlier 2026 reference. The adjustments are “much cleaner” than FIS or GPN’s — reported 2025 earnings were $6.34 per share versus $8.64 adjusted, with $1.90 of the $2.30 gap from First Data acquisition-related intangible amortization — so even ignoring all add-backs the $55 stock is under 10x earnings. “We don’t need a lot of things to go right for us.”
The forward case: ~2.8–2.9x leverage by year-end, $4B+ of free cash flow next year against a business valued at ~$28–29B, and accretive buybacks at these prices; low-double-digit EPS growth deserves “at least 12 times earnings, if not more” on more than $9 of 2027 adjusted EPS — “that’s a triple-digit stock” from ~6x 2027.
Early evidence of Takis’s chess: the StoneCastle acquisition paired with FIUSD, Fiserv’s own stablecoin, which Zhang believed was already being used by North Dakota Agricultural Bank. He argued that StoneCastle liquidity and recirculation through the bank consortium could retain liquidity there rather than allowing it to leave — “that was a brilliant move.”

8. Walker’s incentives pushback — and the Jana/split optionality

The episode’s sharpest challenge: management comp was struck at $180–200, February brought top-off grants, and Lyons’s $18M PSU package has metrics set only at the 2026 investor day — “is there another shoe to drop? … are we six months too early?” Zhang cited three post-reset insider buys: director Lance Fritz (~$650k; his only other insider buy, Parker Hannifin, nearly tripled), the chief legal officer, and the new CFO, whom Zhang thought bought about $1M. Walker also noted that the chief administrator’s $500k purchase followed an ~$800k sale earlier in the year. Zhang’s concession, worth keeping: “this is the weak part of the bull thesis — if the CEO and COO are not buying, why are we as investors so enthusiastic?”
His defense of Lyons — 80% of salary in annual equity awards, 60% of shares as PSUs based partly on free-cash-flow conversion and adjusted EPS, a 12x-base-salary ownership requirement, and “if Fiserv is really just a sham, I can’t believe someone as shrewd and career-driven as him would have jumped ship” — meets Walker’s proxy-based eruption: Lyons made $70M in 2025, so “get the fudge out of here” on the internal claim that “Mike is not that rich,” likened to the NYT couple “struggling” on $500k with a required $10k/month savings line.
Zhang’s TheLayoff.com confession: he posted there posing as an angry employee blasting Lyons for not buying stock in the open market (“Mike, shame on you”) — yet employees defended Lyons anyway (Glassdoor ~70–71% versus Bisignano’s 12%): “any change is a positive improvement.”
Jana surfaced in February with an activist-style case supporting the sale of non-core assets. Lyons said at the Wolfe FinTech Forum that the new team still likes the integrated FIG/Merchant Solutions model, while Zhang sees cross-selling benefits but noted that the new team “didn’t create this structure, so they are not wedded to it.” Zhang believed selling Clover from roughly 6x EV/EBITDA could create multiple arbitrage, though he would not agitate at this stage. Walker’s close: with two big divisions lacking obvious synergies, and with recent payment successes involving split-offs or sales of a company’s last division, “there is a natural split.”

Episode 179 - April 10, 2026 Anthropic's Cybersecurity Shock Wave + Farrow and Marantz on Altman