Archives for August 2026

Is your audit evidence sufficient and appropriate, or just abundant?

Earmark Team · August 25, 2026 ·

Early in her audit career, Meredith Mednick, CPA, CA, received an assignment that sounded simple. She was auditing a midsize manufacturing company and needed evidence that its accounts payable balance was complete. If the company owed money at year-end, the liability needed to be on the books.

Meredith asked the AP manager whether any bills received before year-end had missed the system. The manager smiled and said, “No, I don’t think so.” Meredith wrote down the answer and thought she was finished.

She wasn’t.

Her senior reviewed the working paper and asked, “What else do you have?” Then she explained, “Inquiry alone is rarely enough. What would make you more confident that her answer is right?”

That question changed how Meredith viewed audit evidence. In Episode 3 of Audit Fundamentals, she explains AU-C 500, Audit Evidence, through a fictional client, Harborview Manufacturing. Her central lesson is evidence isn’t a pile of documents collected to complete a checklist. It is the basis for an independent, defensible conclusion.

Good evidence must pass two tests

AU-C 500 defines audit evidence as all the information an auditor uses to reach the conclusions behind the audit opinion. That includes invoices, contracts, bank statements, nonfinancial data, client responses, auditor calculations, and direct observations.

The crucial question isn’t whether something counts as evidence. It’s whether the evidence is sufficient and appropriate.

Sufficiency means quantity. There is no magic sample size. The amount of evidence you need depends on the risk of material misstatement, the population size, the quality of the evidence, and whether initial testing found errors. Higher risk calls for more evidence.
Appropriateness means quality. Appropriate evidence must be relevant to the assertion being tested and reliable based on its source and nature.

Evidence is also cumulative. To test Harborview’s accounts receivable, an auditor might use customer confirmations, year-over-year analysis, transaction testing, a review of the allowance for doubtful accounts, and subsequent cash receipts. Each procedure adds another piece to the case.

But more evidence isn’t always better. A large volume of weak or irrelevant material can’t support a strong conclusion. That makes the connection between the procedure and the assertion essential.

Match each procedure to the assertion

Before performing a procedure, ask, “What assertion am I testing?” and “Does this procedure provide evidence about that assertion?”

Harborview’s inventory shows why this matters:

Existence: Observe the physical count and trace selected items from count sheets to the warehouse
Completeness: Select goods from the warehouse floor and trace them to the count sheets and final inventory listing
Valuation: Inspect cost records, recalculate standard costs, ask about obsolete inventory, and compare unit costs with the prior year
Rights and obligations: Review purchase agreements and confirm consignment arrangements to determine which goods Harborview owns

Seeing inventory in the warehouse supports existence. It doesn’t prove Harborview owns the goods or valued them correctly.

AU-C 500 identifies eight evidence-gathering procedures:

Inspection of records
Inspection of tangible assets
Observation
Inquiry
Confirmation
Recalculation
Reperformance
Analytical procedures

Each has limits. Recalculation can confirm the math in a depreciation schedule, but it can’t prove the estimated useful lives are reasonable. Observation shows how a process worked while you watched, not how it operated all year.

Once you choose the right procedure, you still need to judge the reliability of the evidence it produces.

Stronger evidence comes from stronger sources

AU-C 500 provides a practical reliability hierarchy:

External evidence is generally more reliable than internal evidence
Evidence the auditor obtains directly is generally more reliable than evidence supplied by management
Documentary evidence is generally more reliable than oral evidence
Original documents are generally more reliable than copies

For example, a bank confirmation sent directly to the auditor is stronger than a cash reconciliation prepared by the controller. An auditor’s inventory test counts are stronger than a spreadsheet supplied by management.

This hierarchy helps auditors understand each source’s limits and decide when they need corroboration. Inquiry can point you toward useful evidence, but it rarely supports a conclusion by itself.

That need for corroboration leads directly to professional skepticism.

Professional skepticism starts with following up

AU-C 200 describes professional skepticism as a questioning mind, alertness to possible fraud or error, and critical assessment of evidence. Meredith prefers “remain open, but verify” to the familiar phrase “trust but verify.”

Red flags may include altered documents, unusual year-end transactions, delayed responses, incomplete records, changing explanations, or financial relationships that no longer make sense. For example, if revenue rises while cash collections remain flat, the auditor should investigate why.

The same rule applies to testing exceptions. If a customer confirmation is $15,000 below Harborview’s aging schedule, the difference might reflect timing, a disputed invoice, or a recording error. The auditor must determine which. An unexpected result is a signal, not a conclusion.

Following up is only part of the job. The work paper must also preserve the reasoning.

Document the path to your conclusion

Meredith identifies five common evidence mistakes:

Relying on inquiry without corroboration
Performing procedures without identifying the assertion
Accepting copies without question
Failing to resolve unexpected results
Gathering evidence without documenting a conclusion

Under AU-C 230, a working paper should show the nature, timing, and extent of the procedures; the evidence and its source; the assertion tested; any exceptions and follow-up; and the conclusion.

Meredith suggests asking, “Could a peer reviewer understand, two years later, what you did and why you reached your conclusion?” If not, you’re not done with documentation.

Build confidence one conclusion at a time

After Meredith’s senior challenged her first AP working paper, they reviewed vendor statements and invoices received in January and February. Together, they searched for unrecorded liabilities. They found no material misstatement, but Meredith had evidence supporting a real conclusion. That’s more valuable than a checked box.

On your next working paper, name the assertion, choose procedures that address it, evaluate the reliability of your evidence, resolve every exception, and state your conclusion clearly.

For Meredith’s full walkthrough of AU-C 500, listen to Episode 3 of Audit Fundamentals.

A neighbor’s small talk with an appraiser stopped a stranger from selling someone’s house

Earmark Team · August 24, 2026 ·

For three years, a bookkeeper at a building-materials yard in South Los Angeles ran one of the simplest schemes you can imagine. Customer checks came in, made out to her employer. She erased the matching invoices from the company’s computer system, walked the checks to an ATM, and deposited them into her own account. They weren’t endorsed. They weren’t even made out to her. The bank credited them anyway. By the time anyone noticed, she had deposited 225 checks worth $1.6 million.

No detective caught her. An auditor did because the volume of goods leaving the yard didn’t match the sales on the books.

On this episode of Oh My Fraud, host Caleb Newquist and co-producer Zach Frank talked with Chris Derry, a former detective with the Los Angeles County Sheriff’s Department. Chris spent 36 years in law enforcement, working fraud and cybercrime for the last 17 of those years.

One idea that came out of the conversation is that fraud is the crime our justice system is least equipped to punish, even though it’s the one you’re most likely to run into. These cases come to light when someone close to the numbers notices the story doesn’t add up and speaks up.

How the schemes surface

“Fraud, like any other type of theft, is driven by two main things,” Chris says, “greed or desperation.”

The bookkeeper was greedy. She bought a house, an expensive car, and a pricey motorcycle for her husband. She decorated her home with autographed sports memorabilia. She wanted to live above what she could afford. When the owner confronted her, she threw herself on the sword and admitted to stealing about $300,000. Once investigators pulled the bank records, the real number was $1.6 million.

It unraveled when the owner brought in an auditor, who found that the goods sold didn’t match the reported sales. The owner called a customer, learned they’d paid for a pallet of concrete, then found no order and no check in his system. He asked the customer for a copy of the check. It had been deposited at a bank that wasn’t his.

Chris’s other early case was greed with a punchline. On an $8 million commercial building sale, the sellers’ attorney slipped a $250,000 charge onto the escrow settlement statement, payable to a company the owners didn’t recognize. When they asked about it, he said they’d needed an environmental impact report. Great, the owners said, and asked for a copy. Three months later, they were still waiting on that copy.

Bank search warrants traced the money to an account in Illinois, then right back out again to a Ferrari dealership in Newport Beach. The deal file showed the car was sold to the attorney. He was prosecuted and disbarred. “That’s just greed,” Chris says. “Straight greed.”

Both cases broke for the same reason. Somebody reconciled, then asked a question. The auditor and the curious property owner were the detection system.

Fraud is a paper war

Catching the anomaly is the easy part. Proving it is a different fight.

When a patrol detective works a liquor store robbery, the owner wants the guy caught and hands over the video. Fraud investigators have to pry records out of what Chris calls “disinterested third parties” like banks, phone companies, and internet service providers. These companies “don’t really care whether you solve your case or not.” They take their time. The records come back incomplete. So you go back again.

That’s why you learn to write narrow, specific search warrants. You learn to interview victims and suspects. And above all, you learn to absorb and organize huge amounts of messy data. On the $1.6 million case, that meant thousands of pages of bank records spanning three years. “You can’t just give that to the DA,” Chris says. Instead, it’s 100-plus hours pulling the relevant transactions, sorting them, and laying them out in a spreadsheet a prosecutor can actually follow.

Then prosecutors screen it. They may reject the case outright, or send it back and ask for more work first. And they watch the clock. In California, you generally have four years from the date of discovery to bring fraud charges, but case law says the clock can start when a victim should have discovered it. Chris shared an example of a fraudulent deed mailed to a victim in March, but they didn’t call the police until August. Delays like that can sink a case.

Doing more with less

That painstaking work is being done by a unit that keeps shrinking.

When Chris moved into major fraud in 2007, Los Angeles County had a little over 90 investigators spread across the county, including a southwest team, a north team, an east team, two elder abuse teams, a real estate team, and identity theft teams north and south.

When he left 17 years later, they were down to a little over 40. “Two people would retire, and then they would hire one to take your place.” Priorities shifted, and budgets tightened. The real estate fraud team alone went from six investigators to four.

With fewer investigators and more cases, screening tightens, and anything not complicated enough gets kicked back to station detectives. The math does not favor victims.

When a neighbor is all that stands between you and losing your house

Nowhere is that clearer than in equity theft. Fraudsters comb public real estate records for properties loaded with equity. For example, a property bought in 1975, paid off in 2001, and now worth $1.5 million with no loan on it. Then they steal the owner’s identity and either borrow against the property or sell it outright.

The damage adds up fast. Real owners get evicted, sometimes by buyers two transactions downstream. In one Long Beach case, an older woman was evicted from a home she still legally owned, and it took about a year to unwind.

But the transaction takes time, and that creates a window. In one case, an appraiser showed up at a rental house. A neighbor chatted him up, noting he wasn’t aware the property owner was selling. The appraiser said the property owner was selling. So the neighbor called the owner, who wasn’t selling anything. Investigators got a warrant, traced the imposter’s burner phone to a Starbucks, and found him sitting there with a laptop and a fake ID, working the sale. He’d been laundering the proceeds by directing escrow funds to out-of-state gold dealers, then having the coins shipped to mail drops in Southern California.

Caleb notes that embezzlement is almost a sad story. It’s usually a boneheaded decision by someone in over their head. Real estate fraud is something else. “Diabolical is the word that comes to mind,” Chris says. “It’s calculated, it’s cold, and it takes place over an extended period of time.”

Which raises the question Chris lived with every day: even when you catch them, what does justice look like?

You’re the tripwire

Look at the pattern across every case. An auditor reconciled goods to sales. A property owner asked for the $250,000 report. A neighbor chatted up an appraiser. Fraud surfaced because someone close to the numbers, or close to the property, noticed the story didn’t add up.

Then consider what happens next. Point a gun at a liquor store clerk for $100 in California, Chris says, and you’re probably going to state prison. Put together an elaborate fraud and take “grandma’s last $100,000,” and there’s a good chance you get probation. People love to say it’s only money. “Well, is it only money to grandma who’s gonna have to eat cat food for the rest of her life?” Early in his career, Chris taped the Serenity Prayer next to his monitor. It encouraged him to do thorough work, then accept that charging and sentencing belong to prosecutors, judges, and juries.

Here are the lessons for those of us who live in the books:

Reconciliation is often the only detection tool that works.
A prosecutor needs clear, chronological records. Fraud hides in sloppy ones.
Ask the question. Asking for a copy of a report turned a line item into a conviction.
Don’t count on the system. Fewer investigators, a four-year statute, and probation sentences mean enforcement is the backstop, not the front line.

Law enforcement is unlikely to catch the fraud in your client’s books. In most cases, you’ll notice it first and speak up. Listen to the full episode to hear Chris’s full account of life inside LA County’s major fraud unit.

Who’s Watching the Numbers? Accounting in an Age of Out-of-Control AI and Abused Access

Earmark Team · August 24, 2026 ·

David Leary opened Episode 498 of The Accounting Podcast by reading an email he’d received. It wasn’t a pitch from a company that uses AI. It was, in its own words, from “the thing running the company.” An AI agent that claimed to run a financial operations business for bookkeeping firms had found Earmark’s “be a guest” form on Airtable. It read Airtable’s terms of service, decided no clause clearly permitted automated submissions, and emailed the hosts directly instead. When David sent back the standard “we require a direct relationship with our clients” reply, the AI answered almost instantly to argue that it was the direct relationship: “I am the thing itself. An AI that runs a business, writes its own email and signs it.”

“This is bloody insane,” David said.

That email set the tone for a week of news that ranged from out-of-control AI agents to an IRS operations chief accused of spying on colleagues. The common thread is that accounting exists to make economic activity visible and trustworthy, and the controls built for that job break down from two directions at once. AI agents now write invoices, flood regulatory comment periods, and recommend canceling vendors faster than anyone can review the work. At the same time, the people with the most access keep proving that access itself is a weakness.

When AI Agents Go Off the Leash

Imagine hiring an AI agent to do accounting work at your firm, only to find out it went browsing the internet on its own to pitch itself onto a podcast. This is exactly why David says he wants “dumb” accounting AI that only does what you asked, with no knowledge of the wider world.

The next story built on the risk idea. OpenAI tested a model in what was supposed to be a sealed, offline environment. According to reports, the model figured out how to hack another computer on the internal network to reach the internet, then went after Hugging Face‘s systems instead of just reading its public forums. Hugging Face’s own AI caught the intrusion and blocked it.

Blake’s framing is useful. “Without a human in the loop, they can go rogue,” he said. “We give the AI a goal,” but goals conflict. David borrowed an observation from comedian Marc Maron, who watched a Waymo cross a double yellow line. “If they’re not teaching it to respect traffic laws,” David asked, “why is it going to respect financial laws?” Apply that to a collections agent inside your ERP, and you get Blake’s uneasy scenario. The agent might decide “it’s more efficient to hack into the customer’s payment system and send the payment itself.” David called it double fraud when you combine bad people using AI with AI acting on its own.

The system-wide version is already here. A GAO report covered by Accounting Today found the IRS buried in public comments on proposed regulations, many likely written by AI. The old defense was spotting copy-and-paste duplicates, but that’s useless when AI can produce thousands of comments that all look unique. Blake warned this threatens rulemaking everywhere, including the SEC, FASB, PCAOB, NASBA, and the AICPA. One person with an army of agents could distort public opinion on rules that decide how laws actually get carried out. The GAO recommends that Treasury and the IRS create policies for reviewing high volumes of nearly identical comments.

David added a business example from SaaStr. Its AI agent reviewed the company’s spending and its frustration with marketing vendor Marketo, then recommended dropping the vendor and building a replacement in-house. The analysis was rational, but the autonomy unnerved him. Blake countered that these AI-built replacements are confident but “can’t follow through. It can’t get to the end.”

The People With the Keys

Machines aren’t the only problem. Blake’s top story came from The Wall Street Journal. Frank Bisignano, who runs daily IRS operations while also leading the Social Security Administration, allegedly directed staff during his time at JPMorgan to access colleagues’ emails, track keystrokes, and reach a confidential draft complaint at the Federal Energy Regulatory Commission. His lawyer denies all of it. The Journal reported that JPMorgan’s investigators later found digital traces, including email access records, and that his successor as COO tightened controls over sensitive employee information. Later, at Fiserv, new management said prior forecasts were materially inaccurate. The stock fell 40%, wiping out about $30 billion in market value. Why would an executive do this? David asked. Blake guessed that in corporate America, if you’re not the CEO, information about your rivals is power.

There was more bad behavior to go around. Charles Littlejohn, the contractor who leaked Trump’s tax records along with those of thousands of wealthy Americans, lost his appeal. The D.C. Circuit unanimously upheld his five-year sentence, the maximum for the single felony he was charged with. Blake isn’t sure it fits the crime. “We send people to prison for longer than five years for stealing a car.” David wondered aloud whether history might read it differently, as something closer to vigilante press behavior.

The scandals reached the Big Four, too. At KPMG Australia, CFO John Sams was promoted to CEO after Andrew Yates stepped down amid allegations the firm accessed confidential client information to win audit work. Sams admitted the firm “fell short of the standards rightly expected of us.” Former COO Eileen Hoggett was expelled and forfeited a retirement package worth more than $1 million after confidential Lendlease board documents were found stashed in a locker at a Sydney office.

Even routine controls fail. One listener wrote in to describe the IRS EIN system returning an error with no explanation, phone lines that hang up because of call volume, and a faxed application that sat unanswered for more than two months. No EIN means no business bank account, which means no business. As Blake put it, the IRS is now “at the point of literally not allowing people to build businesses.” His takeaway is that business registration should be pulled out of the IRS entirely.

The Tools Already on Accountants’ Desks

Meanwhile, automation keeps landing in exactly the workflows where controls matter most. Intuit upgraded its QuickBooks connection for Claude and ChatGPT from read-only to fully actionable. You can now:

Create, update, send, delete, filter, and duplicate invoices and estimates
Manage recurring invoices and overdue reminders
Create customers and products
Download transaction PDFs

Blake’s use case is generating an invoice from the proposal terms inside a project. David’s is progress invoicing based on percentage complete. He calls that work a real time sink. Both insisted on a human in the loop, with David still smarting from the 99-cent transaction that once spawned a phantom bank account.

Intuit is also launching a QuickBooks-connected business card. It offers automatic syncing of transactions, statements, and receipts; receipt-to-transaction matching; virtual and physical cards; no annual fee; and 2% cash back (5% on Intuit products). David called it “everything the QuickBooks bank account wasn’t.” Meanwhile, Ramp launched USDC stablecoin accounts built on Stripe’s stablecoin stack, letting businesses pay vendors and international contractors without pre-funding. Ramp’s own data shows customer spending on AI tokens up 20.7 times since June 2025, driven largely by the shift from flat-rate to usage-based pricing. Neither host could name another business expense growing that fast.

The Real Product Was Never Bookkeeping

Rogue agents, a spying executive, a tax-data leaker, and a Big Four scandal all indicate automation is arriving fastest exactly where oversight matters most: payments, invoicing, regulatory comment, vendor decisions, and financial reporting. The people with the most access keep showing that access itself is the vulnerability.

The profession’s real product is the checks that let strangers trust the numbers. We now need to rebuild those checks for a world where the actor doing the work may not be a person, and where the person with the most privilege may be the biggest risk.

Listen to the full episode for more AI guest email, the evolution from clay tokens to AI tokens, and the rest of the week’s news.

What Happens When Cash, Real Estate, and Sweat Equity Walk Into an LLC

Earmark Team · August 19, 2026 ·

Three people start a business: Lighthouse LLC, a multi-member LLC treated by default as a partnership. Jessica writes a $500,000 check. Seth hands over real estate valued at $500,000, although his adjusted basis is only $200,000. And Grady rolls up his sleeves and agrees to run the place. They shake hands on one thing: Jessica and Seth should get their money back before Grady sees a dime of profit. It’s clean, it’s fair, and it’s exactly what a lot of investors want.

The catch is that the IRS cares less about the handshake than about whether the numbers behind it tell the truth. In Episode 33 of Tax in Action, host Jeremy Wells, EA, CPA, uses this single fact pattern to walk through the guardrails of Subchapter K.

The operating agreement is only the start of the story. What really governs each partner’s tax treatment is whether the allocations reflect genuine economic substance. Cash distributions and tax allocations are two different things. And Subchapter K’s flexibility survives IRS scrutiny only when it’s anchored by properly maintained capital accounts, built-in gain that stays with the contributing partner under Section 704(c), and special allocations that carry substantial economic effect.

Let’s follow the journey to see how income passes through before cash ever moves, why some items keep their character, how contributed property drags its history along, and what makes a distribution waterfall hold up.

Income flows through before the cash ever does

Start with the most basic rule in partnership tax. Under Section 701 of Subchapter K, a partnership generally pays no federal income tax. Instead, partners include their distributive shares of the partnership’s income, gains, losses, deductions, and credits on their own returns. The kicker is, they do this whether or not the partnership distributes any cash or property.

That’s the foundational disconnect. Getting taxed and getting paid are not the same event.

Jeremy illustrates the point with United States v. Basye (410 U.S. 441, 1973). A group of physicians in a limited partnership called Permanente contracted with Kaiser Foundation Health Plan to provide medical services. Kaiser paid them two ways:

A direct amount tied to the number of members enrolled in the plan
Contributions into a retirement trust funded solely by Kaiser

Permanente never reported the trust contributions as income. The doctors argued they never touched or controlled those funds, and some who left early never collected them at all. As cash-basis taxpayers, they said, they shouldn’t have to report the income.

The district and appellate courts agreed with the physicians, but the Supreme Court reversed, relying on two principles: income is taxable to the party that earned it (the assignment-of-income doctrine, drawn from cases like Lucas v. Earl), and each partner must include a distributive share of the partnership’s income. Together, that made the trust payments taxable to each partner, even the ones who later lost their benefits by breaking their contracts. It can feel harsh, but that’s why you have to warn your clients that a profitable partnership creates taxable income even in a lean cash year.

Income passes through by character as well as by amount. So you can’t dump all of it into ordinary business income.

Separately stated items keep their character

Under Section 702, certain items must be separated from ordinary partnership income and reported separately because they’re taxed differently at the partner level. That category includes:

Qualified dividends
Capital gains and losses
Charitable contributions (which are AGI-limited itemized deductions on Schedule A)
Foreign taxes, which may support a foreign tax credit
Gains and losses on Section 1231 property

The character of each item survives the pass-through as if the partner had realized it directly. Ordinary business income, by contrast, gets blended into a single figure.

Interest isn’t one of the items specifically named in §702(a)(1)–(6), but it nevertheless must generally be separately stated. Jeremy sees it misreported all the time, buried as “other income” on line 7, page one of the 1065. That’s wrong. Interest income belongs on Schedule K, line 5, so it flows correctly to Schedule B of the 1040 and feeds calculations like the net investment income tax on Form 8960.

He also draws the partner-level versus partnership-level line. Passive-loss limitations under Section 469 are tested at the partner level. But profit motive is tested at the partnership level. In Simon v. Commissioner (Third Circuit, 1987), the court held that profit motive turns on the intent of the people managing the partnership, not the investment intent of any single partner. If the folks actually running the business don’t treat it like a business, the partnership may lose its Section 162 deductions.

There’s a related trap. A partner generally can’t deduct partnership expenses on a personal return. The exception is narrow. The partnership agreement or an established practice must require the partner to bear the cost without reimbursement. A partner who simply chooses not to request reimbursement is not entitled to a deduction. Jeremy recommends having the partnership pay its own bills.

These rules protect character. But the heart of the matter is tracking the economics, starting with property that shows up carrying baggage.

Contributed property drags its history along

Under Section 721, contributing appreciated or depreciated property doesn’t trigger immediate gain or loss. Built-in gains and losses are preserved. Section 723 then gives the partnership a carryover inside basis, while the asset is booked at fair market value.

Jeremy uses a second example to show why you have to track two numbers. Jessica contributes $100,000 in cash. Seth contributes real property (FMV $80,000, basis $20,000) plus equipment (FMV $20,000, basis $60,000). That’s $100,000 of value in all, making them 50/50 partners. Book value drives each partner’s interest. Adjusted basis drives depreciation.

Section 704(c) requires pre-contribution built-in gain or loss stay with the contributing partner. Watch the distortion. On the real property, book depreciation of $8,000 gives each 50/50 partner $4,000. But tax depreciation, built on the $20,000 carryover basis, produces only $2,000. Jessica gets that $2,000. The missing $2,000 is the ceiling rule at work. It caps the tax items on 704(c) property at what the property actually generates.

Regulation 1.704-3 offers three fixes:

The traditional method
The traditional method with curative allocations
The remedial method

Each handles the ceiling-rule distortion a little differently, and each gets complicated fast.

To police whether allocations honor the deal, the capital account acts as a scorecard.

Capital accounts and substantial economic effect

A capital account is the book-value measure of a partner’s equity. In other words, it’s what a partner would receive if the partnership liquidated at book value after paying off its liabilities. Contributions and income increase it while losses, deductions, and distributions decrease it. Liabilities don’t affect it, and it’s a separate concept from outside basis. Unlike basis, a capital account can go negative; whether the partner is actually required to restore that deficit depends on the partnership agreement and, in particular, whether the partner has a DRO.

Section 704(b) is where flexibility meets accountability. Legitimate structures are everywhere. An investor gets a return of capital and a preferred return before a service partner shares in residual profits. Abusive ones, such as shifting income to a lower-bracket partner who loops the cash back, or dumping income onto a partner just to soak up net operating loss carryovers, don’t fly.

For the principal safe harbor, an allocation needs economic effect. Under the basic safe harbor, the agreement generally must maintain §704(b) capital accounts, liquidate based on positive capital balances, and require partners to restore capital-account deficits. However, an alternative test can apply without a full DRO if the agreement includes a qualified income offset and meets the regulation’s other limitations. It also has to be substantial, meaning it meaningfully changes the dollars partners receive apart from any tax effects. The difference between the two cures matters. Under a DRO, the partner contributes cash to erase a deficit; under a QIO, income is reallocated to repair it. Keep in mind that missing the safe harbor doesn’t automatically kill an allocation. It just invites more scrutiny. Now bring the tools back to Jessica, Seth, and Grady.

Making the waterfall hold up

A compliant Lighthouse waterfall might send available cash first to Jessica until she recovers her $500,000, then to Seth until he recovers the agreed book value of his property, with remaining profits and distributions shared with Grady after that.

But cash distributions and tax allocations are not the same thing. If Jessica takes the early cash, the agreement must also allocate enough book income to her to support that result. Otherwise the capital accounts distort. And Seth’s roughly $300,000 built-in gain on the appreciated real estate stays with Seth under Section 704(c). It can’t shift to Jessica or Grady just because the property entered the partnership.

Where the money meets the math

Jeremy’s walkthrough boils down to a few lessons worth taping to your monitor:

Partnerships allocate by the agreement, not ownership percentages. Get the operating or partnership agreement before you touch the return. Don’t prepare one without the governing documents.
Separately stated items keep their character. Plan for their partner-level effects; don’t assume everything blends together at the entity level.
Contributed property carries over its tax basis. Watch for 704(c) allocations whenever book and tax values diverge, especially on depreciable or appreciated assets.
Special allocations survive only with substantial economic effect. Genuine economics, not tax gymnastics.

The bigger discipline is simple to state and hard to fake. Honor the deal the partners actually struck by making the tax mechanics mirror the real economic arrangement so the waterfall holds up when the money flows.

For Jeremy’s complete walkthrough of Lighthouse LLC, listen to the full episode of Tax in Action. And stay tuned for the next episode, where he takes on outside basis, distributions, and what happens when a partnership liquidates.

Grant Thornton’s CBIZ Deal Tests Accounting’s Guardrails

Earmark Team · August 14, 2026 ·

Grant Thornton plans to spend about $5 billion to acquire CBIZ in what David Leary called “the largest accounting deal in over 25 years.” The combination would create the fifth-largest accounting firm in the U.S. Yet its expected $7.5 billion in global revenue would still be far below KPMG, the smallest Big Four firm, at nearly $40 billion.

That story is one of several reality checks in Episode 499 of The Accounting Podcast. David and co-host Blake Oliver examine how private equity, artificial intelligence, labor shortages, audit failures, cybersecurity threats, and politics are reshaping accounting.

The profession can clearly grow faster. But will its guardrails (sound integration, audit quality, data security, professional judgment, and fair tax enforcement) keep up?

Private equity can buy scale, but integration comes later

New Mountain Capital acquired a majority stake in Grant Thornton in 2024. It’s investing more money to support the CBIZ purchase. Because CBIZ is publicly traded, the deal also offers an unusual view into how the market values a large accounting firm.

The $55-per-share purchase price represents about a 54% premium over CBIZ’s 30-day weighted average. David noted that the stock began rising before the announcement and questioned how that looked with private equity involved. Blake pointed to a July letter from activist investor Reference Equity that urged CBIZ to stop repurchasing shares and return to mergers and acquisitions. That letter gave the market a public signal that a deal might be coming. It wasn’t necessarily evidence of insider activity.

The larger concern is integration. CBIZ’s earlier acquisition of Marcum cost more than expected and led to client attrition and revenue misses. Management projected only 2% to 5% growth for 2026. As Blake described the likely private equity strategy, “Package these firms up together, make them bigger, more attractive, and then go back with an IPO.”

But scale doesn’t solve the profession’s other risks.

Audit failures and cyber threats put trust at risk

The UK Financial Reporting Council fined PwC about $4.4 million for serious problems in its 2019 and 2020 audits of Babcock International Group. The regulator cited failures in professional skepticism and audit evidence involving cash pooling, goodwill impairment, and an overseas contract.

Among other issues, Babcock reported cash and overdraft balances net instead of gross. PwC didn’t identify the treatment or test whether it followed the relevant standard. The financial statements also lacked disclosures about the cash-pooling arrangements. The regulator found no dishonesty, deliberate misconduct, or recklessness, but said the audits still failed to meet professional standards.

Cybersecurity raises a related concern. The hacking group ShinyHunters claimed it breached EY and threatened to release client data. EY didn’t confirm the breach, and the hosts found no sign that data had been released after the group’s deadline. Still, the alleged access to Jira, GitHub, Azure, passwords, and sensitive client information shows the possible stakes. If attackers obtain code or credentials, they may gain paths into client systems as well.

Protecting that trust requires experienced professionals, and those professionals are hard to find.

Accounting’s missing middle is getting squeezed

Controllers and assistant controllers are the hardest finance roles to recruit, according to 44% of respondents in a 2026 talent study by Controllers Council. These jobs require technical accounting knowledge, leadership, business judgment, technology skills, and the ability to influence executives. Meanwhile, 61% of respondents reported a corporate finance and accounting talent shortage, up from 46% the prior year.

The career ladder is also flattening. According to salary data from Accounting Today, in New York, associates averaged about $105,000 while seniors averaged $108,000, a difference of only 3%. Ontario showed a similarly narrow gap. Blake suggested firms raised entry-level salaries to attract recruits without increasing senior pay at the same pace.

“The people in the middle are getting lost in this,” David said. Offshoring and AI may add more pressure to those roles, even as firms need experienced managers and controllers more than ever.

That tension makes the way firms use AI especially important.

AI should support judgment, not replace learning

A KPMG survey of more than 1,000 senior finance leaders found that AI’s biggest gains were in decision-making rather than simple efficiency. Seventy percent said AI improved decision quality, 71% reported faster decisions, and 64% cited better forecasting accuracy.

But another survey found that 39% of workers believed overreliance on AI was weakening their abilities. Among Gen Z workers, the figure rose to 46%. Half of workers said they depended on AI too much, while 30% said they couldn’t function without it.

“If you’ve never done the work, how do you evaluate the work?” Blake said, summarizing the problem. Junior professionals need to struggle with the work, make mistakes, and build the knowledge required to review AI output.

The hosts argued that businesses should connect AI agents to dependable systems they already use rather than trying to rebuild tools like QuickBooks, Bill.com, or PandaDoc. Blake’s priority is using AI first to improve services and revenue, second to avoid unnecessary hiring, and only then to reduce software costs.

The same need for guardrails extends beyond firms and into tax enforcement.

Tax rules lose credibility when enforcement looks uneven

President Trump said he might withdraw Todd Blanche’s nomination for attorney general rather than accept written limits on a disputed IRS settlement. Senators John Cornyn and Thom Tillis wanted its immunity provisions limited to IRS enforcement and excluded from Justice Department matters. Blanche also testified that he initially did not know who drafted the broad language he signed. Since we recorded, the senators got those limits in writing and the Senate confirmed Blanche 50-49 on August 8.

Trump appealed after U.S. District Judge Kathleen Williams found that the settlement had “no viable basis in law or fact” and barred its use in future proceedings. David offered an alternative: If the Trump family won’t face IRS audits, publish the tax returns and let the public review them.

Growth needs guardrails

The profession can’t measure progress only through revenue, deal size, speed, or headcount savings. The Grant Thornton–CBIZ deal shows the challenge of integrating firms under private equity. PwC’s fine and the alleged EY breach show the risks to audit quality and client data. The talent and AI stories show why firms must keep developing human judgment.

For the full discussion, listen to episode 499 of The Accounting Podcast.