On-Demand Webinar: “The Bottom Line: Standards Updates, Estimates, and the Growing Impact of AI”

The Bottom Line #1: Standards Updates, Estimates, and the Growing Impact of AI | July 9, 2026

Presenters: Kathryn Good and Tim Partridge

This transcript was generated with AI and may contain errors or omissions. Please refer to the webinar recording for the most accurate version.

00:04 – Introduction and CPE instructions

PYA Moderator [00:04]: Good afternoon, everyone. Welcome to the first episode of PYA’s new accounting and audit-focused webinar series, The Bottom Line. Today’s topic is standards updates, estimates, and the growing impact of AI. PYA is happy to present today’s webinar on these important topics. As this event is for CPE credit, there are some additional steps that we’ll take in order to comply with NASBA’s requests. In accordance with NASBA, attendees will be provided with three polling questions during this session. Responses are documented electronically. Attendees must respond to all three questions within the allotted time in order to receive CPE credit. At the end of the webinar, a survey will also be provided for attendees to submit feedback regarding their webinar experience today. The polling questions and post-event survey must be submitted for proof participation today. Credit will only be earned if the session is attended in its entirety. Your CPE certificate is downloadable via the Continuing Education pane once you’ve completed all requirements. We’ve posted a PDF copy of the presentation slides and the resources pane. Also, later today you’ll receive an email with a link to a copy of the slides and a recording of the webinar. You may submit questions by typing a message into the Q&A pane. All questions posed during the webinar will be responded to via email after the webinar. You can customize your viewing experience by resizing, moving, or minimizing all of the panes within the webinar.net platform. With that, I would like to introduce our presenters, Kathryn Good and Tim Partridge.

01:49 – Presenters and webinar agenda

Tim Partridge [01:49]: Hello, everybody. My name is Tim Partridge. I’m a senior manager with PYA’s Audit and Assurance practice,

Kathryn Good [01:55]: and I’m Kathryn Good. I’m a principal in the Audit and Assurance practice, working alongside Tim, and we’re excited to be here with you on the very first installment of our new webinar series. We have an action-packed agenda for you today. We’re going to be doing a roundup of some of the significant accounting and audit standards that have gone into effect in recent years, with a brief refresher on those standards, some lessons learned now that we’re a few years in, and then some best practices to keep in mind going forward. Certainly, now compared to when these standards initially went into effect, we’re all using AI or artificial intelligence a lot more prevalently. So we’re going to share with you some ideas and use cases on how AI can help you in complying with these standards and in having stronger controls over accounting estimates, and we’re also going to talk about SECURE 2.0, which is a new regulation that’s specific to administration and operations of employee benefit plans. You know, I think each of these individually could probably merit their very own one-hour CPE session, so we’re going to be covering everything at kind of a 50-1000-foot high level as we’re going through.

03:10 – AI risk, privacy, and human oversight

And before we get into the meat of our presentation, since we are going to be talking about some AI use cases and ideas, we wanted to make sure that we provided some guidance on how to use AI safely and securely. I will say this the only time that we’ll be giving this type of disclaimer. This not a cybersecurity presentation, but we want to make sure everyone is aware of data privacy and confidentiality considerations. There’s a lot of financial data like customer information, contracts, forecasts. It’s sensitive, and you would not want to upload that to any type of open, public, unvetted AI tool or any model that doesn’t have good safeguards. You want to be mindful of the risk of hallucinations or inaccurate AI output, making sure that there’s a human validation of anything that an AI tool is spitting out, the AI tool should be transparent. Make the AI tool show its work. It shouldn’t be a black box that you can’t understand the different logic and assumptions that are going into an answer. And AI tools, just like any other software applications, the same type of standard IT controls still apply, so things like access control, segregation of duties, vendor management, keeping those in mind. But the most important thing here would be to keep the human in the loop. AI is a powerful tool; it can be a great resource and a great assistant. But as we all know, it’s not a replacement for human judgment. So now that we’ve got that out of the way, we can talk about

04:49 – ASC 842 lease accounting

our first financial reporting topic, and that’s accounting standards codification or ASC 842 related to lease accounting. As many of you all still remember, this was the lease accounting guidance that required just about all leases to be recorded on the balance sheet, and that was with a lease liability and a right-of-use asset. And as you can imagine, with having additional liabilities on a balance sheet, that’s impacting debt metrics. It’s impacting calculations for debt covenant compliance, it has an impact on EBITDA optics, and so for that reason, having accurate financial reporting as it relates to leases, it’s really important. They can pose risks, and having ongoing oversight is really important. As a refresher, like I mentioned, this was the accounting standard that said that pretty much any lease that’s got a term longer than 12 months goes on the balance sheet as a lease liability and a corresponding right-of-use asset. Just like with the previous lease accounting guidance, there’s the two categories, two buckets of leases: operating leases and what’s known as a financing lease, and something is considered a financing lease if it’s hitting any of the five criteria that we have listed here on the bottom half of the slide.

Kathryn Good [06:10]: So, if ownership is transferring to the lessee at the conclusion of the lease, if there’s a purchase option in the arrangement that the lessee is likely to take, if the term of the lease makes up a significant portion of the remaining life of the asset, if the present value of lease payments is greater than or equal to substantially all of the fair value of the leased asset, or if the asset is something really unique, really specialized that’s not going to have any alternative use to the lessor at the conclusion of the lease. If any of those kind of boxes are checked, that lease would be considered a financing lease. And what’s different now with the lease accounting guidance we’re under now is that the balance sheet impact of the classification is pretty minimal. It’s going to be the liability and the asset on the balance sheet, either way, but the income statement impact is more significant. With those operating leases having the full amount of lease expense categorized as part of operating expense, whereas with a financing lease, the amortization of the right-of-use asset is included as part of operating expense, but the interest expense piece recorded below that operating income line on an income statement, and both that amortization and the interest components are getting backed out of any EBITDA calculation. Since the lease accounting guidance initially went into effect a few years ago, the Financial Accounting Standards Board, or the FASB, has come out with a few different updates, most of which are just clarifying the original guidance. But one of the more significant ones came out in 2023. It was specific for leases between entities under either common management control or common ownership control, and this accounting standard update provided these organizations with a practical expedient where they could rely on written terms of an arrangement instead of the legally enforceable terms of an arrangement to determine whether or not that thing was a lease. The most significant piece of this update was related to useful lives of leasehold improvement assets for these organizations. Now, these leasehold improvement assets can be amortized over the useful life to the entire common control group instead of limiting it to one specific lease term. So it doesn’t necessarily apply to every single related party lease, just those under common ownership or management control. But it’s hopefully alleviating some lease accounting burden for those organizations. We’re a few years into having adopted lease accounting guidance, and now that we’re able to be where we are now and look back, there have been a few different issues and errors that we’ve seen over time now, and one of the more common ones being just issues in having a fully complete lease population. This came up a lot in leases that were embedded in maybe a service contract, an outsourced contract, maybe an IT services deal that had some dedicated equipment. Those contracts had either been overlooked upon initially implementing the lease accounting guidance, or just were missed if there was a new agreement that an organization entered into after adopting the lease accounting standard. We also saw issues in organizations where maybe there was just a lack of communication between the accounting team and whoever was signing a lease, whether that was legal, procurement, could have been any number of team members. So, some best practices and things we’d recommend here would be. To make sure that there’s some type of lease identification step as part of the contract initiation and workflow process, that would help flag any agreement that’s giving control of an asset. Having good training of whoever is signing these leases on what to look for on something that could contain a lease that can also help you maintain a more complete lease population. Taking time to perform a periodic contract review, maybe a quarterly scan of service agreements or other types of vendor agreements to identify any of these embedded leases. And you can also review general ledger detail of lease expense account for sure, but also other expense accounts on the income statement to make sure that there’s not anything flowing through the income statement that should also have a balance sheet impact. This our first AI use case that we’re going to share. An AI model is going to be able to help you in that contract review.

Kathryn Good [10:56]: It’s going to be able to review large volumes of contracts and help you look for those embedded leases. It can help you determine if something would be an operating versus a financing lease, and it can extract a lot of data that would be useful in calculating a lease liability, like the lease term, the payment amount, things like that. AI is also going to be able to review large volumes of general ledger detail and help find anything that has a lease-related expense to help you ensure completeness of your population. We saw for a lot of organizations the next issue that came up was maybe upon initially adopting the lease accounting standard using a manual approach like an Excel spreadsheet worked really well, but then over time as additional leases are added, you’ve expanded your lease portfolio. There’s been additional complexity with modifications, renewals, amendments. It left a lot of room for manual error. So our best practice recommendation here: consider scaling your systems to match the complexity of your lease portfolio. These softwares can help automate a lot of the calculations, make sure that you’re accounting for each lease consistently across the portfolio, and really just help reduce that risks related to manual error. You know, a dedicated software is not going to necessarily make sense for every organization. So, not suggesting that everyone go out and buy it, but maybe if you hit a threshold of 50 plus leases or a high level of complexity, that’s helpful for organizations that are sticking with Excel spreadsheets in a more manual process, there’s still a lot of benefits in just tightening up the controls over those spreadsheets. So things like a dual review of the formulas can be really valuable. Having some type of remeasurement checklist to trigger updates for any type of modification to a lease can be really helpful. Investing time on good controls, investing funds in software that can, at the end of the day, help streamline closing processes and help reduce or eliminate audit adjustments or last-minute adjustments. On that same topic of having errors and problems related to lease modifications, we saw over time that in organizations that treated lease accounting as a very static one-time check-the-box exercise, that modifications, changes that would result in a remeasurement of a lease asset and lease liability weren’t getting processed promptly. That can lead to audit adjustments or potentially even a control deficiency that’s being issued by an auditor. So again, having these checklists can help make sure that the accounting treatment for these leases and the modifications are reflective of the most recent agreement that’s in place, and having that good communication between accounting, legal, procurement, whoever signs these modifications to make sure that they are all being updated and reflected in real time. All right, the last lease-related item we’ll talk through relates to some of the judgments and assumptions used in lease accounting. There’s really not a ton of estimates involved in adopting the lease standard, but there are a few areas that require some judgments and assumptions from management. One being the incremental borrowing rate. The incremental borrowing rate is a refresher to everyone. You can think of that as if you were to have taken out a loan to purchase that leased asset. That leased asset was the collateral for the loan, and the loan was over the same length of time as the term of the lease. The other area for judgment is determining. The lease term in the first place, under the accounting guidance, the lease term would be considered the set period within the lease agreement, and then tacking on any extension periods within that lease agreement that are offered, or any early termination options that are offered, where your organization considers it to be reasonably certain to utilize, and that concept of being reasonably certain is a pretty high bar. It means that you’re highly confident that you’re going to take the lessor up on those options. So we saw a lot of instances where these judgments, assumptions, and methodologies weren’t applied consistently. So that led to having poor support for auditors to look at. Best practices here would be to develop clear policies around lease accounting, so documenting a consistent methodology to derive those incremental borrowing rates.

Kathryn Good [15:58]: Maybe having a documented framework for how to evaluate whether or not you’re reasonably certain to utilize a lease term, so maybe a decision tree type of document could be really helpful. And this another place to potentially utilize AI. AI can at least help in taking a first pass at documenting some of these assumptions and clarifying them, distilling them down with the human in the loop there to specify and tailor it to your specific organization and your lease portfolio. So that concludes our discussion of lease accounting, and now Tim’s going to talk through one of the newest audit standards has gone into effect.

16:43 – SAS 145 and audit risk assessment

Tim Partridge [16:43]: Thank you, Kathryn. So diving right in here to Statement on Auditing Standards No. 145 which is officially titled “Understanding the Entity and Its Environment and Assessing the Risks of Material Mistatement. It’s a giant mouthful. It was issued by the AICPA Auditing Standards Board. Became effective for audits of financial statements for periods ending on or after December 15, 2023 but is still having wide sweeping effects on audits and for those organizations being audited, as the guidance is being interpreted and put into actionable changes in the audit process. At its core, SAS 145 really forces a separate assessment for risk. Auditors must now assess inherent risk and control risk individually rather than grouping them together. The standard introduced the concept of a spectrum of inherent risk, requiring auditors to evaluate both the likelihood and the magnitude of a potential misstatement. Also introduced was a standback requirement. Auditors must now stand back and evaluate all the information gathered to ensure that they have properly identified all significant transactions, account balances, and disclosures. A big change still affecting current audits pertains to IT general controls or ITGCs. Auditors are no longer permitted to audit around IT. They must identify and evaluate the design and implementation of IT controls that address the risks related to financial reporting. So SAS 145 pushes auditors away from checklist-style auditing procedures by explicitly focusing on scalability. Auditors must tailor their procedures to the actual complexity for an entity rather than just an entity size, and this requires more skeptical and critical thinking approach to risk assessment. So in reality, a small business with 20 people could have a riskier, more complex structure than 1000-person corporation. And SAS 145 aims to put that risk where it actually belongs with the more complex entity. So why that matters and what it means if you get audited, you can be staring at mandatory control evaluations. Even if your organization is smaller and you have limited internal controls, auditors must evaluate the design and implementation of controls over significant risks for your journal entries and for financial statement close process. There’s also a focus on professional skepticism. Auditors are mandated to constantly challenge management’s expectations and explanations, and to look for potential signs of management bias or fraud. We also now have stricter scrutiny over estimates. The focus on inherent risk and professional skepticism means that auditors will ask for deeper documentation supporting management’s estimates. For example, allowance for doubtful accounts, inventory valuations, or fair value of investments. We’ll touch a little bit more on some of those later with our discussion with internal controls over estimates. You can also expect more inquiries about business risk. Auditors are now required to understand external factors affecting your business, such as economic conditions, industry-wide disruptions, or new regulations, and they’ve got to tie those risks back to your financial statements. So, what are some changes you’ll see in your audits? Well, as you can infer with the comments on the last slide, increased inquiries are a part of the process now. Standard is aiming, you know. To avoid preparation of those voluminous process narratives that you might be used to, usually those resulted in little to no documentation of identified controls. The controls are where that gold is, right? And SAS 145 aims to get at the heart of that information. The AICPA issued implementation guides for SAS 145 that focus on a clear linkage between the controls and the risks associated in those audit responses, with the standard adding specific requirements to obtain an understanding of controls around journal entries, this often results in auditors adding additional sampling and testing of those journal entries, as well as the associated processes and controls for entries, both during the year and at year end. Auditors will frequently focus on those preparation and review controls, making sure that you have robust documentation in your system. In these areas, will be super helpful in responding. If you’re thinking to yourself, “Well, we may not have very good documentation over these processes or controls, this an excellent opportunity to utilize some AI tools like Copilot or Gemini to help you craft some documentation that follows your current processes, clearly identifies those controls. Also, service organization controls, sometimes known as SOC reports, have received an uptick in focus with SAS 145. How you’re obtaining and reviewing these reports will be important for your auditor to understand.

Tim Partridge [21:18]: If you’re not currently obtaining and analyzing these for service providers like third-party payroll, finance, or software providers, you can count on your auditors inquiring about why that is. Simply obtaining that SOC report for your auditor is not a good control for your IT environment. So considerations must be given to the complementary user controls listed in those reports, as well as any findings in a SOC report should both be documented and addressed in your organization’s files. Auditors are also now specifically required to perform procedures beyond inquiry related to the implementation of other identified controls. They’re now required to perform additional walkthroughs that address the identified controls in these areas. Rather than just filling out questionnaires, you might get more requests for walkthroughs. So some of these inquiries and walkthroughs will likely also involve more time and attention from your IT teams. Some questions that you might get for your IT team could be along the lines of: If you’re allowing staff to work remotely, well, how do you know that unauthorized third parties are not using the same mechanism to access your systems? Also, what about? Can you show me those protections and security mechanisms in action? You might also see questions like, what procedures do you have in place for disaster recovery or backup related to hacking incidents? Can you also provide me a copy of your most recent backup failover or disaster recovery report? So, how can you prepare and streamline for your audit? Well, you can start by mapping your system access and your IT general controls. Be ready to provide IT policies, evidence of who has admin rights to your accounting software, and proof that system-generated reports are reliable. You can also document your processes, including those all important controls. If you haven’t already, write down step by step procedures for month-end close processes, journal entries, bank reconciliations, and major revenue expense streams, including software systems and any service organizations that are utilized to in those procedures. You want to include who’s performing those steps where preventive and detective controls are in place, like segregation of duties, arm’s length review, access system controls, where password and multi-factor IT tools are being utilized. You can also be responsive to those standback inquiries. You can expect your auditors to ask, “Did we capture everything? So, having a clear, organized trial balance, a good chart of accounts ready to review, can help alleviate some of those initial questions. I would say in the past year, our firms identified numerous significant deficiencies and material weaknesses related to those IT general controls that we talked about. This resulted in a lot of conversations with clients about the need to modernize their IT systems and the associated IT controls. And as a result of you know SAS 145, we’re now having those conversations. So we believe it’s helping to bring these important IT concerns to light and can add real value to you know not only your organization but the audit process in general. So with that, let’s take a look at our first polling question.

24:14 – Polling question: ASC 842 implementation

PYA Moderator [24:14]: All right, our first question is: What has been the biggest challenge in implementing the lease standard ASC 842, monthly and yearly accounting entries, determining operating versus finance lease classification, determining what discount rate to use, explaining to governance why these assets and liabilities are on the balance sheet, or I had no problem implementing the standard. Remember, you must respond to the polling questions to receive CPE credit. You will have 30 seconds to answer. Great, thank you for participating. Back to the presenters. Right,

24:56 – ASC 326 CECL

Kathryn Good [24:56]: looks like everyone’s hardships in adopting the lease accounting change. Really, kind of ran the gamut with several of you having no problem implementing, which is wonderful. With that, we’re going to switch gears and talk about another recent accounting standard change, ASC 326 current expected credit losses, or CECL. And when CECL went into effect a few years ago, it really had a massive impact on banks and financial institutions who were issuing loans to borrowers. But the guidance does apply to companies across pretty much any industry, and anyone with financial assets on the balance sheet, like trade receivables, contract assets, which we’ll talk about shortly, notes receivable. We’ve seen for a lot of organizations that the actual dollar impact to the financial statements upon adopting the standard has not been significant, but the overall process and really mindset related to calculating those expected losses is a big shift. CECL really shifted things from taking a very historic, backward-looking approach and really relying on historic losses to set an expected credit loss reserve, to flipping that on its head and having a very forward-looking approach, potentially looking at some macroeconomic data. So expanding on that, under the CECL guidance, you would want to think about a lot of different data points. So thinking about the past, the present, and the future. That historical loss experience is no longer the end all be all, but is still a valuable data point. You would want to think about current conditions to your organization and across the economy at the date of the balance sheet. Financial institutions obviously are thinking about borrower credit quality, collateral values, but others may want to think about the state of the economy and how that could impact collectability. And taking that forward-looking approach with having reasonable and supportable forecasts, thinking about that macroeconomic data like unemployment rates, interest rates, property values. Organizations can consider both internal data specific to their organization, or external data, maybe some industry benchmarks when they’re thinking about setting a CECL allowance or provision, and management has the option to put in a qualitative adjustment based on their understanding of their asset portfolio, their understanding of their organization, their local community, things like that. Just making sure that those qualitative adjustments are supported and justified. There’s been one recent implementation change since CECL went into effect, and that came out last year. This was a practical expedient to really simplify compliance with CECL guidance, and it applied to current receivables on the balance sheet and current contract assets, and it’s allowing organizations to assume that whatever conditions are in place as of the balance sheet date are going to remain that way for the remaining life of the asset, which would be one year or less. This standard went into effect for reporting periods beginning after December 15, 2025 which means that for anyone on a calendar year end, you’re able to apply that in 2026 As you might imagine, with all of the different data points we talked about a few slides back, data has been really a challenge for a lot of organizations in implementing this standard. There were some organizations. This one probably applies more to banks and financial institutions that relied really heavily on some type of complex vendor-provided CECL model without necessarily understanding them. Maybe they didn’t tailor the assumptions that were canned and included with that model to their specific circumstances, their specific organization. There were other entities that just didn’t have a lot of historical loss data to go off of, or struggle to factor in, make macroeconomic or forecast forward-looking information, and that led to having some inconsistent, harder to justify loss estimates. So, best practices here are to own your model, make sure management can fully understand justify the methodology that it’s used, and that the methodology is the same level of complexity as whatever the financial assets are. So, for some organizations, having a pretty simple aging schedule with some qualitative adjustments is going to check the boxes. Whereas others might need a much more advanced model, a much more advanced approach. You would want to make sure that all of the assumptions and methodology. Around your CECL reserve model are documented.

Kathryn Good [30:03]: They’re clear and supportable, easy to follow, and to help improve the data that’s going into those, a lookback analysis or performing back testing where you’re comparing predicted losses to what actually happened can help refine your model, improve the accuracy of what you’ve got. This another great area to utilize AI can help you in analyzing and modeling large volumes of data, maybe even factoring industry factors, macroeconomic factors, and it can certainly help in documenting the assumptions that are going into a model. So those are several best practices and internal controls related to CECL-related estimates. We’re going to switch gears, and Tim is going to talk through some best practices and internal controls for other estimates in the financial statements.

30:57 – Internal controls over accounting estimates

Tim Partridge [30:57]: I’ll actually be tagging on with what you were just talking about, Kathryn. So as we discuss, controls over accounting estimates helps to focus on some of those areas with a higher or elevated risk. And forming good controls over these estimates will help ensure that results are well documented, they’re repeatable, they’re supportable. Accounts receivable and the related allowance for doubtful accounts are a common example, is because they require significant judgment and directly impact those financial statement results. So, as you were just hearing, Kathryn describe under CECL, organizations must estimate you know expected credit losses using both you know the historical experience, current conditions, and a reasonable and supportable forecast. Because the model is forward-looking now, ongoing monitoring and strong controls is going to remain critical. Calculation controls should include items like automated calculation of customer aging buckets, automated calculation of tax with exception reports for unusual tax amounts, and automated calculation of you know customer discounts based on approved terms. The foundation of any allowance estimate is reliable data. Key data controls can include reconciling AR subledgers back to the general ledger, validating aging information, and ensuring all transactions are captured completely and accurately. Controls over the allowance model should include formal documentation, management approval, and periodic reviews of assumptions such as historical loss rates and forecast inputs. Independent recalculation or validation can also provide additional confidence in the estimate, particularly when those models become more complex. Like Kathryn was alluding to, management review controls are an important detective control. Reviewing allowance trends, write-offs, and the allowance as a percentage of AR can help identify unusual fluctuations that warrant further investigation. For example, a declining allowance despite deteriorating aging may signal increased risk. Segregation of duties should include controls that affect adjustments. Adding segregation between billing, cash application, and credit approval activities further strengthens the process. And finally, controls over write-offs should ensure appropriate approval, documentation, and independent oversight. And together, these controls can help make the estimate supportable, consistent, and repeatable. So, moving on here, another estimate area that you know we think is pretty valuable to look at is fixed asset useful lives. These estimates affect depreciation expense, carrying values, and impairment considerations, making them both important to the income statement but also to the balance sheet. Useful life estimates are influenced by factors such as expected usage, physical wear and tear, replacement cycles, and legal or contractual limitations. So the objective of the control environment here is to ensure that useful life estimates are reasonable, they’re consistently applied, and they’re updated when circumstances change. So strong segregation of duties helps support that objective by separating asset setup, useful life assignment, depreciation processing, and final review responsibilities should all be separate. Reliable data and system controls are equally important. Automated depreciation calculations, standardized useful life tables, and regular reconciliations between the fixed asset subledger and general ledger help maintain accuracy and completeness. While analytical reviews of depreciation expense can also identify unexpected trends, so useful life assignments should be governed by formal capitalization policies, and in establishing you know a standard range of asset classes and having that align with industry practices where that’s appropriate. Deviations from those standards should require documented support and management approval. Supporting documentation could include things like vendor specifications, engineering assessments, or historical operating experience. And management review should not stop once an asset’s in place and in service. Useful lives should be reassessed periodically, and whenever significant events occur, such as major repairs, changes in usage, or technological developments. Ongoing monitoring can help ensure that depreciation remains aligned with assets’ actual economic life. And lastly, here looking at fair value estimates, these often involve significant judgments, especially when you know observable market data is limited. And while level one and level two valuations rely on market information, those level three valuations are typically going to require greater management judgment, and therefore they warrant enhanced controls. An effective control framework should focus on four key areas. We want to look at data integrity, valuation methodology, assumptions, and management oversight. The starting point is ensuring that pricing data, investment holdings are complete and accurate. This can be accomplished through independent price verifications, reconciliation procedures, and regular reviews of market activity classifications. Valuation methodologies should be formally documented, consistently applied, and approved by management with evidence of that approval.

Tim Partridge [35:56]: Independent validation and periodic, you know, back testing can help confirm that valuation models remain appropriate, and then they’re reliable, especially when they include proper segregation of duties. And because those assumptions often drive valuation outcomes, organizations regularly evaluate key inputs such as discount rates, growth projections, and comparable company data, benchmarking assumptions to external sources and performing sensitivity analysis can also help identify potential valuation risks, and that also supports management’s conclusions. Ensuring accounting and finance functions have the proper level of input, visibility, and oversight of assumptions is necessary for accurate fair value estimates and the segregation of duties. When third-party pricing services or valuation specialists are used, management still retains that responsibility for the recorded fair value. So our controls should include vendor due diligence processes for investigating pricing differences and management review of those methodologies, assumptions, and significant valuation changes that are utilized by that third party. Ultimately, management review controls should focus on trend analysis, reasonableness for testing, fair value hierarchy classifications, and unusual fluctuations. A well-designed set of controls over fair value estimates helps support the reliability of financial reporting data and strengthens that confidence in the underlying accounting records. So next, let’s take a break here for our next polling question.

37:28 – Polling question: AI use in accounting

PYA Moderator [37:28]: Thank you, Tim. All right, our second question is: How frequently do you use AI in your accounting role? Never, once or twice a week, once a day, multiple times a day, or what’s AI? Remember, you must respond to the polling questions to receive CPE credit. You will have 30 seconds to answer.

Tim Partridge [37:50]: Wonder how many what’s AIs we’re going to get. A

Kathryn Good [37:56]: lot of nevers and a lot of once or twice a weeks.

Tim Partridge [38:03]: You have some frequent flyers.

PYA Moderator [38:05]: All right, thank you for participating in our poll. Back to the presenters.

38:10 – ASC 606 revenue recognition

Kathryn Good [38:10]: All right, I’d be curious if we polled everyone six months from now how those answers might change. We’re now going to cover our final financial reporting topic, ASC 606 This on revenue recognition in contracts with customers. We’ll do a quick refresher to start off with. ASC 606 introduced a five-step framework in recognizing revenue, so we’ll talk through those quickly on this slide, the first step being identifying a contract with the customer. Yes or no? Do we have a real deal or agreement with the customer? It can be verbal or in writing, where both sides are committing to holding up their end of things. There’s a clear rights, clear payment terms, clear business purpose. From there, as the seller, you would identify the performance obligations that you have to the customer. What are the distinct goods and services that you’ve promised to provide to them? Determining the transaction price is determining how much you expect to be paid in total, and there’s a few different kind of components within that. You’ve got to think about variable considerations like rebates, bonuses, discounts, things like that. There could be a financing component to the agreement. So some ways to think about this would be, you know, is this agreement acting as a loan between me and my customer, based on the timing of when cash is changing hands, and is someone at some type of an advantage based on the performance or the payment timing. So, if your customer is paying you years ahead of schedule or years after you’ve provided the service, there’s likely some type of financing. Component, and you would want to separate out the interest expense or the interest income, depending on the circumstances, from the actual sales price of what you’re providing to the customer. Your next step would be to allocate that transaction price across the different obligations. So you’d want to base that on the standalone selling price for each product or service that you’re providing to the customer, because revenue could be recognized differently for each of those. And then, lastly, as you’re satisfying those performance obligations to the customer, you would recognize revenue. It could all be recorded at one point in time when control of the product or service is transferring to the customer, or revenue could be recognized over time, maybe in a subscription type of an arrangement, or based on specific criteria like a percentage of completion. Continuing on our brief refresher, 606 also introduced some concepts and terminology of contract assets and contract liabilities. Contract assets being instances where there’s revenue that’s been earned, you’ve satisfied a performance obligation, but you’re not yet able to bill it to your customer. So an example of this might be that you have an agreement to provide services to a customer for a full year, but under that agreement, you can only invoice your customer on a quarterly basis. So there’s a timing difference between when you’ve actually satisfied a performance obligation versus when you can send an invoice to the customer, and so a contract asset is kind of capturing that revenue. A contract liability, on the other hand, means that your customer has paid you in advance, but you have not yet satisfied a performance obligation to them. So the same thing as deferred revenue or a prepayment. So that’s not necessarily a new accounting concept, but a new terminology that’s with us now. Some of the more common issues that we’ve seen lessons learned after adopting the revenue recognition standard would be having some errors in identifying the distinct performance obligation in a contract. So a lot of contracts have multiple performance obligations. So maybe you’ve offered to sell your customer a software license. That’s one performance obligation. With a year of free support, that’s a second performance obligation. And revenue is going to be recognized differently and at different times for each of those performance obligations. So a leading practice here would be to establish a thorough contract review process on the front end to help in identifying all of the different performance obligations, and having good interaction between sales and the accounting team to help with that process as well. AI can help in performing this contract review to piece out all of the distinct performance obligations.

Kathryn Good [43:01]: I don’t even think it’s necessary to upload every single customer contract to an AI model, but if you have templates or kind of categories of contracts, maybe uploading each of those to determine what the different performance obligations are could be helpful. We also saw some instances where variable consideration elements, like those rebates or returns or performance bonuses, were either overlooked, or there were errors in estimating them, and that meant that organizations ran the risk of recognizing too much revenue on upfront, and then having to go back and reverse it later when those estimates changed, as you can imagine, that could potentially lead some to some volatility in financial reporting, which is not ideal for anybody. So, sticking with that conservatism that we rely on as part of GAAP financial reporting is to recognize only the portion of revenue that’s not likely to be reversed later, and then for these uncertainties, having some type of contra revenue liability on the balance sheet until that uncertainty is resolved, and then at that point taking those amounts into income. This it’s going to apply to organizations in some industries more than others. So if you do have to make a lot of these types of estimates, you would want to document all of the assumptions and the methodology that’s used in it. AI can help in analyzing the data to help in putting together an estimate in the first place and in documenting those assumptions. And the last kind of frequently seen issue related to revenue recognition is the overall timing of when revenue is recognized. So some organizations were recognizing revenue too early before control of a product or service passed to a customer, or too late by failing to update revenue timing when. Contract was changed with maybe an extension, an add-on, or a scope change, and not having any type of trigger to reassess the timing of when revenue was recognized led to booking revenue in the wrong period, which could potentially lead to an audit adjustment. And again, having that volatility in financial reporting, so having some type of control and process to trigger that reassessment upon signing any type of contract modification, having that alignment between sales, accounting, and billing can be really helpful, and helping to make sure that revenue is only recorded when performance obligations have been satisfied and recorded in the appropriate period. That wraps up our conversation on financial reporting standards, and then Tim is going to wrap us up with talking about SECURE 2.0.

45:54 – SECURE 2.0 implementation considerations

Tim Partridge [45:54]: Boy, is there a lot to cover with SECURE 2.0 We will not get to all the changes that came about from this act, but we’ll cover some of the larger items we feel were and still are a struggle for implementation, including those optional provisions that came about. So the SECURE 2.0 Act was signed into law in December 2022 with effective dates ranging from December 31, 2023 so 2024, all the way through January 1, 2026 Keeping track of the various effective dates and those extensions has been a real challenge. Not to mention the time and effort it takes to implement the changes needed, like software updates and personnel training. Reference to the act itself can be found at the bottom of this slide. So starting this off here with auto enrollment, this was effective plan years beginning after December 31 2024 Employers setting up new 401k plans must automatically enroll eligible employees at a default contribution rate of at least 3% scaling up 1% annually to arrive at a final rate of at least 10% but not more than 15% Exemptions do apply for new businesses, small businesses with 10 or fewer employees, church plans, and governmental plans, and so challenges for implementing here would be the intended boost to savings. It creates an onboarding burden, really, for those plan administrators. Employees can now opt out of automatic enrollment within 90 days of that first payroll deduction, which can lead to refunds. In industries with high turnover, this causes an administrative headache, resulting in sometimes hundreds of accounts with very small dollar amounts that sponsors must manage and process. Overall, some businesses are still facing higher administrative oversight, as well as increased pushback or confusion from newly onboarded employees. So, how can you kind of overcome some of those challenges? We can establish a 360-degree integration by moving away from that kind of traditional 180-degree integration where data only flows from your payroll systems down to the record keeper. By establishing a 360-degree API integration with your provider, you can alternatively also engage like a third-party middleware to do this, but that’s going to allow for changes such as opt-outs or auto escalations to then flow instantly between your payroll systems and back to the record keeper. Another tool that we’ve seen that’s been helpful, I think, is that you know, excuse me, utilizing you know low-dollar incentives like a $25 gift card to encourage employees to either make an active or definitive election at the time that they’re onboarding, either opting in or opting out, will allow them to basically keep from slipping into that auto enrollment queue by default. So next, let’s touch on the super catch-up contribution provisions, which were effective for January 1, 2025 Individuals attaining ages 60 to 63 during those calendar years are able to make catch-up contributions to eligible retirement plans of up to $11,250. This in place of their normal catch-up limit, which was $8,000 for this year, 2026. What we found to be the challenge in implementing this the administrative time and ensuring that the right limits are utilized for participants, the provision requires payroll systems to establish those new limit codes, all while tracking participants’ exact birth dates. System tracking must be precise, as the higher limit has to be get you know discontinued there that calendar year that the participant then becomes 64. So systems have to scale as the participant you know that participant limit back down to $8,000 and that’s the max on the January 1 of the calendar year that they turn 64. So ensuring that payroll triggers are set up correctly to you know prevent that excess deferral be key in maintaining compliance with the standard. So how can you overcome those challenges? Well, you can ensure that your systems will identify participants who turn 60 to 63 at any point during the calendar year by either utilizing some software or some additional reporting. So, we would recommend you instruct your TPA or your payroll vendor to create multiple tiers, which will kind of set up different architecture for you. Like a tier one is. Going to be everybody under the age of 50, which are going to be at the standard limit for catch up. You know, you’re going to basically have tier two, which is you know ages 50 to 59 and then 64 plus.

Tim Partridge [50:09]: Those are the standard catch up limits, and then you basically have an you know tier three for ages 60 to 63 which is this new super catch up limit. So that would help ensure that contributions are appropriate by those different age groups. This will also allow that challenge to kind of be overcome potentially by some AI. You know what we could recommend here for an AI usage case would be by tasking an AI agent inside of an Excel spreadsheet to monitor or notify or code by age tier for all of your participants. That kind of can help you avoid a lot of these issues with the different age ranges that’ll be mandatory to track for this super catch up. Moving on here, Roth catch up. This became effective for January 1, 2026 Even though the full IRS regs are not technically in place until 2027 so a mandatory Roth design was added to SECURE 2.0 if that option did not already exist inside of your plans. As of 2026, if you earn more than $150,000 in a prior calendar year, all catch-up contributions to a workplace plan at age 50 or older will need to be made to a Roth account in after-tax dollars. Individuals earning 150 or less, adjusted for inflation going forward, they’ll be exempt from that Roth requirement. So the challenge here in implementing, you know, it’s kind of almost the same as the super catch-up that we just talked about. Employers must coordinate with third-party payroll or record keepers to accurately track FICA wages across prior years, especially for employees transferring between related entities or related parties, or navigating irregular compensation structures, this can prove to be a large challenge. So sponsors must establish systems to either automatically spill over excess pre-tax contributions into that Roth bucket, or enforce separate election methods without causing contribution limits to be exceeded. So, how can you kind of overcome some of those challenges? Well, instead of requiring high earners to complete an affirmative election form every year, sponsors could implement a deemed election policy. This basically an administrative shortcut that lets the plan automatically treat certain catch-up contributions as Roth without requiring each affected employee to then actively elect that, so adopting a deemed election is also a mandatory prerequisite for using the streamlined IRS correction method if a system ever were to ever occur. So moving on to long-term part-time employees here, this effective for plan years beginning January 1, 2025. So anyone who has worked at least 500 hours per the last two consecutive years will be eligible to join an employer’s retirement plan. So this down from the previous three-year threshold. The rule only applies if a plan uses service-based or minimum hours of service exclusions. The challenge in implementing here focuses really on managing complex administrative rule changes. Plan sponsors must track employees who complete at least 500 hours of service in two consecutive 12-month periods, and payroll and HR systems must accurately monitor part-time, seasonal, or variable hour schedules, which can be difficult without automated tracking. Additionally, under IRS guidance, an employee who achieves that long-term part-time status retains that status even if their hours increase above that threshold or 1000 hours in a subsequent year. This basically forcing permanent dual tracking of eligibility investing for these specific employees as sponsors are now required to provide matching or non-elective employer contributions to those long-term part-time employees. This, you know, potentially going to be complicating that plan design. So, what can you do to overcome some of those challenges? Well, really, by mitigating administrative burdens, employers can choose to eliminate hours-based service requirements altogether by lowering a standard plan eligibility to 500 hours or fewer across 12 months for all employees, sponsors avoid the complexities of classifying, tracking, or administering separate rules for those long-term part-time employees. So, looking here at the emergency and hardship provisions, these are effective for December 31, 2023 If your workplace retirement plan allows for it, and if you meet certain other requirements, you may be able to take qualified disaster recovery distributions from your 401k SECURE 2.0 expanded those distribution options and favorable tax treatment for up to $22,000 of qualified disaster recovery distributions from eligible plans. The associated income tax can also be spread over subsequent years, and you may have that option to recontribute funds.

Tim Partridge [54:43]: There are provisions for also for emergency personal expenses withdrawals of up to $1,000 per calendar year, and that’s an optional provision. Also, if permitted, you could apply for a hardship withdrawal for personal situations, including domestic abuse. So. Challenge for those emergency provisions really becomes defining and administrating, you know, administering what constitutes an emergency. So you have to come up with a definition. So something similar to a qualified disaster is a major disaster declared by the president, listed by FEMA. But then, how do you determine if an employee is in an affected disaster area? Well, provisions allow for self-certification by participants, and you know that is both in determining if a residence is in an affected area as well as if there are sustaining economic losses. But sponsors really have to decide how much validation, if any, to require. The risk being that if they’re too lenient, you might end up with plan leakage. If they’re too strict, you’re going to have inconsistencies with application or the standard, or with current laws. Another challenge is tracking repayment windows. Participants cannot take another emergency withdrawal within three years unless they’ve repaid for other withdrawals. The only withdrawal is you know one withdrawal is allowed per year. So you know this becomes an optional provision that sponsors must decide also whether you know they’re going to document and adopt that within their policy and plan. So how can you overcome some challenges here? Well, you can decide whether or not to adopt that $1,000 emergency withdrawal provision. Document that definitive you know definition here of immediate financial need. What does that look like? And establish a rule. You can also allow for self-certification, which is what we would recommend, or limited validation. You can ensure that your record keeper has systems in place to track timeframes and limits. And then our challenges that we’ve been seeing for those hardship, you know, withdrawals are that you know domestic abuse and that certification around that is a real sensitive topic. The administration requires coordination between multiple different departments. Risk of audit scrutiny also exists with the IRS and DOL. So ensuring hardship criteria are applied consistently is a real concern if you want to avoid discrimination or claims of unfair treatment. Easier access increases also the likelihood of reduced retirement balances and higher participant, you know, reliance on those withdrawals, removal of the requirement to take plan loans first, which already is relaxed but reinforced in practice, and then reduced documentation, which could equal fiduciary risk. So sponsors may rely on employee certification instead of collecting proof, but must ensure that process is being consistently applied and it’s prudent. So, how can you overcome some challenges as it relates to those hardship withdrawals? Well, you can work with your plan record keeper to ensure that you’ve got consistent application for self-certification. You can also develop policies on how to handle those sensitive HR issues for you know your domestic abuse requests. Consistent documentation will also help avoid any negative impacts from the IRS or DOL inquiries here. So we’re about to run out of time, so we’re going to skip over I think this last slide here and just move to our our next polling question.

57:59 – Polling question: most challenging standard

PYA Moderator [57:59]: All right. Our final polling question is: Which accounting standard have you found most cumbersome to implement? SECURE 2.0, ASC 842 leases, ASC 606 revenue recognition, or ASC 326 CECL. Remember, you must respond to the polling questions to receive CPE credit. You will have 30 seconds to answer. Right. Thank you for participating in our poll. Back to the presenters.

Kathryn Good [58:47]: Tim, for recording purposes, I know you had one last slide. If you wanted to cover that, and then anyone that needs to drop can move on to the rest of their day. We think thank you for attending, and we would have that last slide covered in the recording that’ll get sent to you.

PYA Moderator [59:03]: If anyone does need, I’m sorry, Tim. If anyone does need to drop, please remember to fill out the survey before you do, and we will. Your your CPE certificate is downloadable via the continuing education pane. So, Tim, back to you.

59:21 – RMDs and student loan matching

Tim Partridge [59:21]: All right. So let’s take a look at RMDs. This had effective date for taxable years beginning after December 31 2023 The age for mandatory withdrawals has been pushed back to 73 Additionally, RMDs are no longer required from Roth accounts and employer retirement plans. So in 2023 a steep penalty for failing to take an RMD decrease to 25% of the RMD amount not taken from previously what was 50% The penalty will be reduced to 10% for IRA owners if the account owner withdraws the RMD amount previously not taken and submits a corrected tax return timely within two years. Also, good to know is that SECURE 2.0 pushes the age at which RMDs must start to 75, beginning in 2023. So for those born in 1960 or later, so challenge for implementation here is that plan sponsors must track the phased increase in RMD age. Those born in 1951 through 59, the age is 73, and for those born in 1960 or later, it’s 75. Now, Roth account exclusions-you know-the designated Roth accounts and employer plans are no longer subject to RMDs during the participant’s lifetime. So, sponsors and record keepers must now segregate and separately calculate RMDs from non-Roth balances, which complicates payroll and withdrawal systems. So, how can you overcome some of those challenges here? Well, you can verify the absolute accuracy of participant birth dates in the HR and payroll systems to prevent automatic RMD calculations from triggering at the wrong age. So you can map how participant birth dates flow from payroll system to TPAs or record keepers. This might also be a great opportunity for AI utilizing an Excel agent like we had previously discussed, and so lastly, here we’ll talk about student loan match effective dates were planned years beginning after December 31 2023 Employers can treat qualified student loan payments as elective deferrals, allowing them to make matching contributions to 401k accounts. Employees struggling to repay student loans often forego those contributions to retirement plans, so they’re missing out on not only long-term saving opportunities but also those potential employer matching contributions. To address this, provisions allow employers to make matching contributions based on employees’ qualified student loan payments, even if the employee does not contribute directly to the plan. This provision allows borrowers and cosigners as well making payments on behalf of themselves or their dependents to qualified higher education loans, provided that they certify their loan payments annually. So the challenge in implementing here is that employers are tasked with verifying that employee payments actually qualify, while IRS guidance permits self-certification tracking loan qualifications, ensuring you know that we have eligible debt, and vetting spouse or dependent loan criteria remains an administrative burden. Third-party administrators as well, you know, and record keepers are still struggling to align to match with this existing payroll system. So, how you can overcome some of the challenges here. Well, we recommend not requesting or manually collecting physical loan statements or bank receipts from participants for this program. One solution is to build an electronic affirmation system. So by establishing an electronic portal where employees can simply just check a box and self-certify their qualified student loan payments, that can alleviate some of the significant risks here, and then keeping any self-certification fields minimal by restricting data collection to the loan amount, payment date, employee legal name, and then the name of the financial lender. And that kind of wraps up what we had to talk about for SECURE 2.0 Kathryn, you want to take us home?

1:03:01 – Closing takeaways and next installment

Kathryn Good [1:03:01]: Happy to. Thank you for attending today’s session. I know we have a bit more housekeeping information to share, but hopefully this was valuable content and helpful to you and your organization. Our next installment of the Bottom Line is coming up in September. This session is on forensic accounting, and hope you’ll hope to see you there.

PYA Moderator [1:03:27]: All right, thanks to our presenters, Kathryn and Tim. Later today, you’ll receive an email with a link to access their information, copies of the slides, and a recording of the webinar. Please visit our website at pyapc.com to learn more about the full range of services offered by PYA. While at our website, you may register for other PYA webinars and subscribe to receive PYA insights. Please remember to stay on the line once the webinar disconnects to complete the short surveys for CPE credit. And once complete, your CPE certificate is downloadable via the CPE pane. Thanks to everyone who stayed on with us, and on behalf of PYA, have a great rest of your day.