Good day and welcome to today's Graham Holdings Company 2022 Virtual Investor Day. Today's conference is being recorded. At this time, I would like to turn the meeting over to Timothy J. O'Shaughnessy, Chief Executive Officer. Please go ahead, sir.
Thank you very much, and thank you to everyone for joining. Welcome to our 2022 Investor Day. I'm Tim O'Shaughnessy, the CEO of Graham Holdings, and our schedule of events brings you two speakers today. I will start with an update on financial results and operations, and then walk you through our other businesses segment in a bit more detail with a particular focus on Framebridge. Jake Maas, our Executive Vice President who leads planning and development, will then provide an update on our automotive business. After the presentation, we will have a Q&A session for as long as time permits. For those that sent any questions in advance, thank you. We will also be answering questions submitted in real-time through the chat in the presentation portal. As 2022 has progressed, the COVID-induced fog on earnings at the company has begun to lift.
Results have improved both compared to 2021 and as the year has progressed. Although the lingering effects of lower COVID-related enrollments at Kaplan International will be with us until the enrollment cycle next fall. Nonetheless, we feel closer to normalized earnings power than we have since 2019. 2022 revenue is up 23% year to date over 2021. Many factors are causing this growth. Continued COVID recovery at Kaplan International and Clyde's, underlying growth at several other business units, political spend at Graham Media Group, and acquisition activity. Adjusted operating income is up 41% for similar reasons. Operating cash flow and free cash flow each increased in 2022 over 2021. After several purchases of property in support of business operations in 2021, free cash flow was impacted by nothing similar year to date.
As for the rest of the year, with the continued recovery at Kaplan International and the influx of political advertising, we expect Q4 operating cash flow to be meaningfully improved over 2021. You'll notice on our free cash flow chart, we add back our pension expense. Many of you may recall the unique position of our pension trust. Our substantial overfunding means the likelihood of near-term company contributions to the trust is very small. While real benefits are being accrued by active pensioners, the existing assets of the trust have historically supported those liabilities, and we anticipate that being true for the foreseeable future. Importantly, these pension expenses flow through our income statement but do not require cash from the company to fulfill our obligations to pensioners. As such, when we evaluate cash flow generation at Graham Holdings, we tend to exclude our pension expense.
This is notable as we continue to find ways to use pension benefits within company operations. We expect our pension expenses to increase over time, but the lack of need to draw on the corporate treasury to remain constant. We remain very comfortable with our balance sheet, although it has changed somewhat over the course of the year. Marketable securities are down meaningfully year to date, primarily due to a decrease in the mark-to-market value of our holdings. Debt has also increased, largely due to new debt that Graham-Ourisman automotive used to help fund the purchase of two auto dealerships this past summer. Despite these changes, as of Q3, cash and marketable securities continue to outstrip total debt. Our approach to capital allocation remains unchanged in terms of philosophy. In the near term, it will likely be similar to 2022.
What I'm about to say should sound familiar. It's nearly word for word from my commentary on capital allocation at last year's event. In addition to maintaining our dividend, our primary near-term capital allocation priorities are: continuing to fund promising internal initiatives such as Framebridge. Repurchasing shares. We continue to believe there is a substantial delta between share price and our view of intrinsic value. If that gap persists, we are likely to be a buyer. If the gap widens, this pace could accelerate. Evaluating bolt-on opportunities, primarily at Kaplan and Graham Healthcare Group, as well as on a case-by-case basis in our other operations. Management's primary focus for 2023 will be continuing to work with our existing operations to grow our earnings power.
Our belief is the next $100 million in operating income at the company is more likely to come from our existing operations than from acquisitions. Let's now move to operations update. Kaplan has begun to emerge from COVID and is on a very good trajectory. As a reminder, Kaplan is comprised of two primary business units, Kaplan International and Kaplan North America, or KNA. At KNA, there are two segments, higher education and supplemental education. The resilience of Kaplan International throughout COVID and the subsequent stop-start recovery has been nothing short of amazing. While the U.S. has been largely open in 2022, this has not been the case worldwide. Australia opened its borders for normal travel only in late February of this year. China has and continues to severely limit travel in and out of the country.
New Zealand opened its borders in late July of this year. Russia and Ukraine were largely shut down as of February 2022. As a global education company, we have no choice but to navigate these challenges directly. Students cannot put a pause on their life goals and dreams. We need to adapt with them. The team has managed to deliver for students throughout 2022, which has been good for shareholders. Financially, the first nine months of the year have shown continued progress. Revenue is up 6% year-to-date. Adjusted operating income is up 22% to $70 million. The operations at Kaplan International are largely headed in the right direction, some with quite a bit of wind in their sails.
While enrollments will not see full recovery in some programs until the fall of 2023, we are optimistic that many of the reductions in the cost base implemented during COVID will allow us to achieve higher levels of profitability due to expanded margins, even before reaching that full recovery. Year-to-date, Kaplan International's operating income is at near record levels. Our languages business accrued substantial losses from 2020 through Q1 of 2022. While travel restrictions have continued to limit language students from many Asian countries, I'm pleased to report that Kaplan's languages group achieved profitability in Q3 for the first time since 2019. While the current profits are modest, we believe 2023 should represent another step forward for languages. Kaplan North America has seen mixed results in 2022. On the higher education side of the business, progress continues.
As a reminder, our higher education segment provides services to universities to help them create, manage, and grow educational programs. In addition to Purdue Global, in recent years, Kaplan has added programs with Wake Forest University, Lynn University, and several other colleges and universities. These additional partners and programs allow us to continue to increase the breadth of offerings we can provide to each institution. A program we build for one university has a reasonable chance of also solving the needs of another partner. Each new partner tends to have a period of operating losses as we build the programs, but within a few years of launch, we expect them to be profitable. Our work at Purdue Global continues to be the largest driver of financial results in the higher education segment.
Purdue Global has continued to improve student outcomes, increase the quality of the student cohorts, and grow the student census slightly above 2021 to 35,400, despite a low unemployment economy in which working adults tend to defer their higher education goals. These efforts should increase the likelihood of collecting our full fee under the agreement that governs our relationship with Purdue. We continue to support Purdue's long-term approach to growth at Purdue Global, helping them build new student support and advising capabilities, additional programs, and improved academic outreach. Purdue continues to set the course for Purdue Global, we are pleased to help them continue to improve both academic and financial outcomes. At supplemental education, the results are more mixed. You've likely seen in the news that requirements for tests, such as the SAT and ACT, have become optional by many colleges.
We are not immune to the reduction in test-takers, and our traditional test prep business has suffered meaningful declines. On the other end of the spectrum, our graduate and professional certification programs remain leaders in their sectors. Student interest in these areas is below average, albeit stable. Like the Purdue Global business, when the unemployment rate is low, students defer the need for additional credentials. With the rate hovering below 4% for much of 2022, enrollments are subdued. While we've seen this cyclicality before, it is never fun to live in the downside of the cycle. As in the past, we are confident of our ability to preserve earnings power in lower demand periods while growing it when the sun shines a bit brighter. Our second major engine, Graham Media Group, a collection of local television and digital businesses, continues to show solid results.
Aided by increased political advertising spend, year-to-date revenue has increased 6%, while adjusted operating income has grown 20%. Income at Graham Media will certainly be down in 2023. With an off year in our election cycle, it is unrealistic to expect to make up for declines in political advertising spend. In 2022, increased spending in Michigan offset lower levels of spending in Florida as compared to previous election cycles. Earlier this year, Emily Barr retired, and Catherine Badalamente took over as CEO. Catherine has picked up where Emily left off. She not only understands our company and our culture, but she also understands the trends of the business and where opportunities for new revenue lie. Catherine and her team are eagerly chasing after new opportunities and have made it a strategic priority to create direct relationships with our viewers, readers, and listeners.
Examples include our Insiders program, which has surpassed 485,000 members, streaming via our YouTube channels and NextGen apps, and creating the leading websites for local news in the vast majority of our markets. The value of local news and content to communities remains high. The public gives local news some of the highest trust scores when compared to other news sources. Catherine and her team are doing a nice job of balancing the heritage of local TV news with the evolving needs of today's audiences. The world of local broadcast will look different 10 years from now, and Graham Media Group is on the forefront of figuring out this change. The manufacturing segment has bounced back from the pandemic-related challenges of prior years. While revenue is up 3% year-to-date, operating income is up 41%.
Among the four businesses, all but Dekko have returned to a pre-COVID operating cadence. At Dekko, one of its largest end markets is commercial office space. Last year, we deemed Dekko as impaired due to the substantially reduced demand for capital expenditures in the office market. 2022 confirmed this belief. While Dekko remains profitable, it is at a reduced level and will be until supply and demand equilibrium is reached in the commercial office space market. A quick note on revenue for this segment. Wood prices have historically had modest volatility. In recent years, the price charts compare well with rollercoaster drawings. Hoover Treated Wood Products is our largest business within the manufacturing segment, and a large portion of its revenue fluctuates with this volatility.
As the price of wood is largely a pass-through cost, our expenses fluctuate in line, meaning operating income, in general, has been minimally impacted by revenue declines or increases tied to wood prices. This largely explains how revenue for the segment can be up 3% while operating income can rise 41%. We're pleased with our manufacturing operations. They have proven to be steady providers of cash for Graham Holdings with strong market positions in their respective sectors. Graham Healthcare Group has become a very important business at Graham Holdings. The co-CEOs, David Curtis and Justin DeWitte, have grown the business organically via acquisition and with joint ventures. Through Q3, revenue at the healthcare business has grown 44% over 2021, with operating income down modestly and is on pace to earn a somewhat reduced amount of joint venture income in 2022 compared to 2021.
Where is that revenue growth coming from, and why isn't income following along? Over the past few years, GHG has entered adjacent service lines that focus primarily on in-home or outpatient care. While the home health and hospice business continues to grow, services such as CSI Pharmacy, our in-home specialty pharmacy infusion service, Skin Clique, which provides in-home dermatology treatments such as Botox, and Surpass, our outpatient autism therapy service, have begun to meaningfully generate additional revenue to Graham Healthcare Group. We are very good at training, managing, and providing operational support for clinical field staff. That, combined with an ever-present focus on our quality of care, can lead to better patient outcomes and above-average business outcomes. We think these go hand in hand, and you cannot have one without the other.
Our expanded view of our business from one with a core competency of home health and hospice to one with a broader core competency of in-home and outpatient care has unsurprisingly led to changes in our revenue profile. Most notably, CSI Pharmacy has grown beyond our expectations. At its core, CSI Pharmacy provides IVIG therapies to patients who can benefit from the administration of treatment through in-home infusions. The skills and capabilities of specialty pharmacists when combined with our expertise with field clinicians, has led to a market offering that meets the needs of patients and doctors. CSI has expanded its geographic footprint in 2022 and plans additional expansion in 2023. Through Q3 of 2022, home health and hospice was 60% of revenue, CSI Pharmacy was 31%, and other healthcare services were 9%.
We expect all three of these categories to continue to grow, but both the pace and how the slices are divided are future opportunity dependent. While revenue growth is usually a better sign than not, if income does not follow, a reasonable question can be asked about whether we've just filled up on empty calories. I can assure you the senior leadership of the company does not believe this is the case, and we will likely prove that in the next few years. Year-to-date results have been in line with our internal expectations. Three things occurred in 2022 that tamped operating income growth, none of which should meaningfully repeat in 2023. Investment in senior leadership capabilities at Graham Healthcare Group. Heading into the year, management believed it could and should bolster its capabilities to position the business to become a larger operation. We built ahead of anticipated growth.
Transaction-related costs. The first nine months of 2022 were unusually active, with five transactions occurring at the organization. Integration costs. Various platform severance and other restructuring costs at acquired businesses impacted initial earnings at those acquired operations. With these items largely behind us, Graham Healthcare Group's operating results should continue to improve into 2023. We feel great about Graham Healthcare Group and think you should too. In-home and outpatient care reduces costs to the system and can provide better outcomes and better patient satisfaction. Macro demographic trends are positive for our service lines and show that we're operating where future growth is possible. While not by master plan, many of our other businesses have naturally fallen into one of three categories: media, retail, or specialty. We are going to begin breaking out revenue within our other businesses category to better inform and track progress.
Let's start with the financial specifics. Year to date, segment revenue has increased 54%, while the adjusted operating loss has increased by 19%. Media revenue has grown 34%, retail revenue by 64%, and specialty revenue by 69%. The reasons for this growth include ownership of Leaf Group for the entire 2022 period, pandemic recovery, and organic growth. Investments in the category have been substantial and are expected to be reduced in 2023. As referenced previously, we view 2022 as the peak investment level for this group of businesses. Losses have primarily come from Leaf Group and Framebridge. Both should improve meaningfully in 2023. Leaf's losses were largely unplanned. Substantial degradation in the business occurred due to both a reduction in purchasing behavior post-pandemic and reduced advertising spend at Leaf Media as advertiser concern about macroeconomic conditions surfaced.
I'd like to spend a bit more time now discussing Framebridge. We receive investor questions around Framebridge at a rate that surpasses its size. I suspect that speaks to the potential that others see in the business. We agree, which is what led to the purchase of the company in 2020. We continue to believe in that potential today. This seems like an appropriate time to give a progress report on Framebridge and share a little more about how we think about that potential and how we plan to achieve it. What is Framebridge today? Framebridge is disrupting the custom framing industry by replacing a cumbersome experience with a delightful one. We are building a beloved brand associated with celebrating the best moments in life. We operate online and in our 17 retail stores currently in D.C., New York, Boston, Atlanta, Philadelphia, and Chicago.
We operate two onshore manufacturing facilities that allow us to deliver a high quality custom product at an affordable price with a quick turnaround. We offer an easier, faster, and more inspiring experience at a better value, we believe we are well-positioned to be the clear category leader. What does the market for custom framing look like? Framebridge primarily competes against a fragmented market of local retailers as well as big box retailers such as Michaels and Hobby Lobby that offer custom framing within their stores. While it is hard to precisely size the existing market, we believe current domestic annual sales to be in the neighborhood of $5 billion. Over time, we believe Framebridge can both take market and make market.
Our ability to gain share in the existing market is driven by both the value proposition outlined above as well as the decline in the number of independent local retailers. An industry study estimates that these shops have closed at a negative 4.3% CAGR from 2017 to 2022 due to COVID, increased competition, and most significantly, the retirement of local store proprietors. Framebridge can also make market by bringing a new class of clients that has previously never custom framed. Providing a 21st-century experience combined with more affordable pricing expands the potential customer base. Based on an internal survey, 33% of Framebridge customers had never custom framed before. How much of this market can Framebridge take? We don't know yet. However, we are investing and operating with a belief that we are very early in our growth curve.
How will Framebridge continue to grow? We expect both online and retail can become much larger than they are today. Growth will be driven by the following. First, continued investment in the brand, which should entrench Framebridge as the default solution for custom framing in the mind of the consumer. Second, expansion of the types and sizes of items we frame. Third, continued expansion of the retail footprint. I'd like to focus on the last of these, retail. Retail offers several primary customer benefits that the online business cannot provide, such as giving customers who are uncomfortable packaging and shipping their own art an option to bring their art to a store or to receive in-person design consultation from one of our associates. The retail footprint has expanded from two stores in one market in 2020 to 17 stores in six markets today.
If our model is correct, there will be many more. Our existing markets can handle more stores. We believe there are dozens of additional markets for Framebridge to operate retail locations. It's hard to predict how many stores this ultimately implies. It is certainly many more than exist today. We open stores in locations where we believe we can achieve a sub-24-month payback on our capital expenditures and pre-opening costs. Over time, we expect to be able to expand with adjacent product offerings in the store, which should further improve the economic profile. However, this is not likely to meaningfully happen until we have a larger store footprint. With the economic profile as described, why not go faster? Well, let me start by saying we're trying.
We will continue to pursue meaningful additional store expansion in the coming years, but there are two main factors that govern the retail expansion of Framebridge. First, we want to be excellent at finding the right location with the correct type of space. It's too important to maintain a consistent model without having to navigate multiple store types or different types of local marketing activation campaigns. We also care an awful lot about our lease terms and will not impair the economic model to chase an additional store. Second, and more significantly, we need to match our demand with our production capabilities.
Because each new store location leverages our existing manufacturing footprint, we must be certain that our framing operations can appropriately absorb additional volume, that staff are properly hired and trained to deliver high-quality products at that increased volume, and that our overall internal operations and supply chains can deliver against that growth. In other words, we'd rather grow half as fast the right way than twice as fast the wrong way. Stores and manufacturing plants each require substantial timelines to open, so it is paramount we are thoughtful in our planning in both of these areas to avoid costly mistakes. 2022 is likely to be the peak investment year at Framebridge.
Escalations in labor, materials, and shipping costs led to sharp increases in production costs. We will have greater losses in 2022 than initial internal forecasts due to these costs and the corresponding conscious choice to slow down growth until we could reset unit economics through price and production efficiency improvements. As we reach year-end, unit economics have stabilized and there is a clear line of sight on further improvements. Increased scale and improved production capabilities should lead to increased gross margins in 2023 and beyond. When does Framebridge generate free cash flow? Competing forces govern this question. The more heavily we build out new retail locations, the more we may ultimately delay positive free cash flow as earnings from existing units are reinvested back into the business to build more stores.
In most scenarios, at somewhere around two times its current business size, the business should be in a position to generate free cash flow. To close out our progress report, we remain both excited and optimistic about the future at Framebridge. Susan Tynan and her team are focused on delivering the vision and building a great business for shareholders. Now, I'll turn it over to Jake for some color on our automotive business.
Good afternoon. I'm Jake Maas. I serve as an Executive Vice President at Graham Holdings. Today, I'm pleased to provide an overview of our growing automotive segment, which over the last few years has become an important business within Graham Holdings and a key new income stream for the company. Currently, Graham Holdings automotive segment consists of six dealerships in the Washington, D.C. metropolitan area. We entered the space in 2019 via our acquisition of two dealerships, a Honda dealership in Tysons, Virginia, and a Lexus dealership in Rockville, Maryland. In 2020, we were awarded a Jeep add point and launched a de novo dealership in Bethesda, Maryland. In December 2021, we acquired a Ford retail and commercial dealership in Manassas, Virginia. Most recently, in July 2022, we acquired a Toyota dealership and a Chrysler Dodge Jeep Ram, or CDJR dealership in Woodbridge, Virginia.
In addition to these six dealerships, we also operate CarCare To Go, a valet automotive repair startup that serves the Washington, D.C. area. I would also note that we own the underlying real estate at our Honda location in Tysons, Virginia, and the Toyota and CDJR facilities in Woodbridge, Virginia. As illustrated in this timeline, Graham Holdings has grown our automotive segment in two ways. First, by acquiring automotive dealerships via M&A. M&A transactions in this space are typically structured in a similar fashion in which the acquirer purchases the assets of a dealership, which include the new and used car inventory, the parts and inventory of the service department, and what is known as the blue sky value of the dealership, which is essentially the value of the franchise. All M&A transactions must be approved by the manufacturer.
Some transactions also include the purchase of the underlying real estate of the dealership as well. In addition to M&A, a dealership can also be established by being awarded a new franchise add point. Essentially, this occurs when a manufacturer decides to expand their franchise network by adding a new dealership location. The franchise is awarded to a chosen operator, with the operator bearing the start-up costs associated with establishing a new showroom and servicing center. In addition, these add point dealerships require ramp-up time to build up a customer base sufficient to operate the dealership profitably. For the right manufacturers in the right locations, we think this can also be an attractive investment and path to obtaining new dealerships. Going forward, we are open to adding dealerships to our platform via both M&A and by earning additional add points for manufacturers.
Graham Holdings automotive segment is operated in partnership with Chris Ourisman and his team of industry professionals. Chris owns a 10% stake in the collective operations that comprise the automotive segment, providing strong alignment between the partnership and our shareholders. I would like to take a minute to expand upon our operating model within the automotive segment as we believe it enhances our ability to acquire and successfully operate automotive dealerships within the Washington, D.C. metropolitan area. All automotive retail operators must be approved by their manufacturing partners in order to be awarded an franchise. For Graham Holdings, our current manufacturing partners that have approved our ownership include Toyota, Ford, Honda, and Stellantis. Rightfully, manufacturers have a high bar for who they are willing to approve as franchise owners.
They want proven operators that will be good stewards of their brands and are committed to providing high quality, trusted service to customers. Often that means a preference for partnering with established local operators that have a strong track record of quality and success. However, there are a limited number of such operators that both meet this high standard and that also have access to sufficient capital to acquire and invest in the ongoing operations of large auto franchises in major urban markets. In other words, the universe of potential owners for large franchises is rather limited by the ownership criteria that manufacturers have established.
Another advantage of our current footprint and partnership approach is that allows us to benefit from regional economies of scale and scope that can be achieved because we operate several distinct dealerships in the same market. Local knowledge and history are important in the automotive retail business, our operators have decades of experience running dealerships in the Washington, D.C. region. There are also less obvious advantages to regional scale. For example, the fact that we operate multiple dealerships in the same market gives us a deeper proven pool of talent to tap into when a mission-critical position like a general manager or service department director turns over. It also correspondingly provides talent development and upward mobility opportunities for our employees. Regional scale also enables us to do things like buy and manage parts inventory more efficiently and optimize our used car inventory across dealerships.
In addition, investments in marketing and technology have a higher ROI when it can be applied to a larger regional platform. Given this context, Graham Holdings is well-positioned to acquire quality franchises in the Washington market. Our long-term orientation, commitment to operating with high integrity and quality across all of our businesses, in addition to our partnership with a strong local operator, is looked upon favorably by manufacturers relative to many other traditional sources of capital, like private equity.
As a result, when a current owner and local operator of a franchise is looking to sell a business that they have owned and operated, often for decades, they can take comfort in knowing that we are a buyer that understands the local market, has sufficient capital to complete a transaction, and is on the short list of buyers that they can have a high degree of confidence will be approved by the manufacturers. This advantage has enabled Graham Holdings to opportunistically acquire automotive franchises in a rather restricted M&A marketplace at reasonable purchase price multiples that we believe provide good risk-adjusted cash-on-cash returns for our shareholders. Before discussing the historical financial results of our automotive operations, I'd like to briefly touch on the business model of a dealership. Automotive dealerships have a diverse set of revenue and income streams.
The result of this diversification has been that well-run auto dealerships have historically been steady cash flow producers, resilient in recessions, and have several moats that insulate incumbent operators. Automotive dealerships have diversified income streams that are commonly classified into two categories, front-end operations and back-end or fixed operations. Front-end operations consist of new and used vehicle sales, finance and insurance products, and the leasing department. As the name implies, these departments are usually found at the front side of the dealership. The back end or fixed operations of a dealership include the parts and service departments, typically at the back end of a dealership. The front-end operations drive most of the revenue of a dealership, but historically had lower gross margins, with very slim margins on car sales, buttressed by higher margins from finance and insurance product offerings.
The back-end operations, however, typically have higher margins and are a more stable source of earnings for dealerships year in and year out. Even when people are buying fewer cars, they still need to service their cars to keep them on the road, thus providing resiliency in the business through up and down cycles. As you will see in our financial results since the pandemic, supply constraints by manufacturers, due largely to chip shortages, have led to increased front-end gross margins relative to historic levels. Coupled with the stable earnings generated from the service departments, this has led to stellar results through the pandemic and our last reported quarter. We know this won't last forever, and that profit margins will eventually normalize down again as the supply chain constraints positively impacting the channel ultimately work through the auto ecosystem.
We also recognize that the automotive dealership model will need to evolve and innovate going forward to meet the needs and preferences of customers and broader stakeholders. Similar to most of our other businesses, our operators will need to adapt to stay relevant in a changing world. That said, we also believe the automotive franchise model in the United States is here to stay. Successful dealerships will need to improve customer service, enhance their digital offerings, and lean into new technologies and ways of doing business. We view this as an opportunity to embrace innovation in ways that will enable our dealerships to thrive alongside manufacturing partners for many years to come. Let's turn to the financial results. Our automotive segment's revenue grew from $236 million in 2019 to $327 million in 2021.
Over that same time period, operating income before amortization grew from $500,000 to $7.8 million. Year to date in 2022, the segment generated $510 million in revenue, a 110% increase over the same period in 2021, and $25.5 million in operating income before amortization, an increase of 189%. Growth in 2022 was attributable to the addition of the Ford dealership in December 2021 and the Toyota and CDJR dealerships in July of 2022, along with growth at the other dealerships, except for declines at Honda due to inventory shortages.
I would like to conclude my remarks on our automotive segment by highlighting a business that we created and launched within the segment in 2020, CarCare To Go. If you live in the Washington, D.C. area, we encourage you to try the service. CarCare To Go is a new concept in car repair that introduces transparency, trust, and convenience into the sector. The valet service will pick up your car directly from your home and deliver it to our automotive repair center, where our technicians will inspect the vehicle, inform you of the service needed, inclusive of pictures and video explanations from the technician, and provide an estimate of the cost and timing of the repairs for your itemized approval.
If you choose, this whole experience can all happen completely digitally via our technology platform, or for you extroverts, you can also speak directly with one of our professionals at any point in the process. Upon completion of your service, the car will be returned to your home. CarCare To Go is an example of Graham Holdings' willingness to invest in technology and innovation to improve upon a legacy service. Whether it is within our healthcare division, at Kaplan, or within our restaurant operations, we are always looking for opportunities to apply technology and digital expertise to our products and services to better serve customers. CarCare To Go is still very small, but we are optimistic about the potential for the business based on traction we are seeing and the fact that new customers that utilize the service absolutely love the experience.
Since our acquisition of the Honda and Lexus dealerships in 2019, our automotive segment has grown to become an important part of the company. We've been pleased to find attractive bolt-on opportunities within the segment over the last couple of years and are hopeful we will be able to find additional attractive M&A opportunities from time to time going forward as well. The more recent results of the segment have been stellar as the business benefited from favorable macroeconomic trends. Thus you should expect the results of the segment to normalize down at some point. That said, we think the automotive segment will continue to be a meaningful part of the earnings power of Graham Holdings for many years to come. Back to you, Tim.
Thank you, Jake. We have been busy working to improve the business and are optimistic about the future. Management believes we now have multiple businesses through which we can deploy capital attractively to augment organic growth. As we end the year, we have a reduced share count, strong operating cash flow, a healthy balance sheet, and promising growth prospects at several units. While cognizant that the unexpected can happen, we are optimistic about the future of Graham Holdings. With that, we'll begin the Q&A portion of the session. As a reminder, you can feel free to enter any questions that you might have in the presentation portal, we will take questions until our allotted time is up or until there are no more.
With that, we'll kick off with the first question that's come in, which is, you have become considerably more leveraged the past couple of years. Why? Will that trend continue? I think it's a good question. I would say that the overall leverage profile of the company is one where we've continued to have a negative net debt standpoint. From an overall kind of fundamental conservatism of balance sheet, the overall cash and securities have and continue to offset any debt that we have. That may not always be true.
I suspect over the long arc of time, that won't always be true, 'cause if we, if we find an opportunity, where you know, leveraging some debt to go and fund that opportunity makes sense, we would look at doing so. It is correct that the growth debt number has gone up. I would say it's gone up roughly in line with both kind of the reported and our view of the underlying earning power of the business. In terms of the overall risk standpoint, I, I don't think that we view that there's been any material change, because the size of the business has grown. Our, and our ability to service that debt, is such where we're very comfortable with that.
That's how we kind of view today. Just putting a little bit more on will that trend continue, I think there's no point in having debt for the sake of debt. If we were to go and add additional debt, you know, beyond mild things tied to working capital and inter-period, that sort of thing, it's largely going to be because we find very compelling opportunity that we think it makes sense to wholly or partially fund with debt. Everything that we would do that would, I think would be a material change would really be something driven by the opportunity set that we have and just not running with a more levered balance sheet for the sake of having a more levered balance sheet.
All right, the next question is about Framebridge. Framebridge, bringing happiness and joy to your walls. Where are the component products of this company manufactured? I will give a summary, and then I'll ask if Susan would like. She's on the call, if she would like to chime in on anything. We have two manufacturing facilities right now. Both are within the U.S., one in New Jersey, the other in Kentucky. That's where the production of the products actually takes place. The components, so the acrylic or the moldings and wood, range from Italy to, you know, U.S.-based moldings. You know, I don't know.
Susan, would you have anything else to add on the manufacturing process?
That's right. Our component parts are from the USA, Italy, Southeast Asia, and very little from China.
Okay. All right, the next question is: Hi, can you please give additional detail on Leaf, possibly breaking out the three different businesses and our hopes for 2023 and beyond for them. Are we close to giving up on any of them? Thank you. As I referenced in the remarks, you know, Leaf had unanticipated losses and, you know, has been performing quite a bit worse than what our hopes and expectations were for the business. If I speak a little bit about each of the three components, that might be the most helpful way for people to understand. The Leaf Media business has generally had what I would call, you know, advertising trends in line with what you've read from a macroeconomic environment.
The back half of the year, there has been some weakness that has come from the advertising and ad sales front. You know, we suspect that like many things, that will recover in time, and, you know, hopefully that with some of our support and stronger brands that we'll be able to kinda gain share and, you know, who knows, maybe even come out better on the other side of it. There's an advertising slowdown. I would say nothing traumatic, but certainly noticeable within that business, and we expect that to recover.
On the Saatchi Art side, which is our marketplace where artists can go, who may want display in a neighborhood gallery, can also get the expanded reach of showing their art to any potential customer via our platform. I would say that has seen a little bit of pullback tied to just overall slight consumer home goods decline. Has been, you know, hanging in there and doing all right. I would say the third business that we have is Society6, and that has been where the largest struggle has been. You know, a variety of factors all happened in that business in a very short amount of time.
Substantial increase in shipping costs, the IDFA changes tied to how Apple would allow advertising to occur through other apps and platforms, really changed the ability to reach and efficiently target folks. You know, frankly, we assumed some degree of pandemic and home goods and decor related pullback, and that pullback has been substantially greater than what our assumptions were. That's a little color on those businesses. You know, we're working very closely with that team to figure out how we can fundamentally, you know, get Society6, but all of the businesses into a better place that looks more like the businesses that we believed that we were acquiring at the time of the acquisition.
That's probably enough good context and color on what's happening in those different segments there. The next question is, what can you do with a greatly overfunded pension plan?
So, this is a, we've talked about in previous years, we are in a very unique position of having an overfunded pension plan, and we have found ways to use it in, I would call incremental fashion, over the years, ranging from assuming pension liabilities as part of an acquisition, to more recently, we rolled out a pension credit, a retention credit for nurses within our healthcare segment, where they can earn up to $50,000 over a certain number of year period of time, if they stay with the company, and that is something that the competitive environment cannot offer and allow.
We are finding ways to use that, and I suspect we will continue to find ways to use that. I did reference in our remarks that we expect that pension expense to go up over time, which reflects our belief that we will find more ways to use that that are differentiated in the market environment. It is challenging the current laws around just withdrawing or using a big chunk of the overfunding are quite punitive from a taxation standpoint. And frankly, you know, don't make a ton of sense unless you believe you will never find ways to use that overfunding. We believe that we will over time. We continue to look at different paths.
We continue to be open to acquisitions where the assumption of pension liabilities are part of that those acquisition discussions. As I said, we've made incremental progress, and I suspect we will continue to make incremental progress and be open to making larger progress as we keep going down different paths. All right. Okay. Just going to next question here. There's a couple of duplicates. Why did you decide to provide so much less information than at last year's meeting? Why did you decide to not continue to provide updates on important matters you talked about last year, such as potential opportunities with the overfunded pension? Well, I think we did not have a viewpoint of that we were providing less information.
I think we look at a high level that we have the annual meeting, then we've had an investor day where we can give substantive updates on the business and allow for substantive Q&A. We look at and assess the things that we think people should know about the business, and that's based on kind of questions that come inbound and our own views. Then we take the content and the agenda to be shaped around those two points. While we certainly would cover different things than we covered on the agenda last year, I don't think we would view it as sort of less information.
On the question on the pension, you know, we're open to questions here for things that we didn't cover. Potential opportunities with the overfunded pension was the other part of your question, which I think I hopefully covered in enough detail in my previous answer. The last thing I would say is that we did last year talk about a potential option we were exploring called a QRP, which we gave an update on over the summer in some of our filings. Maybe, Wally, if you'd like to kind of chime in on the update associated with that, just since that's highly related to the pension.
Sure. As we discussed last year at this time, we were exploring the possibility of looking for a qualified replacement plan as one way to utilize surplus pension assets. This would have allowed us to extend defined contribution benefits. As Tim mentioned, you know, with the complexities of these kinds of transactions and federal excise taxes that can kick in, typically companies will look to get a private letter ruling from the IRS before they would move forward with something like that. In June 2022, the IRS announced that it would no longer issue private letter rulings on this type of transaction. At this time, the company is no longer actively considering that specific transaction.
Thanks, Wally. Next question is, what is the CapEx required to open a typical Framebridge store, and how does the opening of a store change customer acquisition costs in that particular geographic market? The first part of that question, we haven't disclosed a CapEx number. As you would expect, it does vary. If you're opening in a suburban location somewhere, versus lower Manhattan, you have kind of two different numbers. Which is why we, you know, we reference that we do target a sub-24-month payback period on the opening of a store. Hopefully that's enough insight for folks to get a better sense.
The opening of a store, how does it change customer acquisition costs in that particular geographic market? This is a hard thing to precisely answer, but we know it improves, and we believe it improves substantially. The stores themselves serve as billboards, and they are usually in relatively high traffic areas, so they are their own form of out-of-home advertising, where the rent essentially drives customer acquisition. We see, you know, customers coming in to the stores themselves because of that, and we tend to be able to understand and look at particularly in more dense markets, how it can impact the online customer sales as well in a favorable way.
The last thing I would say about, you know, a store and store behavior, you know, is we do see that there are people who have been going online and using Framebridge through just an online version that when we open a store in their neighborhood or within a short distance of them, they expand their purchasing and start to use both online and in-store options. We think these two different pieces of the business work quite well together. What is the company's approach to real estate ownership? Approximately how much is the owned real estate worth? We did touch on this in the annual meeting earlier this year and where we gave a little bit of a rough breakdown.
I don't think we have any newer rough breakdown on the specifics. The approach to real estate ownership is we are not real estate investors. We look at it, very much as part of an economic decision as part of the operations of the business, and whether something makes sense to, you know, lease to buy, to sell, lease back. Those are all analyses that we've run in our various different businesses. We really are very economically oriented around how we, how we think about that. That's the entire way we do. We don't fashion ourselves as real estate investors.
To the extent that there are any businesses that no longer need to use owned real estate, we would likely look to sell that asset, because it would not be providing the operating benefit that it previously had as well. Next question. Is Leaf at risk of an impairment or are the problems more temporary in nature? Please discuss the long-term prospects versus initial thinking. Maybe I'll start off here, and then I'll turn over to Wally on the impairment and impairment analysis side. The long-term prospects, you know, we believe are unchanged.
The media business, we think there's an opportunity to create a group of branded digital media assets that operate and share a similar infrastructure and be able to cost-effectively grow that business. We don't think that there is any change in the long-term thinking of that over time. On Saatchi, we continue to believe that that opportunity exists, and that, you know, Saatchi should exist, it deserves to exist, and it has the right profile, where it will not require very much capital to continue to achieve that growth. Society6 has become materially more subscale and we are looking at various different opportunities to make sure that the business has the required level of scale to be productive.
If that were the case, then I think we would view the initial thinking not to be any different, the facts on the ground are a bit different in terms of the size and scale of that business. As I said, we're you know, we're working hard with that team to figure out how to get it back on more solid footing. Wally, do you wanna discuss anything on the impairment side?
Sure. On an annual basis, the company conducts goodwill reviews at each of its reporting units, and we do this work in the fourth quarter. You know, valuations are, you know, have declined in certain sectors. Discount rates are generally higher depending on the sector too. There could be some increased risks with certain reporting units for potential impairment charges. In the second quarter of this year, the company previously disclosed that impairment reviews were performed at the Leaf Marketplace as well as Framebridge reporting units. At that time, the analysis indicated that the fair values of these businesses exceeded their carrying values by a margin less than 10%.
As we have disclosed over the years, it's always possible that impairment charges could result, and they could result in the fourth quarter of 2022 or in the future. We will report on the results of once we're completed with these reviews as part of the company's year-end financial reporting.
Thanks, Wally. Next question is, what do you expect the incremental operating margins to be over time for Framebridge? We haven't disclosed the specific margin profile there, but it might be helpful if I discuss a little bit about the economic model. It is an economic model of scale. There are essentially, there's a fixed cost structure on the kind of overhead and G&A side, that will grow very, very slowly over time. We do not believe it needs to grow meaningfully from where it's at today. You have a semi-fixed structure, which are the manufacturing facilities, which will need to grow, but will grow much more slowly than overall top line.
Right now we have two manufacturing facilities and I suspect, you know, over time, we likely will add, our manufacturing footprint will continue to grow, but as I mentioned, much more slowly than unit volume. It really is a question of kind of overall growth, and getting that unit volume up, and then what's the margin at which you're producing those units to cover the cost of the manufacturing facility and the fixed cost structure. Each incremental unit, you know, should have a very strong, you know, flow through to the bottom line, which then we can put in the bank or we can choose to reinvest in additional units of or additional retail units if it makes sense. That's a very quick kind of walkthrough of the model.
We have most of the infrastructure built out to become a much larger business with a improved economic profile, and, you know, which is some of what I, what I reference of why we think that Framebridge is 2022 is a peak investment year is, you know, we're rolling out a manufacturing facility, rolling out additional locations. It's unlikely we will need to do nearly as much on the manufacturing and infrastructure side in 2023. Incremental unit volume can flow through quite a bit more. That's how the model of Framebridge works. Right.
The long-term annual free cash flow profile of the company, previously about $200 million per year, seems to be meaningfully improved based on the earnings power of the various key business segments. Please comment. Can this double over the next five to seven years from the $200 million per year? We are not in the nature of making predictions like that. We think that the overall cash flow profile of the business has recovered and is closer to what I would say normalized earning power of the business since 2019. The businesses have continued to improve, and we've added additional segments. That normalized earning power is probably higher than it was back then.
What the next five to seven ye ars look like is just too far out for our crystal ball. We do think that we have a collection of businesses that, you know, in aggregate should grow organically, and that we do have the opportunities to put capital to work in several of them at really good returns. You know, our aim is for that to continue to grow and, you know, I referenced that if we think there's a gap to intrinsic value, we could bring share count down. You know, if that continues to grow and a gap persists and we bring share count down, you know, free cash flow per share should continue to grow.
That's how we think about the business and what we're really trying to go and drive. The rate and pace of that is challenging to always forecast. All right. Next question: With all the opportunities to continue to invest in the wholly owned businesses and buybacks, does that change your thoughts on the publicly held portfolio? Good question, and today, no, but it certainly could. If we could not fund the wholly owned opportunities that we wanted to fund out of just the cash flow of the business, the marketable securities portfolio, you know, would be a place that we could look at, but we would evaluate those things in concert with each other.
The marketable securities portfolio is not off-limits, but if we really like the growth prospects of a particular holding, we obviously will take that into account. To date, we have largely been able to use the balance sheet of the business and existing cash flow operations to fund those. Just to reiterate, you know, the securities portfolio, we view them as businesses. We're buying a piece of business, and we don't expect to sell it soon after we buy it, and we expect to own it for a long time. If we do see other opportunities that we think are a better use of that capital on a risk-adjusted basis, you know, they're not off-limits. The next question is Clyde's Restaurant Group.
Have you provided an update? Post-pandemic, how is this group faring? We have provided a little bit of commentary usually in our quarterly reports on the business, and in some of the updates at Investor Day or annual reports. The business has recovered. You know, I like to say that, you know, potentially the Washington D.C. area was one of the more impacted areas of COVID because we also had the election, downtown closures tied to that, where several of our locations are. That team had a lot to get through, and they've gotten to the other side.
That business was still impacted in the early parts of 2022, but I would deem it as recovered at this point in time. Downtown D.C. is probably lagging behind in overall recovery to other markets, but we're seeing improvements there. Hopefully there's opportunities for continued growth. But that business is back to the business that we thought, and producing cash flow for shareholders. All right. Question around retransmission. Last year you noted retransmission fees were down 10% year-over-year. What is the trend this year? We don't specifically break out our retransmission numbers within Graham Media Group. We do offer some commentary usually about how we're doing in any particular quarter.
In Q3, we did note that, you know, net retrans was down very modestly. I think we are no different than many others in the sector that have a piece of their business that is tied on to linear television and the cord-cutting efforts or the cord-cutting trends rather, that are along the way. We see cord cutting. We see that manifesting in retrans. We also see that we are continuing to be valued for what we provide to those bundles and our ability to drive incremental rates off of the content that we provide has continued to go up. Those things have been balances to each other.
If you kind of read things about the sector, I would say we fall pretty squarely within that, you know, hesitate to offer predictions about what the future looks like on that front. All right. Well, there's another question on Clyde's, which I think I sufficiently answered with the previous questioner. All right. Healthcare questions. Question is, "CSI seems great. What's the long-term opportunity? There seems to be investment required. Home health, I'm assuming you can continue to grow long-term organically in a capital-light manner. Please comment." I'll give a little high level, and then maybe I'll ask David Curtis or Justin DeWitte, the CEOs of that business to go and talk a little bit more about CSI.
In the home health space, we think we're pretty squarely in a service that is desired by society, where the macro trends of an aging society are squarely in our favor. If we can be a good operator, and if we can partner with the right systems occasionally via JVs, then yes, we should be able to organically grow in that, in that space, in a capital-light manner. We do think there are occasionally opportunities to augment and do a little geographical expansion. Historically-- via the acquisitions, historically, these have been relatively small acquisitions, which is more likely than not what that future would look like as well.
There is, there's occasionally an opportunity to put out modest amounts of capital, but most of that growth should be capital light. On CSI, we're very optimistic about that business and the long-term opportunity. Why don't I have David, you or Justin go provide your point of view there.
Yes, this is Justin. Thank you, Tim. We believe today CSI only services a small fraction of the addressable market, we believe around 6%, in the IVIG therapy. There are also other therapies the company is branching into. We believe there's a lot of growth in the addressable market as it stands today. In addition, the addressable market continues to grow, as we go in the future. We're excited about the long-term prospects of this business.
Tim, this is David Curtis. Just to add around home health, Justin and I feel encouraged we can continue to grow patient care volumes for both home health and hospice, assuming we can continue to, you know, build clinical capacity. I think the only offset to those patient care volumes growing from a revenue perspective would be, you know, what does Medicare do with reimbursements going forward, particularly for Medicare home health?
Thank you, David and Justin. Moving to the next question. Why do you not buy additional TV stations? Are you committed to the business? Do you feel having more scale would be helpful in MVPD contract discussions? The same with broadcast network discussions. Thanks. On the TV stations front, you know, we did acquire two stations several years ago as part of a pension transaction that I actually referenced earlier. We have not been a major consolidator in the space. That is, you know, we are unique in the sense that many of the participants in the space are solely in the TV and local broadcast world.
That is where they focus their capital allocation almost I would say, you know, regardless of kind of price or opportunity set. We are in a variety of sectors, and so we have to weigh where we think capital can go on a risk-adjusted basis, relative to those different sectors. Sometimes opportunities in one sector will not be as attractive as opportunities in another sector. That doesn't mean it is a reflection of our view of that operating business. It just means that it's a reflection of our view of the price that we would have to pay to acquire more businesses within that operating business. That's how we think about it. I actually think that's how our shareholders...
Hopefully that's how you hope and expect us to think about it, is we're trying to find the best deal, and the best long-term earnings power growth for the company, and we'll look in multiple sectors that we're in, not just one to do it. I guess in the second part of the question, do we feel having more scale would be helpful in contract discussions? You know, I would say the answer is possibly, but not to the extent where it justifies going and spending a tremendous amount of money on assets, when we have other opportunities in other sectors to also spend that money. That would feel a little bit like the tail wagging the dog for how we run the business and the company.
Overall, I think that we have large stations and large markets and both our MVPD partners and our broadcast partners realize that and acknowledge that and realize that we are an important part of that ecosystem. From that standpoint, we believe and we continue to believe that we have always been able to punch a little bit above our weight on that front. Okay. Are Society6's issues due to external competition or due to increased demand in its end market? What is Graham doing to improve performance here? You know, the Society6 issues, I think are largely due to end market issues and a substantial decrease in consumer demand relative to prior periods.
We are working with the company to make sure that, you know, the business has a view and can adjust to that decreased demand while also trying to go and preserve opportunity and value to chase against things in the future. I would say that it's been compounded by a few external factors tied to shipping costs, and as I mentioned before, the cost of marketing really went up substantially in a short period of time post some of the IDFA changes. The model and the ability to go and reach and acquire customers also changed substantially at the same time.
That has been a lot for a business to take, and we are continuing to work through and work with the management team on making sure that the company is positioned to regain some of that scale. That's, you know, that's what we're doing and if there's anything else on that front, we'll make sure to let people know. Curious, with all your various business and inputs, do you feel like the U.S. is going into a recession anytime soon? Any input would be appreciated. Thank you. Well, we are really in the business of operating our companies day in, day out. We are never going to deem ourselves macroeconomic forecasters.
You know, our businesses provide some level of snapshot. I'm sure Jay Powell and folks at the Federal Reserve and others have a much greater snapshot. Anything I would say would, you know, look foolish relative to folks with a greater set of data inputs than we have. Okay. Next question is why own marketable securities at all? How does this add value to GHC shareholders who can own their own portfolio? Well, I would say the question that we get from time to time, and thankfully, the, you know, the answer doesn't really change, which is over time, we think this has actually been a pretty good thing for the company.
We view these as owning pieces of businesses that we will tend to own for a long time, usually in sectors that are related to ones that we operate in already, where we think that we might have a level of understanding that may not be reflected in the overall market. It expands the universe and scope of things that we think that we can look at. Additionally, you know, if you look at, you know, if we have held various things in cash versus marketable securities over time, the marketable securities performance has been a very good thing for shareholders relative to cash.
It potentially allows us to, you know, view keeping our absolute cash balance at a different level than we may otherwise because we know those marketable securities are likely to grow, but at the same point can be liquidated in a very short amount of time and provide additional flexibility for the company. We think it's been a good thing for our shareholders historically. We think it's another opportunity for us to be able to put capital out if we think we are know something about a sector that isn't reflected in the market. And we think that that will be true in the future as well. Okay. Next question is, what can KNA earn after maxing out the Purdue earn-out?
Andy, I might just turn this one directly over to you.
Okay, sure. I will. We of course don't try to predict earnings. I would just say that the contract that we have with Purdue is just a subset of the overall portfolio of KNA. We are a leader in commissions and professional education that gives us some unique advantages. We have an array of partnerships with universities on many fronts that we think are, you know, provide a basis for some pretty solid growth going forward.
Okay. Next question is, have the possibilities to use the overfunded pension for an acquisition increased with some disruptions in market values of public and private companies? Say we are ever optimistic, and a lot of times people look at things at year-end when they figure out future potential contributions that they may need to make and what they need to report against. We have yet to see that materialize to date. Here's hoping, though. Okay. The next question is for, I'm gonna pass on over to David and Justin. Which is how much of the home health and hospice revenue is from Medicare and Medicare Advantage? How fast is each of them growing?
How does the company plan on offsetting sequestration and behavioral negative adjustment on rate for 2023? I'll just give a quick summary and then pass it over to them. As a reminder, we've broken out the segment revenue, but we don't break out the revenue by specific, you know, Medicare Advantage source. With that, I'll go, and I'll turn it over to David and Justin.
This is Justin DeWitte. I'll start by addressing the payment mix question. They're a little different for home health and hospice. Hospice plans are a little older in age, broadly, and we run around 90%, Medicare or Medicare Advantage. In hospice today, almost all of that is traditional Medicare. Only recently has legislation moved forward to allow Medicare Advantage plans to retain hospice when the patient elects it. Historically, those patients revert to traditional Medicare. Home health is slightly less total Medicare and Medicare Advantage, with a little more commercial payer, a slightly younger audience that accesses home health. It's still ranging, probably in that 80% plus, from a payer mix standpoint.
The split between Medicare and Medicare Advantage varies very much regionally, state by state in terms of Medicare Advantage penetration. I'll hand it over to David to answer the remaining of this question.
Yeah. Thanks, Justin. I would say home health and hospice, we have grown double digits this year, probably 10%, 15%, across the board in that range. In terms of how we're gonna handle, you know, sequestration and certainly the behavioral negative adjustment for home health, some of that, you know, will impact our margins. You know, we've worked hard on efficiency. We've worked hard this year on utilization and having a consistent approach to how we care for patients, to make sure we're visiting for a good outcome, for a compliant outcome, and an efficient outcome.
You know, the challenge for our business is we have, in some respects, more demand than supply, so we're trying to make sure every visit we do to a patient is effective and really efficient towards driving an outcome.
Okay. Next question is I'm gonna pass over to you, Andy, which is how, if at all, has the competitive environment for English language education changed since COVID?
Well, it certainly if anything, the competitive environment has diminished. That is, there certainly were some companies that were not able to make it through the COVID environment. The fewer the thick. Most of the big operators were somehow able to survive. I think to some degree weakened, but they've survived, and I think we come into this post-COVID period with a pretty healthy competitive set once again. Maybe a little less than it was, but maybe a little more than we might've anticipated.
Yeah. The only thing I would add on as a follow-up on that is as Andy mentioned, some of the bigger providers, you know, have made it through to the other side. I think the footprint and the competitive footprint of locations has changed quite a bit. Entities themselves on the bigger side, you know, it's a similar set of competitors, but the number of actual centers and in cities in which they operate has probably changed and been reduced. That city-level competitive set might be mixed a little bit more than kind of if you look at it from a macro standpoint. Okay. Well, we're looking at our queue, and I believe we are out of questions.
I appreciate everybody taking the time to join us for today's Investor Day. We look forward to seeing or chatting with many of you at the annual meeting in this upcoming May. Thank you.