Walker & Dunlop

In this piece we dive into the financials industry and cover a niche player at the intersection of real estate and finance. We hope you enjoy it and learn a lot! Please email us at info@speedwellresearch.com for any questions, concerns, or comments. See our full disclaimer on our website.

In 1937, in the depths of the Great Depression, Oliver Walker and Laird Dunlop founded Walker & Dunlop. Their company would become one of the first to use the newly formed Federal Housing Administration’s (FHA) insurance for single-family loans. This started what would become a long history of utilizing government sponsored programs to help client secure real estate loans. The business stayed a relatively small family-run business throughout their lives, and in 1976, Oliver Walker passed the reins to the next generation with Mallory Walker being promoted to company president.

The family was generally well-connected in Washington DC, which drove a willingness to quickly adopt new government-associated financing programs. Notably, and most relevant to the future of Walker & Dunlop, in 1988, they were named as one of the first Fannie Mae “Delegated Underwriting and Servicing” lenders (also known as DUS lenders). This allows them to underwrite loans without prior review from Fannie Mae, which opened up a new source of capital for their clients that often offered the best rates on long-term fixed loans. The way it would work is that Walker & Dunlop would vet a client and then underwrite the loan. Fannie Mae would then originate that loan, but to ensure Walker & Dunlop stayed prudent with underwriting, they were on the hook for some portion of losses should the loan default. However, they did not need to raise sources of capital to fund the loans like a traditional bank, since Fannie Mae served that purpose.

In 2003, a third generation Walker, William, joined the family business. Willy, as he is known, joined as Executive Vice President. His appointment would in short order transform the company from a small family business with a few dozen employees working out of a single office in Bethesda, Maryland, to one of the largest Government Sponsored Entity (GSE) loan originators with offices plastered around the country.

Walker & Dunlop’s headquarters in Bethesda, Maryland before moving it.

Having shown a “negative” interest in entering the family business, Willy originally wanted to become an investment banker after college. But that all changed when the Ambassador of Paraguay, who was a friend of the family, offered to take Willy with him to Paraguay. He ended up staying there for 3 years. After that, he would return to the US to attend Harvard Business School, only to end up back in South America, although this time, he was working for a Chilean venture capital firm. Willy would spend roughly a decade working in Latin America, where he gained many valuable experiences via a variety of endeavors that included helping set up a new airline, a stint at Texas Pacific Group, and creating the Latin American call center industry through a subsidiary of TeleTech. His work in the call center industry ultimately led him to Europe, where he would eventually be in charge of $250mn in revenues.

Willy Walker worked to set up TeleTech’s international operations before joining the family business.

With experience building up an organization from scratch, Willy started to see a bigger opportunity with Walker & Dunlop when he talked with his father, who was then head of the company. He joined the Board of Directors in 2000 to learn more about the business, and eventually led an offsite with senior management to explore the company’s strategy. Realizing he no longer wanted to live in Europe and any value he created in his then-current role would be trivial relative to the wealth he could build at his family’s business, he told his dad he thought he could help run Walker & Dunlop by the end of that strategy session.

In 2003, he entered the family business as Executive Vice President. They estimated the company was worth $25mn at the time. Two years later he was promoted to President, and it wasn’t before long he started putting his own imprint on the company.

Willy put in simple terms the two main businesses the Walker & Dunlop operated: the brokerage business and the lending business. The brokerage business is just connecting people who want to borrow money for a property to people that want to lend it, such as banks or insurance companies. Walker & Dunlop receives a fee for this service, but about half of that typically goes to the loan broker. Since brokerage is a relationship-based business, the closer the client is to the broker, the more they could negotiate a higher cut of the fees. At the time, the Walker & Dunlop brand was so unknown it was inconsequential to winning deals, which meant that each broker had to rely on their own network and reputation to win business. As a result, Walker & Dunlop was in a weak negotiating position and reliant on their brokers much more than the brokers were reliant on Walker & Dunlop. There were dozens of other real estate finance firms that would be happy to take on a productive loan brokerage team, many of whom had far better brand recognition.

Old Walker & Dunlop headquarters in Maryland.

However, Walker & Dunlop’s other business, “lending”, was more value additive. This business was when Walker & Dunlop helped a client get capital from Fannie Mae (and later Freddie Mac and HUD). While there are still dozens of other financial institutions that have access to the same capital, a broker who relied on Fannie Mae capital to close deals couldn’t leave so easily as his workflow is more enmeshed in a system where the competency of the back office and supporting services to process a loan is critical to serving clients.

A broker at Walker & Dunlop may feel comfortable that his clients would follow him if he left to a competitor, but if his new firm’s back office wasn’t as competent, he could lose business on a slow or tedious loan closing. While Walker & Dunlop was small, they were singularly focused on service and execution of Fannie Mae loans, slightly differentiating them versus competition—most of whom had much more expansive businesses.

Willy’s first priority after joining was going “all-in” on the lending business since it was more differentiated and also had recurring revenue from “MSRs”. MSRs, or Mortgage Servicing Rights, are a contractual right to “service” a loan. Servicing entails sending statements, collecting monthly payments, and remitting them, among other back-office tasks. While it is not exciting, the right to service is contractually guaranteed income over the life of the loan, while also being a high margin activity. For Walker & Dunlop, MSRs are a significant source of earnings that helps even out the vicissitudes of their other deal-driven earnings.

In 2006, Walker & Dunlop bought out AIG, who was a partner with them in their Fannie Mae DUS business. A year later, Willy Walker became the undisputed leader of the company when he became both CEO and Chairman of the Board. While the financial world was crumbling, Walker & Dunlop’s years of conservative originations and limited balance sheet exposure allowed them to not only weather the banking crisis but take advantage of it.

In the midst of all of this, in November 2007 Walker & Dunlop turned 70 years old and Willy set a 5-year plan, which he called the “Drive to 75”. This was an ambitious goal to increase revenue and operating income 5x in 5 years.

In January 2009 they acquired Column Guaranteed LLC, or “Column”, as it was known, from Credit Suisse. This acquisition gave them licenses to originate and service loans for Freddie Mac and HUD. While competitors were getting defensive and reeling from the consequences of unscrupulous loan originations (which often had risk-sharing agreements), Walker & Dunlop was profitable throughout the entire financial crisis and even gained market share. In 2007 they were the 45th largest commercial real estate lender, but by 2009 they were the ninth.

The acquisition helped give them the minimum scale they needed to go public, with their loan servicing portfolio increasing to $13bn of loans from $7bn and operating income doubling to $29mn for 2009. Even so, they were still a very small company when they went public in 2010, and the timing was less than ideal for a real estate finance company. This was made apparent when they had to lower their offering price to $10 from a range of $14-16. After banking fees, they raised around ~$90mn, which they would use to grow the company with a particular focus on adding teams and acquiring complementary businesses. The estimated $25mn business value Willy Walker and his father estimated when he joined had now grown to ~$220mn at their offering price. The loan servicing portfolio he was focused on building grew from $2.8bn when Willy first joined to over $14bn at the time of IPO.

Walker & Dunlop went public in 2010.

But despite helping the company grow significantly since taking over and guiding it through the worst financial crisis since the Great Depression, Willy Walker knew the business was not in a strong a position. Their service level might have been high, but they didn’t enjoy the same synergies other larger real estate finance firms did by grouping together different real estate services, and they lacked the ability to lend on a short-term basis with bridge loans to close a deal. The number of loan originators they had was also a fraction of the of the much larger real estate focused finance firms like CBRE, JLL, and Berkadia.

From Walker & Dunlop’s S1, which shows how they internally conceive of their services.

As Willy would later say, there were no “grand slams”, just hitting a lot of “singles”. They slowly added team after team to Walker & Dunlop to grow their brokerage network and loan originations. However, there were several partnerships and acquisitions along the way that helped them jump ahead.

On the smaller side, in 1Q11 they entered into a partnership with Real Estate sales-focused firm, Cushman & Wakefield. This allowed Cushman & Wakefield to offer clients trying to close deals a new potential source of capital in Fannie Mae, Freddie Mac, and HUD. This in turn essentially acted as a customer acquisition funnel for Walker & Dunlop. The impact was relatively minor initially, as deal flow would need to pick up for this to become a meaningful source of businesses. From 2010 to 2011, originations grew from $3.2bn to $4bn.

An apartment building in California that Walker & Dunlop arranged Freddie Mac financing for.

In June 2012 Walker & Dunlop announced the acquisition of CW Capital for $220mn, financed by 63% equity. This acquisition doubled the size of their servings portfolio to $35bn by the end of 2012 from $17bn the year prior. It also added ~180 employees, and by the end of 2012, they were at 420.

As people and their relationships were at the core of their business, Willy was careful to foster a strong culture. He tried to have a “no jerks” policy, but their small size led to him tolerate the behavior of one star broker who at one point was 70% of their deal flow. He needed to scale the business up so they weren’t so reliant on one person, and he could fire people who weren’t good cultural fits, even if they were big producers. Thus, he relished the moment when their scale allowed them to let him go with de minimis volume impact. The culture and unwillingness to accept common-place industry behaviors like stealing clients from peers helped attract quality talent, which was only helped by their reputation for being quick and competent.

Incredibly, by the middle of 2011 it was becoming clear they would meet the ambitious 5-year targets they set in 2007 to grow the businesses revenue and operating income 5x. But rather than rest on their laurels, Willy would make clear that was but one step towards their goal of making Walker & Dunlop the premier commercial real estate finance company in the US. To keep growing, they started experimenting with adding different proprietary sources of capital. They set up a CMBS (commercial mortgage-backed security) conduit, created a bridge loan program that would allow balance sheet lending for the first time, and explored starting a Mortgage REIT. While only the bridge loan program was ultimately successful (the mortgage REIT never materialized and the CMBS conduit never had significant volume, so they shuttered a few years later), it shows their willingness to experiment and take risks when there is limited downside to failure.

Multifamily property in Pennsylvania that Walker & Dunlop facilitated financing for.

In 2013, the Federal Housing Finance Agency (FHFA), directed Fannie Mae and Freddie Mac to cut their loan origination volume by 10%, which caught Walker & Dunlop off guard. This, combined with interest rates climbing, led to the first headwind since the financial crisis ended: 3Q13 origination volumes fell 19%. Changes to Fannie Mae and Freddie Mac had always been a risk for Walker & Dunlop, but most proposals at the time either had little political backing or contradictory aims (more on Fannie and Freddie later), but this decrease seemed arbitrary with an unclear purpose.

Willy Walker acknowledged at the time that the most unsettling thing for Walker & Dunlop investors was all of the noise coming out of Capital Hill on the future of Fannie and Freddie, with some believing the institutions should be totally eliminated. However, despite the noise, Willy was an adamant believer that “Fannie and Freddie aren’t going anywhere anytime soon”, and their acquisition of CW Capital, which increased their reliance on the institutions, was indicative of that opinion.

Nevertheless, Walker & Dunlop proactively slashed expenses, shuttering both a small loan lending business acquired from the CW Capital transaction and closing a small office. Still though, despite the unexpected shake up with Fannie and Freddie originations, Willy Walker was clear that their strategy is to continue to emphasize these proprietary sources of capital (Fannie, Freddie, and HUD) and grow their brokerage network in conjunction with those capital solutions. (Fannie and Freddie are both GSEs and considered “agency”, but HUD is only considered an “agency” and not a GSE since it is not quasi-privatized. The HUD business is a fraction of the size as the GSE business).

HUD (or the United States Department of Housing and Urban Development) Headquarters.

The reason they continue to highlight the proprietary capital solutions is simple: they would need to do 5 to 6 times the aggregate origination volume to produce the same amount of revenues without the GSEs (or what Willy called their “lending business”). And even if they could achieve that through a larger brokerage network, these revenues would be at a lower margin and at a higher risk of disappearing to a competitor. Despite the hiccups with the GSEs, Walker & Dunlop would not change their strategy, which turned out to be the right decision as GSE volumes increased again in short order.

Despite their confidence in the GSEs, they tried to diversify their sources of capital the best they could. The more differentiated sources of capital they could offer, the easier it would be for Walker & Dunlop to sell themselves to brokers, which in turn allowed brokers to better service clients.

A year later, there was evidence that their strategy was working. On a 2Q14 earnings call, they noted that they quoted a three-property multifamily deal for execution with the GSEs, but the borrower required more proceeds than the agency loan parameters allowed. Instead of losing that business as they would have in the past, they were able to quote the borrower for a deal with their conduit and closed that deal. In another case, their balance sheet lending operation allowed them to offer bridge financing for three other properties that were not stabilized yet (stabilized refers to having the building at least 80% rented). Then, when the building gets rented out, the owner will refinance it through Walker & Dunlop to get a GSE loan with better terms. These were both deals they would have lost without these new sources of capital.

These new capabilities were changing the way Willy would talk about their business and their ability to attract talent. They would not need to pursue “star brokers” and could go after second and third tier teams because the Walker & Dunlop platform could help support them and boost their business (versus relying on the star brokers to boost Walker & Dunlop’s business).

Over the next several years, Willy Walker would extend their platform and business lines with several partnerships and acquisitions: a Blackstone JV in 2017 to create a Mortgage Trust that unlocks a new source of capital for Walker & Dunlop, a 2018 acquisition of JCR Capital which would seed their asset management platform, a multifamily property appraisal business called Apprise in partnership with GeoPhy in 2019 (which they would later acquire in 2022), the acquisition of 75% of real estate-focused research firm Zelman & Associates in 2021, and alternative asset manager Alliant Capital that materially increased their AUM to $16bn.

At the end of 2017, Walker & Dunlop posted their next set of long-term goals, called 2020 vision. It called for revenue growth of ~13% annually to get them to $1bn (from $700mn), increased loan originations to $30-35bn (from $25bn), and the addition of $25bn to the loan servicing portfolio. More ambitiously, they wanted to grow investment sales volume ($3bn in 2017) and Asset Management AUM to $8-10bn.

Their 2020 Vision Goals released in 2017.

Ultimately, they would miss their $8bn in investment sales goal and $8bn in AUM from 2020 Vision, with Covid only partly to blame. Nevertheless, they announced their next set of goals, dubbed “Drive to 2025”.

From Walker & Dunlop’s 2022 10-k. They surpassed their AUM goals with the Alliant Capital acquisition, which added ~$14bn of AUM.

This new set of long-term goals include: 1) double total revenues to $2bn, 2) increase annual debt financing volume to $60bn, 3) grow multifamily property sales volume to $25bn, 4) increase the servicing portfolio to $160bn, 5) take AUM in the fund management business to $10bn, and 6) start three new businesses in small balance lending, appraisal, and investment banking. These goals are by no means padded, and there is a fair risk they fall short of them, but setting ambitious goals means risking missing them.

With the partnerships and acquisitions mentioned, in addition to the many “singles”, Walker & Dunlop has already made material progress towards their goals.

Willy Walker.

While 5-year targets are long-term for most public companies, they are just interim goals for Willy Walker. Ultimately, he wants to make Walker & Dunlop the premier commercial real estate finance firm that is the #1 player in multiple different business lines.

Walker & Dunlop is a commercial real estate finance firm whose services include multifamily lending, property sales, commercial real estate debt brokerage, and investment management among other real estate related offerings. They operate through three reportable segments: 1) Capital Markets, 2) Services & Asset Management, and 3) Corporate. The Capital Markets segment is fairly volatile, as it is directly tied to deal activity. The Services & Asset Management segment, also called SAM, is steadier, but has significantly increased recently from the Alliant acquisition.

We will start by going through the reportable segments to get a better idea of their business activities, and then move to how they categorize their revenues to understand how they make money.

Ignoring the corporate segment, which mostly consists of their treasury operation and unallocated costs, Capital Markets was 55% of LTM revenues, but just 40% of net income. Servicing & Asset Management (SAM) is now 60% of net income, despite being 45% of revenues.

We will start by diving into the Capital Markets segment, which contains five separate activities. The overarching theme of this segment is that all services are either directly related to deal activity or support deals. At a high level, the Capital Market segment is 1) the agency business where they work with Fannie, Freddie, and HUD to originate loans, 2) their brokerage network where mortgage bankers help connect borrowers to potential lenders that are not the GSEs or HUD, and 3) helping facilitate the sales of properties. The other two activities housed here, research and appraisals, were businesses that were recently acquired to expand their real estate-related services. The table below goes into more specifics on each of the five activities within Capital Markets.

The next segment is Servicing and Asset Management, which is split into three segment activities, which can generally be bucketed into:  1) loan servicing, 2) lending, and 3) real estate investment services. The last category is largely the operations of JCR Capital and Alliant Capital. The table below goes into more detail.

Their last segment is corporate, which is described below.

We have just gone through Walker & Dunlop’s reportable segments, which should help clarify their business activities.

We will now walk through key business activities and explain how they are tied to their reported revenues. To do so, we will start by looking at their revenue segments, as shown below.

Above, we see that there are eight revenue segments, and many have some level of activity in each reportable segment. However, for the sake of simplicity, we can think of the revenue segments as tying to the reportable segments as follows:

This is how we will think about the reportable segments tying to the revenue segments. Loan Origination and Debt Brokerage Fees, FV of expected servicing, and Property Sales, are all Capital Markets revenues. Servicing fees, investment management fees, and escrow earnings are all SAM revenues. Both have some net interest income and “other” revenues.

Now with that clarification, we will start by going through Capital Markets revenue items.

1) Loan Origination and Debt Brokerage Fees, Net (For 2022, $348mn in revenue and $44bn in total financing volumes). A loan origination fee is recorded when there is a commitment to originate a loan with a borrower and sell it to an investor, or when a loan from a WD broker closes with the institutional lender. The loan origination fee is recorded net of any co-broker fees, which are commissions WD pays out to the broker (usually equating to 45-55% of the total fee, depending on the broker). Also recorded here are any changes to FV during the loan holding period, and thus no gain or loss is recognized when sold.

2) Fair Value of Expected Net Cash Flows from Servicing, Net (Revenues of $192mn in 2022). We mentioned prior that the servicing income is from carrying out all of the cashier functions relating to the loan, such as billing statements, collecting and remitting loan payments, and related activities. However, this “revenue” is not from when those services are rendered, but rather is a gain related to the expected net future cash flows from servicing income, discounted back to present to derive a “fair value”. Essentially, the accounting is to run a “DCF” on the mortgage servicing rights (MSRs) income and recognize it as revenue in the period the loan is committed. (They technically first record a derivative asset when the loan is rate locked and forward sold, which then is reclassified as a MSR at loan delivery.)

When the loan is rate locked and forward sold, but not delivered yet, they record a derivative asset equivalent to the discounted sum of future mortgage servicing revenue. The derivative asset is then reclassified as an MSR asset when the loan is delivered.

The mortgage servicing rights (MSRs) are contractual and unlike residential mortgages, do not allow prepayment without a hefty penalty, so they have a much higher certainty of occurring. The mortgage servicing fee amount is derived as a % of the remaining loan balance. Thus, the cash flows associated with mortgage servicing are reasonably certain and can be calculated with high precision. However, picking the discount rate is (as always) fairly ambiguous. The range they disclose they use is 8-15%. If a borrower chooses to prepay, the penalty is about equivalent to discounting the MSR at the risk-free rate. This means that if a borrower does prepay the loan, the amount they receive is typically greater than what they record the asset at on the balance sheet.

As clearly laid out in the exchange below, 90% of their servicing revenues are prepayment protected where Walker & Dunlop receives a check if the loan is paid off early. In the other 10% of cases, the prepayment penalty is usually still there, but paid to another party. This still results in some disincentive to prepay, but no compensation in the event it is. The prepayment protection is a key difference between residential and commercial MSRs.

Now, with some understanding of what the “Fair Value of Expected Net Cash Flows from Servicing” is, we can say that this accounting treatment makes little to no sense. Typically, revenues are earned in the same period as a service is rendered, however this GAAP-created revenue item disregards that convention and allows all future associated cash flows to be recognized up front.

While this accounting treatment might be GAAP-mandated, we take issue with it as it results in two revenue streams being recorded for servicing fee in the same period. One from the sum of all discounted future cash flows from the servicing of a loan and a second from the actual cash received to service that loan (which is recorded in SAM as “Servicing Fees”, as we will show in the next section).

Recording the future sum of all cash flows from a contract upfront, even if discounted, is delusive accounting that isn’t mirrored by virtually any other contractual stream of revenues in any other business (That would be akin to allowing a cloud data provider who just signed a 5-year contract or real estate lender to recognize all of their revenues up front. Conversely, it would imply that all interest expenses on debt should be charged in the period the debt is issued. This is all clearly nonsensical accounting that construes the economic reality of the business and obscures the business’ performance in a given period).

Any way you look at it, revenue is recorded twice for the same business activity. This creates a need to offset one of the streams of revenue to avoid overstating earnings. To do this, the accounting dictates they create a Mortgage Servicing Right Asset when the gain is recorded and amortize it over the period of the loan. Thus, as they generate the “real” servicing fee income, they write down the MSR asset to unwind the gain that was recorded when that loan was committed. This is all unnecessarily complex and obfuscates the true cash earnings in a period. We will back these gains out of our estimates and the associated amortization in our adjusted revenue and adjusted operating income estimates to undo these poor accounting practices. (Additionally, for simplicity, we will refer to this item as the FV of MSRs or the MSR gain rather than its full name, “Fair Value of Expected Net Cash Flows from Servicing, Net”.)

3) Property Sales Brokerage Fees (Revenues of $121mn and sales volume of $20bn in 2022). These are broker sale fees Walker & Dunlop earns when their investment sales team completes the sale of a multifamily investment property or land deal. These fees are typically based on a % of the ultimate sale price of the property.

Walker & Dunlop Capital Market Team sold this hotel and helped facilitate financing for the buyer. They would record revenues in all three of the revenue categories listed above for this transaction.

Now we will move to Servicing and Asset Management (SAM).

1) Servicing Fees (Revenues of $300mn and servicing portfolio of $123bn in 2022). These are fees Walker & Dunlop earns from servicing loans, which in addition to the prior listed functions, can include monitoring the physical condition of the property, analyzing the financial condition of the borrower, processing insurance payments, and managing escrow balances. They service essentially all of the loans they originate and some of the loans they broker.

This is the servicing fee revenue line item that is recorded as the payment is received and service is rendered. They provide a relatively stable stream of revenues that are contractually obligated, with any prepayments usually resulting in an amount equal to the lifetime of the fees to be earned from the loan, discounted at the risk-free rate. The average duration of their loan servicing portfolio is usually in the high single digits, nearing 10 years. They currently have ~$125bn of loans that they service, which has been growing every year, increasing their stream of recurring servicing revenue.

While we take issue with the accounting treatment of the servicing rights (the MSR gain that is recorded), those are not Walker & Dunlop’s accounting rules. Accounting treatment confusion aside, the servicing fees are indeed very attractive as they are a recurring revenue stream that is highly likely to occur, with the worst-case scenario being Walker & Dunlop receiving a substantial penalty fee. Growing the servicing fees has been a priority for Willy Walker since he took over and servicing fees have grown from $14mn in 2006 to over $300mn today.

2) Investment Management Fees (Revenue of $72mn in 2022). They manage invested capital from third-party investors through an investment fund structure. The investment fund makes investments in multifamily investment opportunities, primarily as equity in market-rate or low-income housing tax credit (LIHTC) generating multifamily properties. The LIHTCs are federal incentives for owners to reduce the rent on their buildings to a % of the area’s median income. This creates the incentive for a real estate owner to lower their rents to in turn receive tax credits. These tax credits can then be sold to other investors. Walker & Dunlop earns an investment management fee based on a contractual percentage of the invested capital.

3) Escrow Earnings and Other Interest Income ($53mn in 2022). They earn fee income on property-level escrow deposits in their servicing portfolio. Escrow earnings reflect the placement fees net of interest paid to the borrower, if required. Also included with escrow earnings and other interest income are interest earnings from their cash and cash equivalents and interest income earned on their pledged securities.

The following two items are both in Capital Markets and SAM.

1) Net warehouse interest income (Revenue of $16mn in 2022). They earn warehouse interest income from loans held for sale and loans held for investment, which are funded by their warehouse facility. Warehouse interest expense is incurred on borrowings used to fund loans solely while they are held for sale or for investment. Warehouse interest income and expenses are earned or incurred on loans held for sale after a loan is closed and before a loan is sold.

2) Other Revenue (Revenues of $158mn in 2022). Other revenues are comprised of fees for processing loan assumptions (which is when a buyer of a property assumes the seller’s debt), prepayment fee income, application fees, property sales broker fees, appraisal revenues, income from equity-method investments, asset management fees, certain revenues from LIHTC operations, and other miscellaneous revenues related to their operations.

Walker & Dunlop sold this building in Los Angeles and arranged financing for the buyer. This exemplifies the synergies of owning a real estate investment sales firm.

Below we walk through a loan origination example to understand the process and see when revenues are recorded.

The diagram below shows the loan process. First a borrower is sourced, screened, and directed to fill out an application. The loan application then goes through an underwriting process where it is approved or denied. Approved loans are then “rate locked”, which is when an interest rate lock commitment is extended to the borrower with terms set once signed (this is also called “forward sold”).

It is at this point that Walker & Dunlop will record revenue for Loan Origination Fees, as well as a gain from the fair value of expected MSRs, which is balanced by the creation of a “derivative asset”. However, the loan is not funded by the agency yet, and Walker & Dunlop has not received cash revenues. In order to get the borrower loan proceeds quicker, Walker & Dunlop will use a warehouse financing facility (essentially a credit line with a bank), to fund the loan in the interim. They feel comfortable doing this as a Fannie Mae Designated Underwriter, they have the ability to effectively “approve” the loan and there is little to no chance Fannie Mae will not accept it. Since Walker & Dunlop is holding the loan at this point, they earn the loan’s interest income. They may hold the loan for anywhere from 1-2 months on average. They record Net Warehouse Interest Income from the loan’s interest less the warehouse’s interest expense (i.e., the spread between what they borrow at and lend at).

When the loan is delivered, they will receive proceeds to pay off the warehouse loan and all balance sheet impacts are unwound. Their origination fee is received at this point, but no revenue is recognized (as it is recognized at rate lock commitment). The derivative asset that was recorded alongside the gain from the fair value of expected MSRs is reclassified into a MSR asset. Over the life of the loan, the MSR asset will be amortized, creating an expense in that period to reverse the original gain. This happens alongside the “real” servicing fee revenue that will be collected over the life of the loan.

Agency loan process from their May 2023 presentation.

Rolling all of the segments up together, we can see the revenue mix split for 2021 and LTM below, where they generated a total of $1.3bn and $1.2bn in revenues, respectively. Note how with slowed deal activity the Loan Origination Fees and FV of MSRs shrink, while servicing fees grew.

Looking historically at their total revenues, we would expect to see cyclicality, like we see LTM, but for the past decade plus, Walker & Dunlop has consistently grown revenues year-over-year. This is largely because 1) there has not been a major slowdown in real estate activity post the financial crisis (a minor exception is the 2017-2018 period where interest rates started to rise and Walker & Dunlop’s 2018 revenues were +2% y/y) and 2) they were small enough that market share expansion had a material impact.

Operating income (which includes interest expense) shows more volatility, partly owing to how much operating leverage they have in their model, but also because of the impact of the MSR gains. In 2019, the FV gain from MSRs was $181mn and in 2020 it was $358mn, but their servicing fee income only grew from $215mn to $236mn. In our revenue build, we will adjust their earnings to back out this non-cash MSR gain and add back MSR amortization to create a metric that more closely aligns with the economics of Walker & Dunlop’s business.

With that background on the business, we will now move onto the industry to better understand the total addressable market and how it fragments.

Walker & Dunlop’s core market is multifamily, but they service all commercial real estate verticals including hospitality, healthcare, office, retail, and multi-use. (Multifamily is a real estate term for having more than one separate housing unit in the same building). Still though, their existing brokerage network is mostly geared towards multifamily and only in the United States.

This is partly owing to Willy Walker’s time in South America, where he saw how easy it was for a government to interfere or simply take over a business, which meant he was never enchanted by the allure of international expansion. Instead, he decided to focus solely on the US, which is a multi-trillion-dollar TAM anyway.

Below, we see that they estimate the total amount of multifamily debt outstanding to be $2tn and total commercial debt to be more than double that at $4.5tn. This is compared to their ~$120bn servicing portfolio, suggesting just a 6% market share of multifamily debt or 2.6% of total commercial real estate debt.

May 2023 Investor Presentation.

Another way to look at the TAM is to focus on originations instead of debt outstanding. Within multifamily loan originations, we see that their market share has increased in the past couple of years to ~8%. While in 2022 industry non-multifamily loan originations were about twice of that of multifamily ($450bn vs $800bn), WD’s market share there was much lower at ~2%.

Multifamily loan originations peaked in 2021 at $487bn and are estimated to continue to decline into 2023.

It is important to keep in mind that regardless of how real estate capital flows change, every year a large amount of loans come due that need to be paid off. That could be done with a large investor equity injection (which is seldom done) or by refinancing. Refinancing is just when an existing debt obligation is replaced by a new one, paying off the original one in the process. Virtually every time a loan comes due, the real estate investors expects to refinance the property because paying off the debt with cash would require them to have a lot more money than they typically do, and the lack of leverage would drive their equity returns down.

The slides below highlight that $381bn of debt is coming due from the GSEs, life insurance companies, CMBS and credit card companies, which tends to be their prime customers (particularly the GSEs and life insurance companies).

Of the ~$44bn of debt in 2022 that Walker & Dunlop financed, 41% of it was from the GSEs. The split is shown below. “Brokered” is where they help connect a real estate investor to a lending institution for a fee. While brokered is the biggest portion of volumes today, the GSE business is more profitable because they get to retain the MSRs versus the brokered business where they usually do not.

Thus, despite the GSEs being less than half of their volume, they are of critical importance to Walker & Dunlop’s business. In fact, most of what Willy Walker has done over the past decade has been to move them away from reliance on the GSEs through expanding their ability to access proprietary capital (Blackstone Mortgage Trust JV, B/S lending, investment funds) and growing their business lines (asset management, appraisals). While these all help strengthen the value prop of Walker & Dunlop, any dramatic changes to the GSEs could be very problematic for their business.

To better understand the potential issues, we will touch more on the GSEs’ history, purpose, and some potential proposals.

The origins of the GSEs date back to the Great Depression, where financial institutions were traditionally either not in the position to lend or had no desire to. In 1934, the Federal Housing Administration (FHA) was established to offer federal insurance on long-term mortgages loans. To support this aim, in 1938 the Federal National Mortgage Association (or Fannie Mae as it’s now known as) was created to operate a secondary market facility that would purchase said FHA loans. In 1948, Fannie Mae was re-chartered to allow them to also purchase mortgage loans guaranteed by the Department of Veterans Affairs (post-WWII, the GI Bill helped many war veterans buy homes). This was the beginning of opening up who could sell Fannie Mae mortgages, but the private sector would not be included in that for two more decades.

In 1968, Fannie Mae was privatized and split into two entities: the first which kept the Fannie Mae name and its secondary market operations, and the second became a newly established government corporation called Ginnie Mae, which sat in U.S. Department of Housing and Urban Development (or HUD). Ginnie Mae assumed administration of Fannie Mae’s Government-insured mortgage loans and was allowed to provide full faith and credit guarantees on mortgage-backed securities that were issued by private lenders. Two years later, in 1970, Congress authorized Fannie Mae to acquire loans that are not insured by the Federal Government, officially allowing private originators to sell loans to Fannie Mae. At the same time, Congress established the Federal Home Loan Mortgage Corporation (or Freddie Mac) to provide competition to the “privatized” Fannie Mae.

The GSEs didn’t really start growing until the 1980s when regulators increased bank capital requirements on the back of the S&L insolvencies and the Latin American debt crisis. The GSEs’ regulatory capital requirements remained below other private financial institutions, which effectively meant they could earn the same ROE on lower yielding assets. The GSEs owned or guaranteed 8% of single-family debt in 1981, which grew to 25% by the end of the decade. Today, that has nearly doubled to just below 50%. (A White Paper released from the U.S. Treasury in 2019 claims the expanded government role did not lead to much, if any, increase in home ownership).

In the depths of the financial crisis, the guarantees Fannie and Freddie made on the loans made them effectively insolvent as homeowner defaults increased. In 2008, Congress passed the Housing and Economic Recovery Act of 2008 that granted the FHFA the ability to set risk-based capital and also place the GSEs into receivership. It was too late to try to buttress the GSEs’ slim capital base and instead they were quickly placed into conservatorship.

The first day in conservatorship, the Treasury invested in each GSE up to $100bn each (that cap later increased to $200bn) to maintain positive equity. In return, the Treasury received a basket of securities that essentially ensured no common stock investor could ever expect to receive any future proceeds from the company. The conservatorship that happened in the wake of the Great Recession was meant to be a temporary fix. Instead, they both still remain with the same conservatorship status. There were several proposals to change them, but neither party could decide how.

In 2011, the Obama administration put forth three proposals for Fannie and Freddie: 1) Remove the government backstop and par back the FHA to only aid in limited low-income borrowing, 2) create a federal backstop that would only kick in during a severe downturn, 3) act as a reinsurer providing insurance on mortgages to only after a group of private insurers had been wiped out. While the Obama Administration policy was generally interpreted as wanting them to be wound down and replaced, at the same time there was a desire to see them lend more to help the US continue to recover from the recession. No bill with any chance of passing ever moved through Congress.

Years later, the Trump administration would try again. There was a gauntlet of proposals from unwinding them to getting a fee for the government who is effectively insuring the mortgages with the “full faith and credit” backing. There was a notable change, which was that Fannie and Freddie were allowed to start to retain their earnings (instead of paying them out to the government), which can be interpreted as a potential step towards getting them out of conservatorship. However, little else was decided and the issue remains in its same purgatory-like state.

Generally speaking, there is both a desire in Washington to unwind them because they have grown quite large and caused a lot of headaches during the financial crisis. Some also see the GSEs as benefiting wealthy real estate owners with cheap capital. However, at the same time, they also rely on the GSEs to push affordable housing policies, with a portion of loans (50% most recently) earmarked for borrowings that meet affordability requirements. There has also been a push to use the GSEs to promote more green building requirements. And true to their original aim, there might be far fewer providers of 30-year fixed rate mortgages without them (although this pertains more to single family, not multifamily borrowings).

Whether intended or not, the GSEs are now a huge part of the financial system, which makes any large changes seem more unlikely. Since 2008, Fannie and Freddie have grown the amount of multifamily debt outstanding they own or guarantee from 25% to ~45%. Any changes to the system could have broad, sweeping implications that would be hard to tease out a priori. The political will that is required to reform the GSEs just doesn’t seem to be there now. It seems like an issue with little upside for getting right and a lot of downside for getting wrong, which perhaps explains why they have remained in receivership for about 15 years now.

We will now go through some operational differences between Fannie Mae and Freddie Mac.

Fannie Mae works on delegated underwriter model whereby they allow DUS (Delegated Underwriter Servicers) to originate loans on their behalf. With Freddie Mac, loans are effectively underwritten twice, the first time by Walker & Dunlop and the second time by Freddie. Fannie Mae’s DUS model is a little more efficient and quicker for Walker & Dunlop. One of the contingencies that comes with being a DUS lender though, is risk-sharing.

Walker & Dunlop is on the hook for the first 5% of unpaid principal when the loan defaults. Any losses above 5% of the unpaid principal, Walker & Dunlop shares a % of the excess loss with Fannie Mae, with a max loss cap of 20% of the unpaid principal. The idea behind the risk sharing is to ensure the loan underwriting standards are withheld, and by putting WD’s skin in the game, they hope to avoid unscrupulous lending. In return for the risk-sharing, Walker & Dunlop is compensated with a higher than typical servicing right fee (our channel checks have put it at around ~40bps, which is around twice that of Freddie’s and perhaps ~4x that of a large, non-GSE loan).

While the risk-sharing does expose them to potential losses, historically it has been de minimis. From the period following the financial crisis, Walker & Dunlop incurred, cumulatively, only losses of 16 basis points (that is as a % of their loans that have risk-sharing). Last year, they were the #1 Fannie Mae lender with $11.4bn in volume. Assuming that same loss rate as the worst financial crisis in recent history would yield a loss to WD of just $18mn (or ~4.4mn annually over 4 years). With such a low loss rate, it makes sense that their loan loss provision has averaged just ~$3mn.

Walker & Dunlop has risk-sharing agreements that stem primarily from their Fannie Mae business.

Below, Willy comments on the credit culture at Walker & Dunlop that has driven such low loss rates historically.

The risk-sharing agreements on a loan seldom change, but the terms of a loan are set taking into account fluctuations in interest rates. The GSEs publish the interest rates of their products daily and thus firms do not really have to compete amongst themselves to “win” that business. Instead, the brokers just need to find borrowers and ultimately it is the GSEs that will dictate how competitive they want to be. It is typically for them to move in and out of the market according to the caps they want to hit (we explain what caps are below). They may be very aggressive in originations at the beginning of the year, but then slow down their originations (by posting less competitive interest rates) in the middle of the year as their caps are close to filling up. Then, at the end of the year, they may swing back into the market to make sure they fill their caps. This dynamic is why Walker & Dunlop has always noted that analysts cannot really look at their business on a quarterly basis, as it results in a lot of volatility in their volumes.

The governing body of the GSEs, the FHFA (Federal Housing Finance Agency), started setting limits to the GSEs’ lending as early as 2015 to avoid having them crowd out private capital and remove liquidity from the market. However, broad exemptions followed the caps for affordable units and resource efficiency. In 2018, the multifamily lending caps were $35bn for each Fannie Mae and Freddie Mac, but they each originated more than double that because of exclusions, making the caps ineffective.

This chart is from the FHFA and shows how many loans were originated beyond the caps because of the green energy and other exclusions.

In 2019 the caps were unchanged, but the exclusions still existed and again they significantly exceeded the caps with a total of ~$150bn originated between the two. Towards the end of 2019, they moved the caps to $100bn each, but over a 5-quarter period (which is a similar run-rate to the prior year, if you include the loans that had “exclusions”). In conjunction with the much larger cap though, they eliminated the loopholes that were driving originations above the cap and instead set a requirement that at least 37.5% of loans had to be “mission-driven”.

After the 5-quarter cap ended in 2020, they moved back to a normal 4 quarter cap and set it to $70bn each, with no exclusions. In 2022 they increased it to $78bn each, but now with 50% needing to be “mission-driven”. In 2023 they cut the caps slightly to $75bn to “reflect an anticipated contraction of the multifamily origination markets in 2023”.

Taking the full year 2022 multifamily origination figure of ~$460bn, each Fannie Mae and Freddie Mac had a ~17% share of the total multifamily market, or about ~35% total.

In 2022 both GSEs’ volumes came under their $78bn cap, with Fannie Mae at $69bn and Freddie Mac at $74bn. Walker & Dunlop’s volume with Fannie Mae and Freddie Mac was $11.4bn and $6.6bn, respectively, for 2022. This means that WD’s market share of Fannie originations were ~17% ($11.4bn/ $69bn) and ~9% of Freddie’s ($6.6bn/ $74bn).

This pie chart shows the sources of commercial real estate debt outstanding. Agency and GSE is just 21%.

With an understanding of how the GSEs operate and the broader multifamily market, we will now dive into competition.

We bucket Competition into three categories: 1) Real Estate Brokerage Firms, 2) Banks, and 3) Non-bank Financial Institutions. However, some “competitors” can also be enablers in certain circumstances due to the fact that they have a brokerage network as well as proprietary capital. On the brokerage side, they are just connecting lenders and borrowers, so it is possible their debt brokers are helping a bank or life insurance company get a deal, even if Fannie Mae is one of the sources of capital (which would pay WD the most). Conversely, a non-Walker & Dunlop broker may end up bringing WD a deal that they arrange the financing for, even if in other contexts that broker competes with their own for clients.

Generally speaking, the big Real Estate Brokerage firms are considered their most direct competitors, which includes JLL, CBRE, and Berkadia. We will expand on each of those businesses below, but their competition set is much wider than that and varies based on the service. These peers tend to be competitive not just for debt brokerage and property sales, but also in competing for the best talent.

CBRE: CBRE, founded in 1906, is the largest global commercial real estate services and investments firm. They generated $30bn of revenue in 2022 and have a market cap of $23bn, which was ~$34bn at peak. Their global business is much more expansive than Walker & Dunlop’s in terms of both geographical footprint (44% of revenues are ex-US) and services.

In addition to capital markets, mortgage services, and investment services, they also have leasing services, property management, facilities management, and development services, among others. In 2022 their origination and sales volume was ~$63bn, which was the same as Walker & Dunlop’s. Their loan servicing portfolio exceeds $380bn, which is substantially larger than Walker & Dunlop’s ~$134bn though. CBRE (and JLL) both have large management businesses, including property management, facility management, and project management. These make up a large portion of their revenues, but are smaller margin compared to the core capital markets business (these three represent ~$9.5bn in revenues). A lot of their businesses are “cost plus” businesses where they hire 3rd parties to render various management services, and as a result, ~$12bn of their revenues are pass through costs recognized as revenue. Even backing those out though, their revenues would be ~$19bn with EBIT of ~$1.5bn for an 8% margin. While Walker & Dunlop may have only 6% of CBRE’s revenues, they have 18% of CBRE’s EBIT because of their focus on more profitable businesses.

JLL: JLL, or Jones Lang LaSalle, is one of the largest real estate services firms with ~$21bn in 2022 revenues and a current market cap of ~$7bn (or $12bn at peak). While technically founded in 1999 with the merger of Jones Lang Wotton and LaSalle Partners, Jones Lang Wotton dates back to 1783. They are a global firm with ~40% of the business outside of the United States. They have a much more expansive service offering than Walker & Dunlop that includes property management, leasing, construction oversight, real estate software, and various on-premise real estate services.

Their loan servicing portfolio is just slightly larger than Walker & Dunlop’s, at ~$134bn. However, they also report $286bn in client transactions (unclear exactly what they count in that) versus Walker & Dunlop’s $63bn (which is debt financing volume of ~$44bn plus property sales of ~$20bn).

Berkadia: Berkadia was built from the seeds of GMAC Commercial Mortgage Corporation, which was established as a separate operating unit in 1994. It became a large real estate finance company, and in 2006, private equity bought it out and changed its name to Capmark Financial Group. In 2009, Capmark Financial voluntarily filed for bankruptcy. That same year, Berkshire Hathaway and Leucadia (which was later bought out by Jefferies), created a joint venture called Berkadia. Berkadia then purchased Capmark Financial Group’s assets out of bankruptcy.

Financial performance on Berkadia is sparse, but in 2021 they had a $337bn loan servicing portfolio, which is larger than WD’s. Their investment sales volume was $27bn and debt financing was $40bn in 2021, which compares to Walker & Dunlop’s $19bn and $47bn, respectively.

In some sense, Berkadia is WD’s most direct peer because they also are decently focused on real estate finance without large property management businesses. However, they still sit within two larger conglomerates (Berkshire and Jefferies), which means financial outperformance is relatively inconsequential to their larger organization and they cannot incentivize with equity ownership.

To be blunt, there are no notable moats to any of Walker & Dunlop’s offerings and for virtually every service they provide, there are numerous alternatives. The GSE lending business is arguably the only business with a barrier to entry as Fannie Mae and Freddie Mac restrict how many lenders are approved to work with them. However, even in that business, there are still over 20 approved lenders for each, which translates to plenty of competition.

Willy Walker has argued that the only way to stand out in a commoditized business is to be #1. However, once a brokerage firm is well known and ranked amongst the top, like Berkadia or CBRE, the individual client’s relationship matters far more to win and keep business. Firm reputation may be important to hire talented people and, in some cases, can help convince clients to work with you, but most professional real estate investors will have several lenders that they have worked with in the past. It is ultimately a relationship game, where the firms that can draw the best talent, and retain them, will do the best.

Willy launched a webcast at the beginning of the pandemic that ultimately ended up being very popular, resulting in growing brand awareness and mind share.

Still though, it is very common for a borrower to have 2-3 lenders that they regularly work with and rotate between. From the real estate investors’ perspective, it is relatively easy to get multiple people to send them quotes for a deal, which will help them know they are getting the best terms available. They may, over time, prefer the Walker & Dunlop debt broker, but anyone who works on multiple deals a year is going to want to have more than one debt broker.

Walker & Dunlop consistently ranks as a top Fannie Mae and Freddie Mac lender. They historically are stronger with Fannie Mae, though.

While Walker & Dunlop has a very high service level, other top real estate service firms (JLL, CBRE, Berkadia) are on par with them. The large banks though, like Wells Fargo, arguably have worse service levels because they are very bureaucratic, which can make them slower to process a deal. The Regional Banks are a little nimbler and can be particularly strong within a niche of their geographical footprint, but by and large do not have the focus or reputation to draw and support top talent. What really sets the banks apart though, is the ability to offer other banking services in return for their lending business (for instance, offering higher yields on deposits). Given the competition for each deal, there is fee pressure on the loan economics. Historically though, it seems to have a limit.

Walker & Dunlop makes money off a loan in three ways: 1) The loan origination fee, 2) a gain on sale if the loan was originated at a higher interest rate than what they can sell it at, and 3) the MSRs. Depending on the type of loan product and clientele, the economics can differ. Sometimes they will reduce their loan origination fee but increase the interest rate so it roughly nets out to the same if the borrower prefers that. The origination fees and the interest rate are the most common way lenders will compete with each other.

However, despite heavy competition, the industry’s fee rate tends to hold up fairly well, particularly amongst top bracket brokers. If the borrower keeps hammering Walker & Dunlop on the fees, at some point they will just pass on the deal. This is in part because a broker may not consider all the time-intensive paperwork to be worth the fee, but also because they generally feel like they know their value.

It is similar in investment banking, where the typical 7% fee for an IPO may be negotiated down, but Goldman Sachs and Morgan Stanley aren’t going to race to zero for it. At some point, they would feel the fee is beneath them and devalues their services. A small boutique with a new investment banking team might be happy to get the deal done at 2%, but would you really want to hire them? There is some aspect of getting what you pay for, and if you want a top firm to handle your business, they will only capitulate on the fees to a point. This is what has helped keep these services fairly lucrative despite having numerous alternatives.

In general, origination fees start at 100bps but drop off as the loan size increases. A $100mn loan may only have a 35bps origination fee. Fee competition typically happens on the margin. Maybe a $40mn loan with a 70bps origination fee gets competed down to 65bps or 60bps, but a reputable firm is not going to drive it down to 30bps (unless they make it up elsewhere). For the borrower, a slight decrease in the origination fee isn’t nearly as important as actually winning the deal, and having a more reputable financing partner can be critical to winning the deal. In fact, it is common for a seller to take a slightly lower price for their property in exchange for higher confidence that the deal actually closes, which helps give the real estate financing partner some leverage.

The other competition-constraining factor is time. Every top broker tends to have more activity than time, so it is easier for them to hold out on fees. However, when the market turns and deal activity slows, this creates an environment where not only does deal volume slow, but incentivizes fierce competition on fees in order to win the little volume that there is. This may result in a double whammy of simultaneously lower volume and lower fees.

The other part of the loan economics is the MSRs. For GSE loans, the mortgage servicing rights are far richer. This is in part because it is how they compensate the firms for sourcing deals for them, but also because they wanted to make it economic for smaller firms to participate as well. The MSR fee will vary based on the type of loans and duration, but generally speaking for Fannie Mae they bill ~40bps annually of the loan outstanding. This is the highest of any mortgage servicing rights because it is in part compensation for risk sharing on Fannie Mae loans. Freddie Mac also varies, but ballpark might be about 30-50% less than Fannie Mae. This is still higher than non-GSE MSRs which may be in the high single digit to low 10s bps range.

Despite Walker & Dunlop operating in a “commoditized market”, they still have gained material share with the GSEs and in other areas of real estate services while maintaining a similar economic profile. Frankly, it seems like a bit of an enigma because you would expect commodity markets with intense competition to eventually drive returns down, but their operating margins have stayed well over 20% (and usually closer to 30%) with competitors also seeing similar margin profiles in their brokerage businesses.

From what we can tell, this is simply because the top firms seem unwilling to negotiate on fees beyond a point. While theory might dictate that the industry fees should be whittled down over time towards their marginal cost of product, fees have remained remarkably resilient for many decades.

Walker & Dunlop’s competitive position has improved over the past decade though. As they became a bigger organization with more name recognition, it in turn made it easier to attract talented employees, which in turn helped grow the organization further. Relative to peers, Walker & Dunlop is considered to be a more flexible and less bureaucratic organization, largely owing to its smaller headcount. They hit above their size though, with revenue per employee that is almost twice their closest peers at close to ~$1mn. Their proprietary sources of capital that they added also start to differentiate their services. The bridge loan program on non-stabilized properties is a helpful solution for clients that not all debt brokers can help with. Even so, it is an industry with many competitors and minimal competitive advantages, but an industry where fees have broadly remained resilient.

Willy Walker has continually expanded the business into new areas beyond their original GSE-focus. Most recently, Walker & Dunlop has made acquisitions into the appraisal industry in combination with the acquisition of GeoPhy and more recently into Low Income Housing Tax Credit (LIHTC) Syndication with Alliant Capital. We will say a bit about each new business line below.

Small Balance Loans. Walker & Dunlop has traditionally focused on commercial originations >$10mn with an average loan size of ~$20mn. Typically, it is only professional real estate investors that are borrowing large loans, and thus the client base is more sophisticated and easier to work with. This is particularly important for GSE loans which have a lot of requirements like monthly P&L statements that smaller real estate owners usually are missing.

Part of the push to focus on more small loans comes from Fannie and Freddie directives, but those have been there for a while. The real impetus is that Walker & Dunlop realized this was a strategic offering. Loan brokers noticed that some small loan borrowers were borrowing loans that were much larger in a few years’ time, but since they were already working with lenders, it put WD at a competitive disadvantage to win that business. In other circumstances, they were seeing that even their big real estate borrowers would occasionally have a small deal and they had to turn these core customers away to competitors.

WD’s original hesitation with small balance lending is that it would be much less profitable as they tie up valuable employee time with low value deals. However, they have recently acquired and implemented more technology to streamline this process. The hope is that by leveraging technology they can help service more small balance borrowers in a shorter amount of time to make up for the poorer economics.

If successful, this could be a nice business as these borrowers tend to pay full ~100bps origination fees versus larger loans where that slides down. However, smaller borrowers also tend to be more inexperienced and could come with a higher default risk. They plan to grow this business to $5bn by 2025 in annual originations, up from about ~$700mn at the end of 2022. So far, it seems that volumes have grown slower than they would need to for them to hit their goals. Either way, we see this business as incremental to their overall earnings power.

Apprise. Apprise is the name of their appraisal business that they launched 4Q19 in partnership with data focused firm GeoPhy (which they later acquired in 2022). The idea was to leverage data to offer property appraisals in under five days, instead of the typical two to three weeks. Appraisals are typically required for lending, sales, or for accounting purposes of larger organizations. They have set the goal of making $100mn in revenue off of their Apprise business, but it has lagged behind with only ~$11mn (+14% y/y) thus far.

Zelman & Associates. This is a real estate research and investment banking firm that WD acquired 75% of in 2021. They wanted to use their research to help grow their sales platform as well as debt financing businesses. Separately, they want to grow the investment bank and hope to benefit from synergies of plugging their clients into Walker & Dunlop’s network.

Investment Management. Walker & Dunlop first entered this business through the acquisition of JCR capital in 2018 (which they later renamed to Walker & Dunlop Investment Partners). They had $800mn in AUM and were in the process of raising their 4th fund at the time of acquisition. The idea is to create synergies between their brokerage network that can find properties for the investment fund, while also allowing WD brokers to have proprietary capital. The fund business also creates a steady stream of cash flows to help balance out their deal-driven earnings. They had the goal of reaching $8-10bn in AUM by 2020 but were unable to get there with JCR alone.

However, in 2021 they acquired Alliant Capital and their $14bn of AUM, which helped push their AUM to $16bn by 2022, exceeding their five-year goal (detailed later). This was their largest acquisition thus far at ~$700mn, including the assumption of $155mn of debt and a $100mn earn out (They issued $90mn in stock—which was at a much higher price at the time– and covered the rest with cash).

Alliant Capital specializes in a subsector of investment management called LIHTC (or Low-Income Housing Tax Credit) syndication. As briefly mentioned prior, when a real estate owner leases out a certain portion of their building at a below market rent (which is based off of a % of the area’s average median income) they receive Federal Income Tax Credits. These tax credits can then be purchased by investor groups, which is what Alliant facilitates. As Fannie and Freddie increasingly focus on affordable housing, this creates natural synergies to help arrange the financing of a property, as well as syndicate out the tax credits.

These businesses all help increase the Walker & Dunlop value prop – and have the potential to be large profit centers in and of themselves.

Walker & Dunlop gaining material share and growing their business over ten-fold since 2010 shows how fragmented the industry is with relatively fluid competitive positions. Ultimately, it’s a people business, and people can always move, which means industry share is never settled.

Willy Walker is the first to call their services commoditized. With many competitors offering essentially the same product, ultimately it is just 1) the brokers’ relationships and 2) their reputation for consistently executing deals competently that set them apart. For years after Willy took over, there were nothing but small wins to slowly progress the business forward. The more they worked on the back-end execution and supporting their deal makers, the better able they would be to draw in more talent. The more top brokers they could bring on, the more they could grow the business and their name recognition, which in turn helped to bring on more top brokers.

While Walker & Dunlop was slowly building up their firm’s name and reputation, they simultaneously started embarking on tangential offerings to buttress their relatively commoditized, but very profitable core GSE business. Layering in a bridge loan program that could finance unstabilized properties contingent on doing the refinancing when they were stabilized would incrementally help them win more deals. Partnering with investment sales firms and later building their own in-house sales operations allowed them to funnel some deal flow into the financing business. The Blackstone JV gave them proprietary capital to win deals with. Their GeoPhy acquisition built up their real estate data and analytics capabilities in order to better support investment sales. No single initiative on its own made a material difference competitively, but the sum of each effort helped Walker & Dunlop offer a better, more holistic service, while continuing to attract top talent.

Even so, it is a fiercely competitive industry. Every borrower typically has 2-3 lenders they work with who in turn may source 5-7 bids each for every deal. With such a level of competition, one might think that fees get bid down to a de minimis figure, but the economics of the industry have been withheld throughout time. This is largely because those most likely to compete on fees are those with the least opportunity cost, who are disproportionately younger, less experienced brokers. Certainly, for some buyers knocking another 15bps off the origination fee is paramount, but for many more experienced real estate investors, they value working with someone who will not drag out the process by continually returning for follow-up information they didn’t know to originally ask for. They also know that having a financing package from a more reputable firm will make their offer stronger and help them win the deal in the first place.

Now, to be clear, when there are multiple reputable lenders, they will compete on fees. But it’s only to an extent. Berkadia may drop their origination fee from 70bps to 60bps against Walker & Dunlop’s original offer of 65bps, but neither is going to match a Northmarq or NCB bid for 40bps (both are small GSE originators that do not make the top 10 lenders list).

Price is negotiable, but at some point, the top brokers “know their value” and are unwilling to engage in a race down to the bottom. For an experienced real estate borrower, they probably interpret a low bid from a broker they have never worked with as indicative of their quality of service.

Thus, on an industry level, the fee rates of the top brokers have historically been preserved, which has allowed Walker & Dunlop to compete on service. Their advantage is size and focus. Wells Fargo, who is consistently in the top ten Fannie Mae lenders, has never proclaimed ambitions of being #1. With over $100bn of home loan originations in 2022 alone (which itself is down 50% y/y), the ~$5bn more of financing volume Wells Fargo would need to take the #1 spot is trivial to their bottom line. Willy Walker in turn is willing to get on the phone or fly to a client himself to win a deal. While Wells Fargo may be able to compete in ways Walker & Dunlop can’t by offering higher interest rates on deposits for instance, the big bank’s size and bureaucracy has made them unwieldy competitors.

Other real estate finance companies like JLL and CBRE are strong competitors, but they too have a motley of global businesses and a massive work force of over 100,000 each. Berkadia is a more focused competitor, but they are just a piece of two much larger businesses, and that tends to show up over time in how the business is prioritized. Still though, altogether their total market share of commercial real estate is small enough that Walker & Dunlop can take share elsewhere.

The smaller real estate finance players, of which there are many, may be focused, but they lack the other capabilities that Walker & Dunlop has built out, in addition to a reputable brand, which makes them competitively weaker.

Whatever conclusions you want to make on the intensity of the competitive environment and Walker & Dunlop’s “right” to win, you have to take into account their history: they have grown market share from essentially 0% to 2% of commercial and 8% of multifamily originations, so clients clearly must see a strong value prop in their offering. In some sense, it is like Constellation Software, which does not have much, if any, competitive advantage in buying companies. As Mark Leonard says, the barriers to entering the business are a telephone and a check book, but yet they still consistently win and have a phenomenal track record. Just because many other firms vie for that business with ostensibly similar offerings, there is informational value in the fact that a single firm—Walker & Dunlop—continually wins their disproportionate share of deals.

Walker & Dunlop grew total originations, whereas all but 1 of the other top 10 brokers failed to in 2022.

We will not overstate the moats of the business – there are few. But by the same token, it is hard for someone who started in WD’s position circa 2003 to grow to the position they have, not only because they would have to compete against Walker & Dunlop, but WD has already acquired many of the companies that a newcomer would want to buy as a wedge.

Do not confuse the company’s moats with the brokers’ moats, which are formidable – it is very hard to tear away a long-time client from a seasoned and trusted broker. While this means that the broker’s negotiating leverage and ability to extract economic value from a company is high, it is already accounted for. Half, if not more, of all origination fees goes to the broker. Star teams will always have the option to go to other firms, but if they want to keep producing at the level they are, they will have to settle at one of the few large firms, and Walker & Dunlop’s culture stands out here.

The Walker Way, with a “no jerks” policy and a less internally combative environment where trying to steal a colleague’s client is sure to get you fired, is generally preferred to others’ wheeling and dealing cultures. The culture also helps keep the company’s many different divisions integrated, which is critical to ensuring synergies between property sales and financing, for instance. While the firm has grown to ~1,500 employees, Willy still tries to give it a smaller company feel by writing everyone hand-written birthday cards. He also stays accessible so if an employee needs to get him on the phone to win a deal, it happens. This all shows up in company rankings that consistently place them as a top firm to work at.

It should not be missed that Willy’s instincts for where to drive the business have been spot on so far. In the depths of the financial crisis, with both GSEs in conservatorships and their status totally unknown, Willy doubled down on the business with the Credit Suisse acquisition of Column and then with CW Capital. At the time people thought it was only a matter of time before the GSEs were dramatically reformed or disappeared entirely, but Willy was willing to go against group think. Simultaneously though, Willy worked to reduce their reliance on the GSEs and grow their service offerings. In his roughly two decades as leader, he has grown the value of the firm 100x.

The business itself suffers from cyclicality owing to the boom-and-bust cycle of the real estate industry. However, the GSEs’ capital is more resilient, and typically grows in time of turmoil. Their mortgage servicing rights are penalty protected, giving them a contractually obligated ~10-year stream of cash flows. Their new investment management division has captive funds which they can take recurring fees on. And no matter how bad the real estate market may be, borrowers will always need to refinance their loans when they come due.

Despite Walker & Dunlop’s business growth and expansion, Willy is only 54 and unlikely to stop anytime soon. They have goals to reach $2bn of revenue and $13-15 of GAAP EPS by 2025. After that, we can expect a new set of long-term goals.

However, the business is not without notable risks. We will now first go through the revenue build and valuation and then conclude with risks.

We start our revenue build with Walker & Dunlop’s 2025 target goals. As mentioned earlier, they set these long-term goals every five years. They are somewhat reach goals that are not “padded” like most companies’ guidance. In their last set of goals for 2020 (called Vision 2020) they missed on two of the five (Hit loan originations of $30bn, >$100bn in servicing portfolio, and total revenue of $1bn. They did not meet their $8bn property sales goal or $8bn in AUM).

Above we see the volumes they are targeting, and below they tie that to revenue.

The revenue build below is informed by Walker & Dunlop’s Drive to ’25 goals. We show the low to high range of their estimates. Note that their revenue goal for Debt Financing includes the FV of MSRs, which as discussed, is an accounting requirement that does not constitute a real revenue in that period. To account for that in our build, we come to an adjusted revenue figure which excludes the FV of these MSRs. In order to do that, we take the Debt Financing Revenue goal of $850-950mn and subtract the FV of MSRs to get to the loan origination and debt brokerage fee.

We had to make an assumption as to the FV of MSRs as a % of volume to do that though. They also grouped several businesses (WD Investment Partners, Investment Banking, and Small Balance Lending) together in their last revenue target of $275-400mn, but those goals do not comport with their revenue segments, so we had to recategorize them (the Investment Management Fees were kept as a separate line item and backed out of that $275-400mn target). For historical periods, we put in “Other Revenues”, all revenue items that were not included above (i.e., Net Warehouse Interest Income, Escrow Earnings, and the “Other Revenue” item).

All in all, they are targeting $1.7-2bn of revenues or about $1.34-1.64bn of adjusted revenues, which backs out the FV MSR gain. From 2022 adjusted revenues of $1.1bn, this implies a 9-16% revenue growth CAGR. This is particularly ambitious given that revenues are likely to decrease in 2023 (it is hard to gauge revenues on a quarter-to-quarter basis given the lumpiness of originations, but 1Q23 adjusted revenues are -22% y/y).

Despite revenue growth lumpiness, they have grown revenues at a 20% CAGR since 2010 and a ~14% CAGR since 2015.

The slide below shows their progress relative to 2022 numbers. They have already hit their AUM goal, driven by the Alliant acquisition, and made strong progress on property sales and debt financing volume, but are quite behind on small balance lending and Apprise revenue. The two areas they are most behind on are least impactful to their revenue growth target of $2bn though.

Note that WD’s targets imply the servicing fee rate on the portfolio increases slightly and the origination fees are up in the top end of the range. For conservatism, we would want to assume the servicing portfolio fee rate compresses as a lower portion of incremental loans are Fannie Mae (which carry the highest servicing fee rate). However, this is offset by the fact that they do not seem to be including Escrow Earnings and Net Warehouse Interest Income, which were $30mn in 2021 and $69mn in 2022.

Below, we take the low end of Walker & Dunlop’s drive to ’25 revenues and adjust the FV of MSRs out and round down to get the low-end of the adjusted revenue range at $1.3bn. This compares to $1.07bn in adjusted revenues for 2022.

We then apply an adjusted margin range (that does not include amortization of MSRs) of 22.5-30%. Since 2015, their adjusted margins have averaged 26%. Excluding 2020, where margins were 17%, the lowest adjusted margins were 24% and the highest was 34%. Certain business lines like loan originations and investment property sales should have higher incremental margins than the corporate average, but newer businesses like Small Business Lending and appraisals are likely lower as they ramp those up. We hone in on the 25-27.5% adjusted margin scenarios as being fair.

Instead of a reverse DCF as we typically do, we have opted to use a multiple approach. The reasoning for the reverse DCF is to see what the implied assumptions are for the business going far out in the future. However, if you only need to go out a few years to see that the owner’s earnings yield to cost is already producing an adequate return, then we feel it is unnecessary to extrapolate out the value implied in the retained earnings.

In other words, at the low end of the revenue range with a 25-27.5% adjusted margin, we are getting $7.70 to $8.50 in adjusted EPS. At today’s market cap of ~$2.6bn ($78 stock price), an investor would be earning over an 11% return on their stock purchase over roughly three years (includes annual dividends of $2.52). As long as the earnings they retain are invested in a non-value destructive way, this is the investor’s floor on their return. The discount rate output in a reverse DCF would attribute value to the retained earnings portion, but under this method you are effectively attributing zero value to Walker & Dunlop’s ability to grow retained earnings accretively (which they historically have).

To get a sense of how retained earnings have compounded historically though, we show ROE since 2015 below. ROE has averaged ~13%. However, if you back out the MSR asset, which is an accounting creation with no commensurate capital outlay associated with it, the average jumps to 63%. The decrease in ROE (less MSR) has been driven mostly by the fact that if you back out the MSR, the equity base is very small and minor increases in retained earnings have a disproportionate impact. Still though, a 30% ROE is very strong, which makes sense as their core business is high margin and requires minimal capital. If they lean into more balance sheet lending, then we would expect this to fall. Clearly though, their retained earnings have historically created value.

Aside from internal investments and acquisitions, they have paid out ~40% of earnings in dividends (~$80mn for 2022) and have embarked on periodic stock buybacks. Their stock repurchases have roughly matched their equity issuance and they tend to prefer growing to stock repurchases. This does not seem likely to change in the interim as their stock repurchase authorization is $75mn right now, which would constitute just a <3% reduction in shares.

We use ROE instead of ROIC because their warehouse activities and loan origination activities create balance sheet distortions that are not indicative of the actual capital “invested” in the business. For most financial firms, their “working capital” is recorded as items that are traditionally included in invested capital.

To show a range of returns though, we apply a 10-17.5x multiple. (You can think of the multiple as a valuation if it is what you would pay for it, and a pricing if it is what you think someone else would pay for it). This yields a wide range of annual returns from 0-26%, which includes the dividend yield of 3.2%. We circle in on the 12.5x to 15x multiple scenarios, which seems like a fair valuation (although some investors may argue that the implied return is too high to consider it fair). It is on the investor to figure out what they consider to be a fair valuation for Walker & Dunlop, but some guideposts are below.

We do not explicitly give any value to retained earnings over the next ~3 years, since we are essentially assuming the 2025 incremental earnings growth is funded through retained earnings. We calculate the annualized return assuming a 3.5-year period (which suggests the multiples are “TTM”), but an investor can adjust the calculation as they wish. Of course, it is possible they also achieve these results later than 2025. Adding a year to the annualized return scenarios highlighted lowers the TSR from 9-18% to 8-15%, and adding two years lowers it even further to 7-13%.

Below we run the same math again, but for the high end of Walker & Dunlop’s revenue estimates. A conservative investor would not want to assume this, but we show it as an investor is free to make their own judgements.

These assumptions assume all goes well for Walker & Dunlop, but of course, as with any investment, there are risks.


1) GSEs cut volume: One of the biggest risks is that Fannie Mae and Freddie Mac leave their annual caps at current levels. They reduced their limits this year by $3bn to $75bn, but that was fairly trivial as neither Fannie nor Freddie hit the prior year’s $78bn limit. In 2013, they cut the caps 10% without warning, and it adversely impacted Walker & Dunlop. A much larger cut could not just make reaching their 2025 goals impossible, but permanently pressure their business model.

2) GSEs cap Walker & Dunlop. While there is no stated maximum for a single company’s volumes with Fannie and Freddie, that could change if a single company becomes too large. It is thought that the GSEs would not allow a single company’s market share to exceed much beyond 20% (simply because the highest observed market share for a full year was CBRE around 22%). So, in theory it is possible for their market share to continue to grow, but they are nearing the “speculated” limit with Fannie at 17% for last year (Freddie is lower at 9%).

If their market share is limited, then WD will become more reliant on them raising the caps. In 2019, 61% of WD’s volume was Agency Lending, but that fell to 41% by 2022. As noted, GSE loans (particularly Fannie Mae) come with more lucrative servicing fees as well as typically higher origination fees. The more they are forced to grow volumes with non-GSE volume, the more revenue growth will be pressured. A growing mix of brokered loans would mean their origination fees will be pressured and it will be increasingly harder to maintain the same servicing fee rate on their servicing portfolio.

2) GSE Reform or Elimination. If the GSEs are reformed, eliminated, or replaced, it could spell disaster for Walker & Dunlop. Their whole business model places their GSE origination abilities at the center of the business and is by far their strongest service offering versus competitors. However, as we mentioned, there does not seem to be any political will to eliminate the GSEs now as they are a large part of the mortgage market and elimination could do anything from moderately pressuring the economy to distorting the entire debt market. If the GSEs were replaced by similar institutions, that could also be disastrous for Walker & Dunlop, as a reshuffle could allow new competitors to weasel into the top ranks. The outcome from reform could range from adverse to neutral or positive. While there is no current push to change the GSEs and no agreement on how they should be changed, it seems somewhat inevitable that there will be some changes to the GSEs eventually. This remains the largest risk to WD.

3) Culture erodes as they scale. Walker & Dunlop’s culture is an important differentiator for them to attract top brokers and talent from competitors. Many like the small company feel and less bureaucracy that WD has. However, as they scale, it will be harder and harder to maintain their culture that has allowed them to attract brokers in the first place.

4) Key Man Risk. Willy Walker has been instrumental to growing Walker & Dunlop into what it is today, and it is not clear how much the business and culture would change if he left. Howard Smith is another important figure who is likely to leave soon. He is President of Walker & Dunlop and has worked there for over 40 years. He was critical in teaching Willy the business and losing him would be a loss of a lot of institutional knowledge.

5) Credit Losses are worse than expected. It only takes a relatively small amount of losses to wipe out the capital base. While this is true for any lending operation (or originator that retains risks-sharing), their history has suggested they are highly prudent with credit losses.

6) Big Teams leave. The business is only as strong as their people, and if teams defect for any reason, it will adversely impact WD.

7) Capital Requirements Change. Capital requirements change to make them carry more capital in their lending operations.

8) Prolonged Downturn. The longer the downturn, the more risk the business becomes impaired. WD has already laid off about 8% of their workforce, and if the real estate market slows more, they will need to cut more employees. Laying off employees will impair their ability to grow volumes when the market returns and could create ill-will with existing employees as well as turn off potential future applicants.

9) High Interest Rates Make Alternative Sources of Capital More Competitive. As interest rates increase, lending becomes more attractive, and it is possible new sources of capital enter the market. This would be a problem if these new sources of capital made the GSEs uncompetitive, or if other brokers had a stronger relationship with these sources of capital. For instance, Berkadia is 50% owned by Berkshire Hathaway, and if Berkshire’s insurance operations decided to allocate more capital to real estate loans, then those deals could become “proprietary capital” for Berkadia.

10) Internet-based Platform Disruption. The brokerage business has remained very fragmented with most real estate borrowers needing to go through 1-3 intermediaries to feel they have a sense of the market and get the best deals. In theory, if there was an internet marketplace for borrowers where they could put their requirements and just receive bids without a middleman (think an eBay for debt), than that would intermediate WD and other brokers. A similar thing could happen (and does to a small extent) on the Real Estate sales side, where listings are public. However, they still require a broker to close the deal. If a platform tried to take that position, it could cut the real estate broker out of the sale of property as well. This is a risk to monitor, but on such large purchases, trust is important, and people tend to want to talk to someone.

Below is our Summary Model for WD. The model below is provided so investors can get a sense of how items flow through the P&L, but should not be taken as our estimates. There is a much broader range of potential outcomes than what we show below, and we do not spend much time in “predicting” a single year’s estimates, which we think is a futile exercise. However, the model can be useful to see how different assumptions change various outputs. Member Plus can of course download the excels and change the model as they wish.

Thank you for reading our Walker & Dunlop report, we hope you learned a lot! Please follow us on Twitter @Speedwell_LLC, Threads @speedwell_research, and subscribe to our newsletter at speedwellsnippets.substack.com for periodic updates on Walker & Dunlop and more business content. Feel free to reach out to us at info@speedwellresearch.com if you have comments or questions.

*At the time of this writing, one or more contributors to this report has a position in Walker & Dunlop. Furthermore, accounts one or more contributors advise on may also have a position in Walker & Dunlop. This may change without notice.