Merchants Bancorp (NASDAQ: MBIN) — Credit Heals Faster Than Capital Rebuilds

Published: 2026-09-14 · Verdict: Hold · Entry price: $44 · Price target: $61 · Research confidence: High (91%)

Executive conclusion

Analyst Take

Merchants Bancorp merits a HOLD at the September 14, 2026 closing price of $53.14. A disciplined accumulation level is approximately $44, and the central 15-month value estimate is $61. The prospective return from the current price is therefore only modestly above an estimated 11.5%-12.5% cost of equity, whereas the lower entry level would provide a meaningful margin against unresolved collateral, funding and capital risks. These are analyst estimates, not company guidance.

The central investment tension is straightforward. Merchants owns a valuable specialist housing-finance platform spanning multifamily bridge and construction lending, agency permanent financing, retained servicing, residential mortgage warehousing, custodial deposits and affordable-housing tax-credit syndication. Its fifth-place 2025 ranking in Fannie Mae affordable-housing production, more than $25 billion of serviced or subserviced loans at year-end 2025 and $66.3 billion of 2025 warehouse funded volume demonstrate relevance and scale. Yet the same integration creates correlation. Higher rates can simultaneously impair property values, reduce permanent-loan proceeds, strand bridge exposure, weaken gain-on-sale activity, pressure mortgage-company customers and raise Merchants’ funding cost. The business is diversified by product but concentrated around one housing-finance ecosystem. [S3][S9]

The recovery evidence is meaningful. Q2 2026 net income reached $78.3 million, diluted EPS was $1.48, tangible book value per share rose to a thirtieth consecutive quarterly record of $39.93, criticized loans fell 12% sequentially and nonperforming loans fell 17%. Management disclosed that the NPL decline came primarily from loans paid in full, which is stronger evidence than a decline caused principally by charge-offs or foreclosure. The confidential memorandum of understanding with the FDIC and Indiana Department of Financial Institutions was terminated during Q1. These facts reduce the probability that the 2025 episode becomes a prolonged regulatory or solvency crisis. [S1][S2][S4]

The counter-case is equally concrete. The $43.8 million year-over-year reduction in provision accounted for most of the Q2 earnings improvement; net interest income increased 6%, noninterest income fell 10% and gain-on-sale income fell 44%. NPLs of $205.6 million were still above the $197.8 million reported at year-end 2025, OREO increased to $72.4 million and the $75.8 million allowance covered only 36.9% of NPLs. Nearly 95% of Q2 charge-offs came from two multifamily relationships. Only about 6% of criticized loans overlapped disclosed credit-risk-transfer protection. Brokered deposits rose 46% sequentially to $1.3 billion with a 51-day weighted-average remaining maturity, while total capital declined to 12.5%. [S1][S2][S3]

Valuation is no longer distressed. Correctly rolling reported common earnings—including preferred-redemption effects—produces approximately $228.7 million of trailing common income and $4.98 of diluted EPS, rather than the slightly higher common-income figure obtained by subtracting preferred dividends alone. At $53.14, the stock trades near 10.7 times trailing earnings and 1.33 times company-defined tangible book. The peer group’s roughly 9.6-times earnings median points below the market price, while its approximately 1.67-times tangible-book median points materially above it. That disagreement is the valuation thesis: Merchants looks inexpensive per dollar of book but not per dollar of currently demonstrated earnings. [S1][S3][S6][S16]

Investment conviction is moderate. Filing evidence is strong for financial statements, credit migration, capital, ownership, compensation and insider transactions. Fannie Mae, FHFA and MBA provide authoritative industry evidence. Evidence is weaker for forward management intent because Company Financials returned no earnings-call transcripts for the latest two quarters, and no public Q&A record was located. Release commentary is therefore treated as management’s hypothesis, not independent corroboration. The industry tailwind is also less powerful than an older forecast suggested: MBA’s April forecast projected 2026 multifamily originations of $422.1 billion, up only 2.2%, versus January’s now-stale expectation of 20.8% growth. [S6][S10][S25]

The call would improve if NPLs, criticized loans and OREO decline through cash repayment or economically sound takeouts; annualized charge-offs fall below 0.35% of loans; recurring fee income grows without favorable servicing marks; and CET1 and total capital rebuild despite tangible-book growth. It would deteriorate if NPL improvement depends on write-offs or OREO transfers, another small group of relationships produces outsized losses, normalized EPS falls below roughly $4.50, NIM drops below 2.6%, or new common or preferred issuance is required. The differentiated conclusion is not that Merchants remains broken. It is that the franchise has recovered faster than capital and credit uncertainty, while the stock has recovered faster than both.

Stock Price Action — Five-Year Event Map

The relevant price series must be adjusted for the January 2022 three-for-two stock split. On a consistent split-adjusted basis, the five-year closing low within the controlled September 14, 2021 through September 14, 2026 window was $16.04 on the first date of the period—not the $21.50 intraday low reached during the 2023 regional-bank panic. The highest close was $55.72 on August 3, 2026; $56.72 was that day’s intraday high. The current $53.14 close is 4.6% below the five-year closing high. Over the trailing 52 weeks, the closing range was $30.42 to $55.72, with an intraday low of $30.37, putting the current close roughly 90% of the way from the closing low to the closing high. [S6]

• 2022 rate shock. After trading above $32 on January 18, 2022, the shares weakened as mortgage activity contracted and deposit and borrowing costs rose. The price change is directly observed. Attributing it to the rate and mortgage cycle is an interpretation supported by the 2022 filing’s discussion of market contraction and the company’s asset-liability structure. [S6][S21]

• March-May 2023 regional-bank panic. The stock reached an intraday low of $21.50 and closed at $22.27 on May 4. Merchants did not disclose a contemporaneous institution-specific failure. The move therefore demonstrates sensitivity to regional-bank liquidity sentiment, not proof that Merchants suffered the same deposit dynamics as failed banks. A March 2023 release emphasized available borrowing capacity and rapid asset repricing, but that was management’s reassurance rather than an independent stress test. [S6][S21]

• July 2024 earnings reversal. The shares closed at $52.10 on July 29 and $44.26 on July 30, a 15.0% decline. Reported earnings remained profitable, but the reversal showed that investors were questioning durability, mortgage-cycle exposure or emerging credit risk before the full 2025 losses became visible. Price alone cannot distinguish those explanations. [S6][S7]

• October 2024 reversal. The shares closed at $43.89 on October 28 and $36.03 on October 29, down 17.9%. This preceded the large 2025 multifamily charge-offs and is consistent with rising concern about asset quality and operating volatility. It is not evidence that the market already knew the later relationship-level facts. [S3][S6][S7]

• April and July 2025 credit shocks. The stock fell 8.6% around the Q1 release and 10.8% around the Q2 release. These declines had direct company-specific support: provisions and charge-offs rose sharply, lower multifamily appraisals increased loss estimates, and management disclosed investigated or suspected borrower fraud. Full-year net charge-offs ultimately reached $124.1 million, with 70% arising from five relationships. [S3][S6]

• January 2026 relief rally. The stock closed at $34.95 on January 28 and $40.44 on January 29, a 15.7% increase following annual results and the announcement of a $100 million common repurchase authorization. The move reflected lower perceived tail risk and a capital-return signal, although subsequent execution was limited to 73,164 shares in Q1 and none in Q2. [S3][S4][S6]

• July-August 2026 rerating. The shares closed at $48.77 on the July 28 earnings date and $55.51 at the next available observation, July 30, before reaching the $55.72 closing high on August 3. The release supplied genuine evidence: $1.48 EPS, record tangible book, lower criticized and nonperforming balances and an 83% year-over-year decline in provision. The counterweight is that provision relief contributed much more to the earnings increase than revenue growth. [S1][S2][S6]

The map rejects two easy narratives. Merchants is not merely a regional bank still trading at panic valuation, because the shares are near their five-year high and above the 2024 common-offering price. Nor has it been a smooth compounder interrupted only by sentiment: repeated earnings-date gaps correspond to real mortgage, credit, funding and capital questions. The factor model reinforces the distinction. Its market exposure is 0.97 and small-size exposure 0.83, but the model explains only 27.4% of return variation. The sector coefficients are statistical co-movements, not legal classifications or causal business explanations. [S15]

Verdict. The latest rally is supported by better credit and earnings evidence, but the five-year range and current multiple show that historical cheapness arguments must be recomputed rather than inherited.

Business Overview

Merchants Bancorp is an Indiana bank holding company operating principally through Merchants Bank of Indiana and specialized mortgage, servicing, investment and asset-management subsidiaries. MBIN is ordinary Nasdaq-listed U.S. corporate common stock. It is not an ADR, partnership, MLP, mortgage REIT or K-1 issuer. Common holders nevertheless sit behind three public series of non-cumulative preferred stock, so consolidated net income and total equity cannot be used as if every dollar belonged to common shareholders. [S3][S5]

The company is easiest to understand as an integrated housing-finance chain. A multifamily or healthcare developer may begin with bridge or construction financing while a property is built, renovated or leased. After stabilization, Merchants can arrange permanent financing through FHA/Ginnie Mae, Fannie Mae, Freddie Mac, USDA or another investor, sell or securitize the fixed-rate loan, retain servicing and recycle its balance sheet. An affordable-housing project may also require low-income-housing-tax-credit equity, debt-fund capital or asset-management services. Separately, a residential or multifamily mortgage originator needs short-duration warehouse advances between closing a loan and selling it to an agency or investor. Merchants provides the advance, controls collateral and may capture custodial, escrow or operating deposits.

That chain produces several forms of customer value. Developers receive a lender able to coordinate interim debt, permanent debt, tax-credit capital and servicing. Mortgage companies receive dependable funding capacity, rapid collateral processing and settlement infrastructure. Agencies and investors receive an approved counterparty with specialized underwriting and servicing systems. Tax-credit investors receive access to structured affordable-housing projects and ongoing asset management. The economic test is not the breadth of the product list; it is whether integration produces faster execution, repeat mandates, more fee revenue per relationship, lower acquisition cost, stable deposits or superior loss experience. [S3]

The company reports three operating segments:

1. Multifamily Mortgage Banking originates and services multifamily rental and healthcare loans. Fixed-rate loans are commonly sold or securitized, often with servicing retained. Revenue includes gain on sale, contractual servicing fees, tax-credit syndication and asset-management fees, and changes in the value of mortgage-servicing rights and related derivatives.

2. Mortgage Warehousing provides revolving, short-duration credit to nonbank mortgage companies, generally secured by agency-eligible single-family and multifamily mortgages. The advance is repaid when the underlying mortgage is sold. Economics depend on funded volume, utilization, spread, collateral control, customer quality, fraud prevention and Merchants’ marginal funding cost.

3. Banking holds bridge, construction and adjustable-rate multifamily and healthcare loans, as well as commercial, residential, agricultural, SBA and community-bank credits. It bears the principal traditional credit risk when a project’s planned permanent takeout is delayed, reduced or unavailable. [S3]

In 2025, segment net income was $40.2 million for Multifamily Mortgage Banking, $96.9 million for Mortgage Warehousing and $122.0 million for Banking, offset by $40.3 million in Other. At year-end, segment assets were approximately $0.5 billion, $7.3 billion and $11.3 billion, respectively. Q2 2026 segment earnings were $10.3 million, $30.6 million and $47.3 million, offset by $10.0 million of Other losses. The small reported asset base of Mortgage Banking understates its economic footprint because loans move off balance sheet after sale while servicing, customer relationships and contingent obligations remain. [S1][S3]

Revenue is diverse by accounting line but not fully stable. Company Financials’ standardized series shows revenue of approximately $428 million in 2021, $440 million in 2022, $559 million in 2023, $667 million in 2024 and $677 million in 2025. First-half 2026 net interest income was $265.2 million and noninterest income $92.3 million, making spread income roughly three-quarters of pre-provision revenue. Contractual servicing is the most annuity-like fee stream, but the reported servicing line includes fair-value movements. Gain-on-sale and syndication income are transactional. Warehouse income depends on closing volume and utilization. Banking spreads depend on asset mix, deposit beta and borrowing costs. Provision can overwhelm pre-provision stability, as 2025 demonstrated. [S1][S3][S6]

Servicing is an important asset and a source of analytical confusion. At year-end 2025 Merchants serviced $21.3 billion for others, subserviced $3.7 billion and serviced another $0.7 billion in related portfolios. The balance sheet recognized $217.3 million of servicing rights at year-end and $236.9 million at June 2026. The platform also generated restricted servicing escrows exceeding $2 billion. Agency approvals, operating data, repeat-borrower relationships and servicing processes are not fully captured by ordinary tangible book. Conversely, the recognized servicing asset is exposed to prepayment, rate and valuation changes and requires compliance and operating expense; it should not be treated as a costless unrecognized moat. [S1][S3]

The balance-sheet strategy provides offsets rather than immunity to rates. Approximately 96% of held-for-investment loans repriced within three months at year-end 2025. Higher rates can raise yields on floating assets and support servicing values, but can reduce property values, borrower coverage, permanent-loan proceeds and transaction volume. Lower rates can improve refinancing feasibility and originations but compress floating-asset spreads, escrow income and servicing values. The speed of the move, the yield-curve shape and the timing of deposit repricing determine the net result. Q2’s NIM compression despite record assets is evidence that nominal growth does not guarantee better spread economics. [S2][S3]

Geographically, the company’s economic exposures are overwhelmingly domestic even though its warehouse and multifamily activities operate nationally. Seven Indiana branches are not a good measure of market reach because online deposits, national warehouse relationships, agency lending and loan sales extend beyond the branch footprint. The model is nationally scalable but remains dependent on U.S. housing policy, U.S. bank funding, agency programs and local property economics.

The main simplification is correlation. Segment accounting may suggest three businesses, but bridge lending, agency takeout, servicing, warehouse finance and related deposits frequently depend on the same mortgage-liquidity chain. A financing disruption can reduce new originations, lengthen bridge duration, increase credit exposure, pressure warehouse customers, slow servicing additions and weaken relationship deposits at the same time. Diversification is real relative to a mono-line bridge lender, but less powerful than segment labels imply.

The business can be readily understood as an integrated housing-finance chain whose principal drivers are origination volume, net interest spread, agency takeout, servicing economics, credit loss and regulatory capital. Net interest and contractual servicing provide a recurring base, but gain-on-sale, syndication, mortgage volume, fair-value marks and credit provisions make common earnings materially cyclical. Agency permissions, affordable-housing expertise, customer relationships, servicing data and escrow or custodial deposit channels are economically valuable assets only partly captured by book value. MBIN is ordinary U.S. corporate common stock, but common investors must deduct the substantial preferred layer when measuring earnings, book value and return. [S1][S3][S5]

Verdict. Merchants has a coherent, understandable and potentially high-return franchise, but it should be underwritten as one correlated housing-finance ecosystem rather than three independent businesses.

Industry Dynamics

Merchants participates in several connected U.S. markets: multifamily and healthcare construction finance, bridge lending, agency permanent lending, mortgage servicing, residential mortgage warehousing and affordable-housing capital. Profit pools move along a chain. Developers and owners need interim capital; banks hold or syndicate transitional exposure; approved lenders arrange agency permanent debt; Fannie Mae, Freddie Mac, FHA, Ginnie Mae and USDA provide liquidity and program rules; servicers earn fees; warehouse banks fund originators; and tax-credit investors supply equity to qualifying projects. A company able to serve several links can capture more economics but also becomes exposed to several failure points.

The newest market forecast is materially less bullish than an older estimate. MBA’s January 2026 forecast projected $399.2 billion of multifamily originations, up 20.8% from an estimated $330.6 billion in 2025. Its April forecast revised 2025 multifamily activity to $413.1 billion and projected $422.1 billion for 2026—only 2.2% growth—followed by $419.8 billion in 2027 and $422.0 billion in 2028. The dollar forecast rose while the growth rate collapsed because the comparison base was revised sharply upward. Using the January growth rate without its vintage would materially overstate the current industry tailwind. [S10][S25]

Actual Q2 data are positive but similarly need the right denominator. MBA reported that all commercial and multifamily originations increased 16% year over year and 12% sequentially. Multifamily-property originations specifically rose 8% year over year and 15% sequentially. Depository originations increased 61% year over year, whereas GSE activity declined 17% and life-insurer activity declined 27%. Thus transaction liquidity improved, but the mix favored depositories and CMBS rather than every channel. [S11]

Two 2025 market totals should not be treated as interchangeable. MBA’s April annual summation estimated $413 billion of multifamily lending within a broad commercial-real-estate universe. Its August multifamily report measured $381.8 billion of new mortgages on properties with at least five units, up 32%, across approximately 2,530 active lenders. The August report found that 45% of lenders completed five or fewer loans, evidence of a fragmented long tail, while the top five were JPMorgan, Wells Fargo, Walker & Dunlop, Berkadia and CBRE. The $31 billion difference reflects survey universe and estimation method, not necessarily an error. [S12]

Government policy is simultaneously a barrier, demand source and concentration risk. FHFA set 2026 multifamily purchase caps of $88 billion each for Fannie and Freddie, or $176 billion combined, up 20.5% from $146 billion in 2025. At least half of each enterprise’s business must be mission-driven affordable housing. Qualifying workforce-housing loans remain outside the cap, while most other mission-driven loans remain within it. The higher capacity and affordable mandate benefit specialists such as Merchants, but pricing, eligibility, appraisal, reserve, servicing and compliance rules can alter economics independently of customer demand. [S13]

The maturity wall is a transaction opportunity, not automatically a profitable refinancing wave. MBA estimated that $875 billion, or 17% of $5.0 trillion of commercial-mortgage balances, was scheduled to mature during 2026. That figure covers all commercial mortgages, not multifamily alone; 13% of multifamily balances were scheduled to mature. Only $39 billion, or 4%, of agency, FHA and Ginnie multifamily and healthcare balances were scheduled to mature, compared with $396 billion of depository-held balances and $200 billion of CMBS, CLO and other securitized balances. Some maturities generate new fees. Others require extensions, equity infusions, modifications or losses because current property cash flow cannot support the debt. [S22]

The industry’s capital cycle is mixed. Contraction in mortgage activity caused some warehouse providers and mortgage originators to exit, creating opportunities for surviving platforms. Merchants’ warehouse funded volume rose from $33.0 billion in 2023 to $45.6 billion in 2024 and $66.3 billion in 2025. The 46% increase in 2025 exceeded the approximately 22% single-family market growth cited in its filing, consistent with share gain. Management attributed part of the gain to customer acquisition and competitor exits. The volume is a fact; the cause and durability remain management claims to test. [S3][S8]

Supply can return. Warehouse lending requires funding, collateral control, fraud systems and capital, but attractive spreads invite banks and private capital. Agency products are standardized, and sophisticated borrowers solicit multiple lenders. Producer compensation can rise with volume. In an early recovery, competitor exits may lift spreads and share; later, stronger expected volumes can attract capacity and compress economics. Investors should therefore monitor warehouse segment income and return on allocated capital, not funded volume alone.

Barriers to entry are meaningful but not absolute. Scaled agency lending requires approval, specialized underwriting, regulatory compliance and servicing capacity. Warehouse lending requires a bank or equivalent funding source, collateral-control technology, settlement processes and the ability to survive rapid utilization changes. Tax-credit syndication requires institutional investor relationships, legal structuring and asset-management capability. These assets take years to assemble. They do not create monopoly power because large banks, national brokerages and dedicated mortgage firms possess comparable permissions and deeper resources.

The market is overwhelmingly domestic. Foreign low-cost labor can reduce back-office expense, but it cannot cheaply reproduce a U.S. bank charter, FDIC-supervised funding, agency approval, local appraisal review, title and collateral control, regulatory capital or institutional borrower relationships. The credible disruption threats are domestic technology, alternative capital and policy change—not offshore production.

Credit conditions contradict a simple volume-recovery thesis. Fannie Mae’s multifamily guaranty book had a 0.60% serious-delinquency rate at June 2026, down from 0.78% at March, but the reduction was driven principally by modification of a previously forborne portfolio and foreclosure activity. Multifamily REO increased from 181 properties carrying $1.0 billion at year-end to 218 properties carrying $1.5 billion. Fannie recorded a $259 million Q2 multifamily provision and $157 million of net charge-offs and warned that ongoing market challenges could create additional delinquencies. The 2022 vintage had a 1.26% serious-delinquency rate versus 0.60% for the total book. [S14]

This industry evidence matters for Merchants in two ways. First, improving activity can lift origination, warehouse and servicing additions. Second, wider market liquidity does not cure every older transitional loan. A property originated against peak values may remain impaired even if a newly underwritten loan is attractive. New production and legacy credit must be analyzed as different vintages.

Merchants addresses a large, predominantly domestic market, but the newest April forecast implies only modest 2026 multifamily growth rather than a 21% surge. Competition eased during the mortgage contraction, although renewed capacity and producer demand can restore spread pressure as activity improves. Scaled agency, servicing, warehouse and tax-credit businesses can earn attractive returns, but agency approval, regulatory capital, compliance, data and specialist labor create barriers rather than monopoly power. Foreign low-cost production cannot readily undermine the franchise because U.S. regulation, funding, agency permissions and local collateral processes dominate labor cost. [S3][S9][S10][S13]

Verdict. The industry setup is better for transactions than in the trough, but forecast growth is modest and legacy multifamily credit remains stressed. Merchants must create value through share, mix and underwriting rather than relying on an indiscriminate market rebound.

Competitive Position

Merchants’ strongest competitive feature is integration. A developer can obtain bridge or construction credit, permanent agency execution, tax-credit capital, servicing and related treasury services within one group. A mortgage originator can obtain warehouse funding and maintain operating or custodial deposits at the same institution. Integration reduces coordination cost and gives Merchants more observations across the life of a transaction. The potential advantage is faster execution, repeat business and more revenue per relationship.

The distinction between capability and moat is essential. Owning adjacent products does not automatically produce superior economics. Integration becomes a competitive advantage only if it creates measurable outcomes: repeat mandates, lower acquisition cost, lower deposit cost, faster asset turnover, better cross-selling or lower credit loss. The 2025 losses show that extensive customer knowledge and specialist systems did not prevent lower appraisals, suspected fraud, subordinated structures and borrower-specific failures from producing $124.1 million of net charge-offs. [S3]

External evidence supports a genuine affordable-housing niche. Fannie financed approximately $73.7 billion of multifamily activity in 2025, including more than $8.3 billion of affordable housing. Merchants Capital ranked fifth in affordable-housing production. It did not appear in Fannie’s overall top-ten DUS producer list, whose leaders included Walker & Dunlop, Wells Fargo, CBRE, Berkadia and Newmark. Nor did it appear in Fannie’s top-five structured, small-loan or manufactured-housing categories. The defensible conclusion is specialist relevance—not broad agency dominance. [S9]

Warehouse growth also validates operating capacity. Funded volume doubled from $33.0 billion in 2023 to $66.3 billion in 2025, and the 2025 increase materially exceeded market growth cited by the company. That suggests customers trusted Merchants with more settlement volume and that the platform could process it. However, volume can be won with price, broad limits or favorable covenants. Persistent advantage requires segment profit and capital return to grow along with volume. Q2 2026 Mortgage Warehousing segment net income of $30.6 million was strong, but continued growth also contributed to lower-spread asset mix and higher borrowing needs. [S1][S3]

Servicing creates longer-duration relationships. A borrower whose loan is sold to an agency may retain Merchants as servicer for years. The servicing platform produces contractual fees, operating information and escrow deposits; at year-end 2025 restricted servicing escrows exceeded $2 billion. These are more durable than one-time gain-on-sale revenue. Yet servicing contracts are regulated, operationally demanding and exposed to prepayment. Q2 reported servicing income nearly doubled year over year, but roughly $6 million of the line was a favorable valuation adjustment. The recurring cash component must be separated from the mark. [S2][S3]

Brand matters as institutional reputation. Developers, mortgage companies, agencies and tax-credit investors care about certainty of closing, program expertise, balance-sheet capacity, servicing quality and regulatory standing. A credible institutional brand can lower customer-acquisition cost and win repeat mandates. It does not create consumer-style price inelasticity. Evidence of brand value is agency ranking, retained servicing and repeat volume—not public name recognition.

Switching costs are moderate. A warehouse customer changing lenders must alter collateral-control procedures, reporting, covenants, settlement instructions and operating deposits. Multiple warehouse facilities are nevertheless common, limiting lock-in. Servicing rights can remain in place for a loan’s life, but borrowers can shop the next financing. Repeat developers benefit from a lender familiar with their projects and capital stack, yet standardized agency products allow competitive bidding. Operational friction supports retention; it does not guarantee pricing power.

The competitive set changes by activity. Walker & Dunlop, Berkadia, CBRE, JLL, Greystone, Newmark, PGIM, Wells Fargo and KeyBank compete in agency multifamily. Customers Bancorp, Western Alliance, The Bancorp, Axos, Texas Capital and large national banks provide more relevant comparisons for specialized banking, deposits or warehouse finance. Arbor Realty is useful for bridge-credit and agency questions but is a mortgage REIT, not a deposit-funded bank. No single peer captures Merchants’ preferred-capital layer, servicing marks, warehouse scale and multifamily bridge exposure.

Relative positioning has trade-offs. Compared with Walker & Dunlop, Merchants takes more balance-sheet, deposit and regulatory-capital risk but controls its own funding. Compared with Arbor, it has a bank charter and broader funding but less direct comparability to a distributable-earnings model. Compared with Customers Bancorp or Western Alliance, Merchants adds agency multifamily, tax-credit and servicing capabilities but carries more visible multifamily problem assets. Compared with The Bancorp or Axos, its demonstrated common return is lower and more credit-sensitive.

The supply-side capital-cycle test is particularly important. Competitor exits made warehouse growth easier in 2024-2025. If returns remain attractive, capital will re-enter, experienced producers will demand more compensation and customers will bargain more aggressively. A true advantage would let Merchants preserve segment return while peers add capacity. Falling spreads, rising compensation or higher funding intensity would show that growth came from favorable industry supply rather than a durable moat.

Competitive deterioration would appear through operating outcomes before it appeared in labels. Warning signals include warehouse volume rising while segment income falls; higher agency volume without gain-on-sale or servicing growth; servicing assets increasing without contractual-fee growth; loss of affordable-housing rankings; core deposits losing share; producer compensation outgrowing revenue; or repeated credit failures in portfolios described as specialist. Validation would require above-peer common returns, lower credit losses on remediated vintages, stable relationship deposits and repeat agency production.

Brand matters as institutional execution credibility with agencies, developers, mortgage companies and tax-credit investors, but it does not create consumer-style pricing power. Competition is based on price, certainty and speed of execution, agency expertise, funding capacity, servicing and the ability to provide several layers of a housing capital stack. Servicing contracts, collateral-control systems, operating deposits and repeat-underwriting knowledge create moderate switching costs, while standardized agency products and multiple warehouse lines prevent lock-in. [S3][S9]

Verdict. Merchants has a defendable specialist franchise and real cross-selling potential, but the evidence supports niche differentiation rather than a wide moat. The unresolved test is whether integration improves risk-adjusted return, not merely volume.

Growth History and Forward Opportunities

From 2021 through 2025, Company Financials’ standardized revenue measure increased from approximately $427.7 million to $677.2 million, while assets expanded from $11.3 billion to $19.4 billion. Assets reached $21.2 billion by June 2026. Common earnings were much less linear: approximately $206.2 million in 2021, $193.7 million in 2022, $244.6 million in 2023, $283.7 million in 2024 and $173.6 million in 2025 after preferred dividends and redemption effects. Diluted EPS rose from $4.76 in 2021 to $6.30 in 2024 before falling to $3.78 in 2025. Scale grew; earnings durability did not improve proportionately. [S1][S3][S6][S24]

The first opportunity is transaction-market normalization, but the size of the tailwind must be disciplined. MBA’s April forecast calls for $422.1 billion of 2026 multifamily originations, up 2.2%, not the 20.8% growth in its older January forecast. Q2 multifamily originations were nevertheless up 8% year over year and 15% sequentially. A healthier market can lift gain-on-sale income, warehouse utilization, tax-credit fees and future servicing additions. It cannot convert every overleveraged maturity into a profitable new loan. [S10][S11][S25]

The second opportunity is warehouse share. Funded volume doubled from 2023 to 2025 as customers were added and competitors exited. Short-duration agency-eligible collateral can turn quickly, and customers may bring operating or custodial deposits. The economic constraint is the marginal balance sheet. Q2 average borrowings rose 28% sequentially and NIM fell 11 basis points to 2.81% as growth favored warehouse and held-for-sale loans. If incremental assets earn less after funding, operating expense, expected loss and capital, volume can dilute value while increasing reported size. [S2][S3]

Affordable housing offers the clearest structural runway. FHFA’s $176 billion combined 2026 enterprise caps and 50% mission-driven requirement provide capacity. Merchants can earn debt-placement, tax-credit, servicing and asset-management economics on one project. Its Fannie ranking establishes relevance. The constraints are construction cost, local rent economics, developer execution, tax-credit pricing, agency rules and the availability of permanent takeout. [S9][S13]

Servicing converts transactional production into contractual revenue and escrow funding. The year-end serviced and subserviced footprint exceeded $25 billion. Contractual servicing fees were $33.1 million in 2025, while the reported servicing line was affected by asset amortization and valuation. Q2 2026 servicing income increased approximately $5.9 million year over year, but a $6.0 million positive valuation adjustment explains much of the apparent growth. The opportunity is growth in contractual fees and low-cost relationship balances, not capitalization of favorable marks as an annuity. [S2][S3]

Credit-risk transfer can free capital or reduce selected tail exposure. In June 2026 Merchants executed a credit-default swap on a $169.9 million multifamily pool, adding to other protected pools. Total loans within disclosed protection structures were approximately $2.2 billion, but contractual protection covered only 13%-15% of each pool’s unpaid balance and only about 6% of criticized loans overlapped protected pools. Protection can improve the referenced pool’s capital or loss distribution without validating the underwriting of unprotected legacy loans. [S1][S2]

A further growth avenue is balance-sheet recycling. Selling fixed-rate production, using agency securitization and transferring selected credit risk can support more customer activity per dollar of common capital. The risk is that management uses freed capacity to increase gross exposure faster than retained earnings. Total assets grew 9% from year-end through Q2 while total capital declined. Growth should therefore be judged by incremental common earnings and tangible-book growth per share, not gross loans alone.

Management also discontinued some subordinated products associated with prior losses and strengthened monitoring and concentration controls. If post-remediation vintages have lower leverage, stronger recourse, independent appraisal review and better fraud controls, new production may carry a better loss distribution than 2021-2023 cohorts. Public disclosures do not yet provide sufficient vintage-level evidence to prove that improvement. [S3]

The outlook for agency origination, warehouse finance, servicing and affordable-housing syndication is improving with transaction activity, but growth creates value only if spreads, credit quality and common capital remain disciplined. The best near-term opportunity is better utilization of the existing platform, not another step-change in leveraged balance-sheet growth. [S1][S10][S13]

Verdict. Merchants has credible growth channels, but the latest industry forecast is modest and internal capital is constrained. Incremental risk-adjusted return, rather than market growth or funded volume, should control the outlook.

Financial Quality

The historical record combines strong reported profitability with abrupt credit volatility. Banks do not report one universally comparable GAAP revenue line, so the table uses Company Financials’ standardized revenue series while reconciling net income and common EPS to filings.

The common-income correction is material. In 2025, consolidated net income of $218.8 million was reduced by $41.1 million of preferred dividends and a $4.2 million preferred-redemption effect, leaving $173.6 million for common. In 2024, the comparable figures were $320.4 million, $34.9 million and $1.8 million, leaving $283.7 million. Subtracting preferred dividends alone overstates common earnings. [S1][S3][S6]

The 2025 downturn was severe. Diluted EPS fell 40%, ROA declined to approximately 1.16% from 1.79%, return on tangible common equity fell to 10.49% from 20.16% and the efficiency ratio worsened to 44.01% from 33.37%. Net charge-offs increased to $124.1 million from $10.6 million. Seventy percent of the 2025 charge-offs arose from five relationships, and multifamily net charge-offs were 2.29% of average multifamily loans. Management linked losses to lower property values, updated appraisals and investigated or suspected mortgage fraud; some subordinated products involved are no longer offered. [S3]

Concentration cuts two ways. It supports the bull case that a limited set of loans caused an unusually bad year. It also shows that a handful of underwriting, control or borrower failures can overwhelm diversified pre-provision earnings. A bank cannot treat relationship concentration as reassuring unless remaining collateral, sponsor exposure and underwriting controls are independently demonstrated.

Q2 2026 was profitable but provision-heavy. Net income was $78.3 million, up $40.3 million year over year, and diluted EPS was $1.48. Net interest income increased $7.8 million, or 6%, to $136.5 million. Provision fell $43.8 million, or 83%, to $9.2 million. Noninterest income declined $4.8 million, or 10%, and gain-on-sale income fell $10.2 million, or 44%. Expenses declined 5%. The provision change exceeded the pretax increase in recurring revenue, so most of the earnings rebound came from lower credit cost. [S2]

First-half results also contain fair-value assistance. Q1 servicing-rights valuation increased income by $8.9 million and derivative marks added $2.7 million; Q2 corresponding gains were $6.0 million and $1.9 million. These are legitimate GAAP changes in asset and hedge values, but they are not equivalent to cash servicing fees. Conversely, excluding only favorable marks would be asymmetric if one did not also exclude unfavorable marks in other periods. The correct approach is to show GAAP and separately evaluate contractual fees and multi-period economic return. [S2][S4]

Trailing common earnings through Q2 were approximately $228.7 million. The reconciliation is $173.6 million of reported 2025 common income, less $70.3 million from first-half 2025, plus $125.5 million from first-half 2026. Four-quarter diluted EPS sums to approximately $4.98. That is higher than 2025 but below the 2024 peak. It also shows why $1.48 cannot simply be multiplied by four without assuming that provision, fee mix and fair-value marks remain favorable. [S1][S3]

Conventional industrial ROIC is not an appropriate primary metric because deposits and borrowings are operating inputs for a bank, not financing to exclude from invested capital. Company Financials correspondingly provides no standard ROIC value for the bank. Sector-appropriate measures are ROA, common ROE or ROTCE after preferred claims, tangible-book growth plus common distributions and incremental earnings relative to regulatory capital. Q2 ROTCE of 14.95% is attractive; 2025 ROTCE of 10.49% was modest for the observed risk; 2024’s 20.16% subsequently proved to contain more latent credit risk than the contemporaneous provision suggested. [S2][S3][S6]

Credit remains the central quality test. Criticized loans fell to $444.7 million, or 3.6% of gross loans, from 4.4% at Q1 and 4.6% at year-end. NPLs fell to $205.6 million, or 1.67%, from $247.5 million, or 2.16%, sequentially. The company stated that the decline came primarily from loans paid in full, a stronger cure mechanism than charge-off. Nevertheless, NPLs remained 3.9% above the $197.8 million year-end balance. Q2 gross charge-offs were $16.5 million, with almost 95% arising from two multifamily relationships, after $23.0 million of Q1 charge-offs. [S1][S2][S3]

The $75.8 million allowance equaled 36.9% of NPLs and approximately 0.61% of gross loans. Specific reserves on substandard loans were only $3.9 million. Low NPL coverage is not automatically inadequate: collateral-dependent loans can have recoverable value outside the allowance, and some loans have government or credit-transfer protection. But updated appraisals, liquidation timing, taxes, senior claims and fraud can reduce net recovery. The auditor identified the allowance, particularly qualitative adjustments, as the sole critical audit matter while issuing unqualified opinions on the financial statements and internal controls. That is evidence of estimation risk, not evidence of misstatement. [S1][S3]

OREO is a disconfirming indicator. It increased from $60.1 million at year-end to $72.4 million at June. A loan moving from NPL to OREO can reduce one credit statistic without completing economic recovery. Management’s disclosure that Q2 NPL declines were primarily repayments is encouraging, but the exact waterfall among repayment, refinancing, charge-off, foreclosure, sale and OREO transfer is still not fully disclosed. Investors need both balance migration and realized recoveries.

Funding improved structurally but weakened at the margin. Core deposits rose from 58% of deposits in 2023 to 79% in 2024 and 87% at year-end 2025. At Q2 they were $13.0 billion, or 91% of deposits—up year over year but down from 93% at Q1. Brokered deposits rose $411.3 million, or 46%, sequentially to $1.3 billion and had a weighted-average remaining maturity of 51 days. The coexistence of a high core ratio and rising short-duration brokered funding is not contradictory: relationship deposits fund most of the balance sheet, while marginal asset growth still required wholesale funding. [S2][S3]

Disclosed liquidity of $13.0 billion, or 61% of assets, is not $13 billion of cash. It includes $5.5 billion of unused borrowing capacity, warehouse lines, loans held for sale and assets available for securitization. Cash and cash equivalents were $314.7 million. The broader measure is relevant because high-quality collateral can be sold, financed or securitized; it remains contingent on market access, collateral eligibility, haircuts and operational execution. [S1][S2]

Interest-rate risk is managed but not eliminated. Approximately 96% of held loans repriced within three months at year-end 2025. Q2 NIM nevertheless declined 11 basis points sequentially to 2.81% because growth favored lower-margin warehouse and held-for-sale loans and average borrowings rose 28%. Average interest-bearing deposit cost remained substantial and borrowings cost more than 4%. Rapid changes in the curve can alter asset yield, deposit beta, servicing value and borrower credit in different directions. [S2][S3]

Capital is adequate rather than abundant. June CET1 was 9.3%, total capital 12.5% and tangible common equity to tangible assets 8.64%. Total capital was down from 13.4% a year earlier and from 12.8% at Q1. Total assets reached $21.2 billion while common capital had to absorb credit costs and support growth. The MOU’s termination removes a regulatory overhang, but it does not create capital or cure assets. [S1][S2][S4]

Cash-flow statements require bank-specific interpretation. Company Financials reports operating cash flow of approximately negative $49 million in 2021, positive $976 million in 2022, negative $356 million in 2023, negative $835 million in 2024 and negative $341 million in 2025. These swings are dominated by loans originated and sold, not by ordinary industrial working capital. A conventional CFO-minus-capex calculation would incorrectly classify mortgage inventory funding as recurring operating consumption. Cash flow remains important for liquidity, but common earnings, tangible-book growth, distributions and regulatory-capital generation are more useful measures of distributable economics. [S3][S6]

Off-balance-sheet and contingent obligations include unfunded loan commitments, servicing duties, representations and warranties on sold loans, derivatives, tax-credit commitments and interests in variable entities. Loan sale removes an asset without eliminating every repurchase, servicing or compliance obligation. The relevant loss is not the gross notional alone; it depends on contractual triggers, project performance, collateral and whether Merchants is principal, guarantor, servicer or manager. [S1][S3]

Accounting is neither demonstrably aggressive nor uniformly conservative. Loan-sale accounting appropriately recognizes sale economics and a servicing asset. CECL recognizes expected losses before charge-off but depends on forecasts, appraisals and qualitative factors. Credit protection is accounted for separately rather than automatically netted against gross exposure, which avoids disguising the underlying loan balance. Fair-value servicing and derivative marks add volatility. No material error correction or adverse internal-control opinion explains the earnings swing. [S1][S3]

Earnings were at a credit-cycle low in 2025 and recovered in first-half 2026, but Q2’s provision and favorable marks should not yet be treated as through-cycle levels. Sector-appropriate profitability is attractive when credit normalizes, although the gap between 2024 and 2025 returns shows why risk-adjusted ROTCE matters more than peak reported ROE. Unfunded commitments, servicing and repurchase obligations, derivatives and affordable-housing fund structures remain economically relevant even when associated assets are not consolidated. Mortgage Banking can recycle sold loans, but Warehouse and Banking require substantial funding and regulatory capital. GAAP operating cash flow diverges from net income because loans held for sale pass through operating cash flow, making industrial free-cash-flow measures misleading. [S1][S2][S3][S6]

Verdict. Financial quality is recovering, but concentrated charge-offs, year-end-to-date NPL growth, rising OREO, short-duration marginal funding and declining capital ratios prevent classification as a high-quality bank.

Capital Allocation

Capital allocation has principally funded organic growth and regulatory capacity. In May 2024 Merchants sold 2.4 million common shares at $43 and received approximately $97.7 million net. Common shares outstanding rose from roughly 43.3 million before the transaction to about 45.9 million in 2025-2026. The recent record is therefore net issuance, not sustained common-share shrinkage. [S3][S17]

Preferred capital is material. At June 2026 the carrying values of Series C, D and E totaled approximately $551.3 million, compared with about $1.84 billion of common book equity. Aggregate liquidation preference was approximately $568.7 million. Annual preferred dividends are roughly $41 million. These securities support capital but rank ahead of common and reduce common earnings. Any comparison using consolidated net income or total equity without deducting them overstates common economics. [S1][S3]

The company redeemed $52 million of Series A in April 2024 and $125 million of Series B in January 2025. It issued $230 million of 7.625% Series E in November 2024, receiving approximately $222.7 million net. Series C became callable in April 2026. Redeeming its approximately $196 million liquidation value would save roughly $11.8 million of annual dividends, a cash return near 6% before regulatory and refinancing effects. That creates a real allocation hurdle: common repurchase, preferred redemption and loan growth compete for the same capital.

The common dividend is $0.11 quarterly, or $0.44 annually. At the controlled price its yield is approximately 0.8%, and the annual cash cost is about $20 million. Combined common and preferred distributions require roughly $61 million annually, well covered by trailing consolidated and common earnings. Coverage can tighten quickly in a credit shock because preferred claims are economically senior and regulators may restrict distributions even when the preferreds are legally non-cumulative. [S1][S3]

The January 2026 board authorization permits up to $100 million of common repurchases through December 2027. Merchants bought 73,164 shares for $3.0 million during Q1, an average near $41.03, and none during Q2. That was rational price discipline relative to tangible book. It also means the authorization has not produced material net shrinkage. The pause may reflect valuation, growth, capital planning or several factors; it should not be attributed to the former MOU without evidence. [S1][S4]

The acquisition record requires one correction to a simplistic no-M&A description. Merchants did not complete a transformative acquisition during the trailing five years, but it did unwind an older transaction. In January 2024 it sold the four Farmers-Merchants Bank of Illinois branches and collapsed the remaining charter into Merchants Bank. Buyers assumed approximately $230 million of deposits and $47.8 million of loans. Merchants recognized a $0.7 million gain, including a $10.1 million deposit premium and the extinguishment of $7.8 million of goodwill and $0.5 million of intangibles. The disposal sharpened focus on national mortgage banking, but public disclosure is insufficient to calculate a full return on the 2018 acquisition after intervening earnings. [S7]

This also explains why an automated cash-flow classification can look like acquisition spending: transferring deposit liabilities and assets in a branch sale can produce a large net cash settlement. The economic event was a divestiture, not a new $170 million acquisition. Primary transaction notes should control the interpretation.

Insider evidence is mixed. Merchants Capital CEO Michael Dury bought 16,000 common shares in the open market on July 30, 2025—11,765 at a weighted average $31.45 and 4,235 at $31.86. CFO Sean Sievers bought 2,500 shares at $29.75 on August 1. These code-P purchases near the trough are meaningful because they committed personal capital. Chairman and CEO Michael Petrie subsequently sold 15,000 shares at approximately $44.05 in February 2026 and 39,030 direct or spouse-trust shares near $46.75-$46.86 in May. Grants, vesting, tax withholding, gifts and director retainers should not be conflated with open-market purchases or sales. [S18][S19][S20]

Petrie and associated spouse-controlled trusts beneficially owned approximately 27.4% of common shares, and directors, nominees and executives collectively controlled approximately 44.9%. Seven of the twelve nominated directors were classified as independent, and the board has a lead independent director, but Petrie combines chairman and CEO roles. High ownership aligns wealth with long-term outcomes while concentrating voting control and limiting minority influence. [S5]

Related-party disclosures deserve a valuation discount, not an allegation of wrongdoing. They included compensation to family employees, legal fees involving a director’s former firm and a $2.2 million land transaction with an entity related to the CEO. The transactions were disclosed and described as reviewed or conducted on market terms. Minority investors cannot independently reproduce every economic comparison.

Compensation design is more credit-sensitive than a surface reading suggests, but still imperfect. The 2025 annual targets were proxy-defined revenue of $714.0 million, EPS of $6.02 and ROAE of 14.23%. Actual values were $563.7 million, $3.78 and 9.88%, producing an average score of 70%. Proxy-defined revenue equals pre-provision operating revenue less provision, so credit costs did affect that metric indirectly. Nevertheless, there was no explicit multi-year charge-off, criticized-loan, risk-adjusted-return or tangible-book measure. [S5]

Despite the below-threshold result, the committee used discretion to award Petrie, Dunlap and Sievers 75% of target cash and equity. Dury received no equity because his unit contributed to credit problems, while warehouse leader Aaron Schroeter received a discretionary $900,000 cash bonus. Positive features include three-year RSU vesting, clawbacks and a well-capitalized gate. Weaknesses include discretion after missed thresholds, no minimum stock-ownership requirement and no general ban on hedging or pledging without approval. [S5]

Bank free cash flow is best measured through common earnings, tangible-book growth, distributions and regulatory-capital generation; management has primarily retained capital for growth while paying modest common and substantial preferred dividends. No transformative recent acquisition drives returns, although the 2024 Illinois branch divestiture completed the unwind of an older acquisition. Only modest Q1 repurchases occurred, so no sustained net-share reduction exists. Common issuance has been material because the 2024 offering was much larger than ordinary employee awards. Executive incentives include provision-sensitive revenue, EPS and ROAE but lack explicit long-horizon credit and risk-adjusted-return metrics. Large founder ownership supports alignment, while control, related-party dealings, compensation discretion and founder sales support a minority-shareholder discount. [S3][S5][S7][S17]

Verdict. Allocation has preserved growth capacity and avoided transformative M&A, but common dilution, expensive preferred capital, compensation discretion and concentrated control offset the evidence of price discipline in the limited Q1 buyback.

Changes and Headwinds — Last Two Years

The past two years moved Merchants from peak profitability through a concentrated multifamily credit shock and into an early recovery. In 2024, common earnings and ROTCE reached records, the company sold 2.4 million common shares and issued Series E preferred capital, while investors twice reacted sharply to quarterly results. In January 2024 it completed the Illinois branch sale, reducing conventional community-bank exposure and increasing the strategic concentration on national mortgage finance. [S3][S7][S17]

During 2025, updated appraisals, lower property values, borrower performance and investigated or suspected fraud converted latent risk into $124.1 million of charge-offs. Seventy percent came from five relationships. Management discontinued certain subordinated products and strengthened appraisal, monitoring, concentration and capital practices. Those are internal responses. Interest rates, property NOI, permanent-loan proceeds and transaction liquidity remain external variables that management cannot control. [S3]

The June 2025 confidential MOU with the FDIC and Indiana Department of Financial Institutions covered capital, concentration management and risk mitigation. Because its complete terms were not public, investors cannot independently score every requirement. The company disclosed that regulators terminated it during Q1 2026 after progress addressing its provisions. Termination is strong evidence of regulatory remediation; it is not a regulator’s certification that every legacy borrower, appraisal or recovery is sound. [S3][S4]

The company also expanded credit-risk-transfer activity. Selected multifamily pools are protected through credit-default swaps or related structures, potentially reducing risk-weighted assets and referenced-pool loss. Public terms show limited protection percentages and little overlap with criticized assets. The structures improve capital flexibility before they prove broad legacy-credit repair. [S1][S2]

The 2026 environment is more supportive for transactions than the 2023-2024 trough, but not uniformly benign. Q2 multifamily originations rose year over year and sequentially, and enterprise caps increased. MBA’s newest forecast nevertheless implies only 2.2% full-year multifamily growth. Fannie’s rising REO, $259 million multifamily provision and warning of additional delinquencies contradict a fully healed property cycle. [S10][S11][S13][S14]

Company credit indicators also give a mixed signal. Criticized and NPL balances fell sequentially, with NPL reductions primarily from repayments. Relative to year-end, NPLs were still higher and OREO rose. Provision normalized but charge-offs remained concentrated. The correct description is early repair, not resolution. [S1][S2]

Funding changed materially. The core-deposit ratio improved from 58% in 2023 to 91% in Q2 2026, supported by operating, custodial and escrow balances. At the margin, however, record asset growth required more borrowings and a 46% sequential increase in brokered deposits. Assets reached a fifth consecutive quarterly record while total capital declined. The balance sheet is more core-funded than in 2023 but also larger, more capital-consuming and reliant on wholesale funding for incremental growth. [S2][S3]

No material accounting-policy change explains the earnings swing. CECL, loan-sale accounting, servicing-rights fair value and derivatives continued under established policies. The 2025 auditor reported effective internal control and no material error correction. The change was economic and estimation-driven: losses, appraisals, provision and fair-value movements changed, not the fundamental accounting framework. [S3]

Leadership remained founder-centered. Petrie continued as chairman and CEO, Dunlap as president and bank CEO, and the operating response occurred without a strategic leadership transition. The material management evidence is behavior: capital issuance, product exits, regulatory remediation, selected risk transfer, restrained buybacks after the rally, discretionary pay and insider purchases and sales.

Results reflect both external rates and property values and internal underwriting, product design, funding, credit remediation and expense decisions; neither explanation alone is sufficient. The last two years moved Merchants from peak profitability through a concentrated multifamily credit shock and into an early recovery under tighter capital oversight. No material accounting-policy reset or correction explains the volatility. The important changes were national mortgage conditions, the Illinois branch divestiture, regulatory remediation, capital issuance, balance-sheet expansion and product tightening rather than a facility build or CEO transition. [S3][S4][S7]

Verdict. The environment and internal controls have improved, but the short interval since the 2025 losses and continued year-end-to-date NPL and OREO pressure leave remediation durability unproven.

Risk Analysis

The central risk is concentrated multifamily credit. A small number of relationships generated most 2025 and first-half 2026 charge-offs. If collateral values, borrower cash flow or recovery timing deteriorate, the $75.8 million allowance may prove insufficient relative to $205.6 million of NPLs and $72.4 million of OREO. Affordable-housing or agency intent does not eliminate construction, bridge, lease-up, subordinated or fraud risk before permanent takeout. [S1][S2][S3]

The second risk is business-model correlation. High rates can impair property values and refinancing proceeds while slowing originations. A frozen securitization or agency-takeout market can reduce gain-on-sale, extend bridge loans, pressure warehouse clients, slow servicing additions and weaken related deposits simultaneously. Segment diversity reduces single-product dependence but does not eliminate common exposure to mortgage liquidity.

The third risk is funding and capital transmission. Core deposits fund most liabilities, but marginal growth required brokered deposits with a 51-day remaining term and higher average borrowings. CET1 of 9.3% and total capital of 12.5% provide buffers but less flexibility than a year earlier. Renewed losses could force slower growth, preferred issuance, common dilution, dividend restraint or asset sales. [S2]

The fourth risk is fraud, control and appraisal failure. Management cited investigated or suspected fraud in some 2025 relationships. Fraud can bypass historical credit models, and appraisal errors can concentrate losses when recourse is weak or senior claims exist. MOU termination reduces the probability of unresolved supervisory concern but cannot prevent a new borrower-level failure. [S3][S4]

The fifth risk is rate convexity. Higher long rates can support floating yields and servicing values while damaging refinancing capacity. Lower rates can help borrowers and originations but reduce MSR value, escrow income and asset yields. Funding may reprice at a different speed from loans, and derivative hedges may not perfectly offset servicing or basis risk. Q2 NIM compression shows that record volume can coexist with weaker spread. [S2]

The sixth risk is agency and policy dependence. Fannie, Freddie, FHA, Ginnie, USDA, tax-credit law and banking regulators influence eligibility, caps, pricing, capital and compliance. Higher caps are currently supportive. A reduction in approval status, mission rules or program economics would impair both revenue and institutional reputation. [S3][S13]

The seventh risk is governance. Founder-family ownership creates economic alignment but concentrates voting power. Related-party transactions, family employment, compensation discretion and the absence of minimum ownership rules create agency questions. Founder sales after the recovery do not predict performance, but they reduce the informational value of insider alignment at the current valuation. [S5][S18]

A catastrophic loss would most plausibly require several failures together: large collateral write-downs or fraud, correlated borrower defaults, reduced market access for warehouse and held-for-sale assets, deposit or borrowing pressure and an inability to rebuild capital without severe dilution. That path is plausible but not the base case because the company is currently profitable, remains above regulatory minimums, has substantial collateralized capacity and has reduced criticized assets. [S1][S2][S4]

A literal total loss for common equity is remote but possible. It would require losses and liquidity pressure sufficient for receivership or a recapitalization that eliminates existing common value. Preferred claims and holding-company obligations would rank ahead of common recovery. The more realistic severe outcome is a fall toward or below tangible book accompanied by capital issuance, suspended repurchases and materially lower common earnings.

The stock could decline on renewed multifamily losses, weaker gain-on-sale revenue, NIM compression, higher funding costs, capital dilution, policy change or governance-related multiple contraction. Catastrophic loss would most plausibly arise from correlated credit, fraud, funding and capital failures rather than one weak origination quarter. Total common-equity loss remains remote but can occur through receivership or a deeply dilutive recapitalization after losses overwhelm capital and liquidity. [S1][S2][S3][S5]

Verdict. Expected downside is manageable under a gradual credit cure, but the transmission from a handful of relationships into earnings and capital makes tail risk materially greater than a benign agency-housing description implies.

Valuation Discussion

At $53.14, 45.94 million common shares imply a market capitalization of approximately $2.44 billion. June total equity was $2.394 billion; subtracting $551.3 million of preferred carrying value leaves approximately $1.843 billion of common book equity, or $40.11 per share. Company-defined tangible common equity was $1.834 billion and tangible book value $39.93 per share. Corrected trailing diluted EPS was approximately $4.98. The resulting multiples are about 10.67 times trailing earnings, 1.32 times common book and 1.33 times tangible book. [S1][S2][S3][S6]

The historical comparison argues against describing the stock as distressed. Company Financials indicates that MBIN’s annual average P/E generally ranged from the mid-single digits to high-single digits during 2021-2025 and average price-to-book from roughly 0.9 to 1.2 times. Historical multiples are imperfect because the company grew rapidly, changed its capital structure and passed through unusual mortgage and credit cycles. Even so, the current earnings and book multiples sit above most of that history. [S6]

A specialized peer set illustrates why no single multiple is decisive. Prices are September 14 closes, while earnings and book denominators are based on the latest reported quarter and reconciled to primary releases.

The five-peer median is approximately 9.6 times earnings and 1.67 times tangible book or the closest disclosed proxy. Applying those medians mechanically gives conflicting answers: roughly $48 from earnings and $67 from tangible book. A tighter Customers Bancorp and Western Alliance book comparison points closer to $49. That conflict is informative. Merchants earns less consistently than the highest-book-multiple specialist banks, while the market has not dismissed its book value or franchise. [S16][S26][S27][S28][S29][S30]

Peer definitions must be treated carefully. The Bancorp release provides book value rather than a fully reconciled tangible-book measure. Axos excludes servicing rights in its tangible measure, whereas Merchants’ company-defined tangible book retains its separately recognized MSR. Western Alliance and Customers carry different loan and deposit concentrations. The table is a valuation framework, not a claim that all banks deserve the same multiple.

Reverse valuation is more useful. At the 9.6-times peer earnings median, $53.14 requires normalized EPS of approximately $5.54. At 10.5 times, it requires $5.06. The current price is already supported by reported tangible book at 1.33 times, but a larger book multiple would require confidence that common returns remain in the mid-teens and credit does not consume retained earnings. The embedded earnings expectation is therefore roughly $5.00-$5.55, above trailing $4.98 but below a simple annualization of Q2.

The scenario framework uses both EPS and tangible book because either can mislead after a credit event:

The bear case assumes credit remains costly, fee recovery is weaker, capital return is minimal and the market values the bank near tangible book. The base assumes criticized assets continue to cure, industry activity supports fees, credit cost normalizes above Q2 but far below 2025 and retained earnings rebuild book. The bull requires both operating recovery and a quality rerating: recurring fee growth, no new concentrated losses, stable funding, limited dilution and sustained mid-to-high-teens common returns.

Using 25%, 50% and 25% probabilities produces probability-weighted arithmetic of approximately $61.3 before dividends. The most fragile bullish assumptions are the $20-$30 million provision, 1.65-times book multiple and absence of new issuance. The most fragile bearish assumption is continued tangible-book growth while provision remains high. None of these figures is company guidance.

The market is right that acute regulatory and solvency risk has declined. It is also right to withhold the highest specialist-bank book multiple until capital rebuilds and post-remediation vintages demonstrate lower loss. The current price no longer offers the asymmetry available near $30-$40; future performance depends on earnings conversion rather than survival.

Verdict. Merchants is inexpensive on peer tangible book but slightly expensive on peer trailing earnings. The current valuation embeds substantial credit normalization and requires durable common EPS above approximately $5 without another capital raise.

Variant Perception

The strongest bullish view is that 2025 represented a finite cluster of troubled multifamily relationships rather than a broken franchise. Evidence includes the concentration of losses in five 2025 relationships, lower sequential criticized and NPL balances, repayments driving most Q2 NPL improvement, the terminated MOU, record tangible book, strong warehouse earnings, higher agency capacity and officer purchases near the trough. [S2][S4][S13][S19][S20]

The strongest bearish view is that Merchants grew too quickly into correlated multifamily exposure, funded expansion with common and preferred issuance and remains underreserved relative to problem assets. The provision-driven rebound, NPLs above year-end, rising OREO, 36.9% NPL coverage, concentrated first-half charge-offs, short-duration brokered funding, lower capital ratios, founder control and founder sales support that skepticism. [S1][S2][S5][S17][S18]

A more useful variant is narrower: the franchise is better than the 2025 loss episode implied, but normalized earnings are weaker than the Q2 headline and the current rerating imply. Agency standing, servicing scale, warehouse volume and relationship deposits validate the platform. Provision relief, fair-value income and capital consumption argue against annualizing the latest quarter. This view expects positive book compounding without assuming an immediate return to 2024’s peak return.

Five load-bearing assumptions determine the outcome:

1. Legacy credit cures faster than new problems emerge. Bull evidence is declining NPLs, criticized loans, OREO and realized loss for four quarters, with reductions driven by repayment or sound takeout. Bear evidence is migration of previously pass-rated loans, new fraud disclosures or additional relationship-level concentration. [S1][S2]

2. Mortgage activity produces operating leverage. Bull evidence is higher gain-on-sale, contractual servicing fees and warehouse profit while NIM and efficiency remain stable. Bear evidence is volume absorbed by spread compression, producer compensation and wholesale funding. [S2][S3]

3. Capital rebuilds despite asset growth. Bull evidence is rising CET1, total capital and tangible-common-equity ratios alongside TBVPS. Bear evidence is another preferred or common issue, lower capital ratios or no capacity to repurchase even when valuation becomes favorable. [S1][S17]

4. Remediation changes vintage loss behavior. Bull evidence is low loss on post-2025 originations with stronger leverage, appraisal, recourse and fraud controls. Bear evidence is recurrence of subordinated-product, appraisal or sponsor-control failures. [S3]

5. Normalized common EPS exceeds roughly $5.25. That supports the market price without an exceptional multiple. Results below $4.50 would expose overvaluation; results above $6 accompanied by higher capital ratios would support a better quality assessment.

The factor model supplies positioning context rather than a business conclusion. As of September 11, 2026, market exposure was 0.97, small-size exposure 0.83, growth exposure negative 0.51 and CreditRisk exposure positive 0.16. The Financials coefficient was positive 0.23, but every sector coefficient is a statistical return exposure, not an industry classification. R-squared was only 0.274, residual momentum 0.070, residual Sharpe 1.05 and residual volatility 0.329. Most return variation remained unexplained by the model, consistent with—but not proving—company-specific credit relief. [S15]

Management commentary is difficult to pressure-test. Company Financials returned no transcripts for either of the latest two quarters, and no public earnings-call Q&A record was located. The releases contain prepared statements from Petrie and Dunlap about platform strength, liquidity and opportunity. Those statements should be treated as management claims. Filed balances, realized charge-offs, appraisals, capital and subsequent migration deserve greater weight than optimistic adjectives. [S2][S4][S6]

The most decision-useful investor questions concern NPL cure mechanics, collateral values, post-remediation underwriting, warehouse incremental returns, brokered funding, recurring servicing cash fees and the capital threshold for repurchases. Specifically: what proportion of NPL reductions came from cash repayment versus refinancing, foreclosure, charge-off or OREO; what current LTV and DSCR support criticized assets; what was the risk-adjusted spread on new warehouse volume; and why did short-duration brokered deposits rise when core funding was described as strong? [S1][S2][S3]

Two reusable analytical rules survive testing. First, common bank returns must deduct preferred dividends, redemption effects and preferred claims; applying that rule lowers reported 2025 common income from the dividend-only approximation to $173.6 million. Second, valuation must be recomputed after a large price gap; the move to $53.14 removes the distressed multiple that existed at the 2025 trough. A mortgage-REIT rule separating distributable income from comprehensive return is only partially transferable because Merchants is a bank, although the need to separate contractual servicing fees from marks remains useful. [S1][S3][S6]

Most retrieved cross-company concepts—biotechnology research accounting, foreign regulatory transfer, restaurant launches, chemicals, utilities and contributed-data scale—have no decision-useful application here and should not be imported. The absence of relevance is itself a quality control: analogies should follow mechanisms, not keyword similarity.

Verdict. Consensus can be wrong in both directions. The finite-loss bull case has credible evidence, but the market is capitalizing an earnings recovery that still depends heavily on provision normalization and successful capital rebuilding.

Fact vs. Interpretation

Several apparent contradictions resolve when definitions are corrected. The controlled five-year price low is $16.04 on a split-adjusted closing basis; $21.50 was the 2023 panic’s intraday low. The newest MBA forecast has a higher dollar estimate but a much lower growth rate than January because the 2025 base was revised. Proxy-defined revenue includes the effect of provision, whereas Company Financials’ standardized revenue is a different analytical series. The 2024 cash settlement associated with transferring branch assets and deposits was a divestiture, not a new acquisition. [S6][S7][S10][S25]

Verdict. The facts establish a valuable platform and sequential credit improvement. They do not establish that current EPS, collateral recovery, funding mix or valuation is sustainable without further evidence.

Open Questions

1. How much of the Q2 NPL decline was cash repayment, permanent refinancing or agency takeout, and how much came from charge-off, sale, foreclosure or transfer to OREO? Management disclosed that repayments were the largest factor, but not the complete waterfall. [S1][S2]

2. What are current appraised values, debt yields, LTVs and debt-service coverage ratios for criticized multifamily and healthcare loans, net of taxes, senior claims and liquidation costs? [S1][S3]

3. What deficiency claims or recourse remain after the large 2025 and 2026 relationship-level charge-offs, and what recoveries are expected?

4. How do post-remediation vintages differ in leverage, appraisal review, borrower concentration, recourse, subordinated exposure and fraud verification? [S3]

5. What proportion of warehouse growth represents durable share, temporary competitor exits or market recovery, and what is the incremental return after funding, operations, expected loss and allocated capital?

6. How much servicing income is contractual cash fee, initial capitalization, runoff, amortization, valuation movement and derivative result? [S2][S3]

7. Why did brokered deposits rise 46% sequentially despite the 91% core ratio, and at what cost will the 51-day book renew or be replaced? [S2]

8. What consolidated CET1, total-capital and tangible-common-equity ratios must be reached before the board materially uses the remaining repurchase authorization?

9. Will the compensation committee add multi-year charge-off, criticized-loan, risk-adjusted-return or tangible-book measures after exercising discretion in 2025? [S5]

10. What premiums, attachment points, exclusions, counterparty limits and capital benefits apply to each credit-risk-transfer structure, and will future protection cover legacy criticized assets or only performing pools? [S1][S2]

11. How much holding-company cash is available after preferred and common dividends, and what restrictions apply to upstream dividends from the bank?

12. What evidence would management provide to show that the Illinois divestiture and focus on national mortgage finance improved long-term risk-adjusted return rather than simply increasing concentration? [S7]

What Must Be True

Bull tests

• Criticized loans must fall below 3% of gross loans and NPLs below 1.25% by mid-2027, with reductions driven primarily by cash repayment, sound refinancing or agency takeout rather than charge-offs or OREO transfers. The starting points are 3.6% and 1.67%. [S2]

• Annualized net charge-offs must remain below 0.35% of average loans for four consecutive quarters, with no new relationship producing more than 15% of the period’s losses. This directly tests the 2025 and Q2 concentration pattern. [S2][S3]

• OREO must decline from $72.4 million and realized recoveries must validate current collateral assumptions. A lower NPL ratio without lower OREO and loss would not pass. [S1]

• Common ROTCE must exceed 15% for four quarters after separately showing material servicing and derivative marks. Q2’s 14.95% is the initial benchmark, not proof of durability. [S2][S4]

• Gain-on-sale, contractual servicing fees and warehouse segment profit must grow with transaction activity while the efficiency ratio remains below 42% and NIM at or above 2.75%. Volume growth alone is insufficient. [S2][S10][S11]

• CET1 and total capital must rise from 9.3% and 12.5% while tangible book grows and diluted shares remain near or below 46 million. [S1][S2]

• Post-remediation multifamily vintages must avoid the appraisal, subordinated-credit and suspected-fraud failures disclosed for older relationships. [S3]

• Common repurchases, if made, must occur at a price offering a higher expected common return than preferred redemption or incremental balance-sheet growth.

Bear tests

• The bear case strengthens if NPLs remain above $200 million through mid-2027, criticized balances reaccelerate or OREO exceeds $100 million. [S1][S2]

• A rolling four-quarter provision above $90 million without a comparable decline in criticized balances would indicate that 2025 was not a finite episode. [S3]

• Normalized common EPS below $4.50 after separately identifying favorable servicing and derivative marks would show that the current price embeds excessive recovery. Trailing reported EPS is approximately $4.98. [S1][S3]

• CET1 below 9%, total capital below 12% or another material common or preferred issue would indicate that growth and losses are outrunning internal capital. [S2][S17]

• Brokered deposits above 12% of deposits, a material renewal-cost spike or NIM below 2.6% would challenge the funding advantage. [S2]

• Loss of agency approval, disappearance from affordable-housing rankings or warehouse volume growth accompanied by falling segment profit would weaken the competitive thesis. [S3][S9]

• Repeated discretionary compensation after credit and return thresholds are missed, or another material related-party transaction without compelling economics, would require a larger governance discount. [S5]

The bullish thesis is falsified if lower reported NPLs do not translate into lower realized loss, lower OREO and stronger capital. The bearish thesis is falsified if post-remediation vintages remain clean, recurring fees rise, capital rebuilds and common returns remain above 15% without provision releases or favorable marks. These thresholds should be checked against the next filed 10-Q, earnings release and insider filings rather than management tone. [S1][S2][S23]

Public source appendix

• S1: Merchants Bancorp Form 10-Q for the quarter ended June 30, 2026 — SEC filing; published 2026-08-07; Part I, Items 1-2; financial statements; credit quality, servicing, credit-risk-transfer, segment and capital disclosures
• S2: Merchants Bancorp Reports Second Quarter 2026 Results — Company earnings release; published 2026-07-28; Financial highlights; asset quality; deposits; liquidity; margin; provision; servicing and capital tables
• S3: Merchants Bancorp 2025 Form 10-K — SEC filing; published 2026-02-27; Items 1, 1A, 7 and 8; Notes 5-7, 17-18 and 23; five-year, credit, servicing, preferred and segment tables
• S4: Merchants Bancorp First Quarter 2026 Results and MOU Termination Disclosure — SEC-filed earnings release; published 2026-04-28; Regulatory update; repurchases; operating results; servicing and derivative fair-value disclosures
• S5: Merchants Bancorp 2026 Proxy Statement — SEC filing; published 2026-04-10; Ownership; board independence; compensation definitions, targets and discretion; related-party transactions
• S6: Company Financials — MBIN profile, statements, ratios, valuation, prices and transcript availability — Third-party financial data, reconciled; published 2026-09-14; NASDAQ:MBIN resolved before retrieval; multi-period API data reconciled to filings; split-adjusted price history through September 14, 2026
• S7: Merchants Bancorp 2024 Form 10-K — SEC filing; published 2025-02-28; Items 1, 7 and 8; Illinois branch-sale note; securitization, financial and segment tables
• S8: Merchants Bancorp 2023 Form 10-K — SEC filing; published 2024-03-12; Items 1, 7 and 8; financial statements, warehouse volumes and risk factors
• S9: Fannie Mae 2025 Multifamily Lender Rankings — Government-sponsored-enterprise release; published 2026-02-04; Overall DUS and category rankings; affordable-housing production and definitions
• S10: MBA April 2026 Commercial and Multifamily Forecast — Industry-association forecast; published 2026-04-21; Forecast table for 2025-2028 total commercial and multifamily originations
• S11: Commercial and Multifamily Borrowing Increased in Second Quarter 2026 — Industry-association data; published 2026-08-06; Overall and multifamily-specific year-over-year and sequential originations; investor-channel mix
• S12: MBA 2025 Annual Report on Multifamily Lending — Industry-association data; published 2026-08-13; New mortgage volume, lender count, market fragmentation, GSE share and leading lenders
• S13: FHFA Announces 2026 Multifamily Loan Purchase Caps — Regulator release; published 2025-11-24; 2026 enterprise caps, mission-driven requirement and treatment of workforce housing
• S14: Fannie Mae Second Quarter 2026 Multifamily Financial Supplement — Government-sponsored-enterprise results; published 2026-07-29; Multifamily guaranty-book composition, delinquency, REO, provision, charge-offs and vintage metrics
• S15: the factor model — MBIN exposure snapshot — Quantitative diagnostic; published 2026-09-11; Exposures, residual signals and diagnostics dated September 11, 2026
• S16: Company Financials — specialized-bank peer valuation snapshot — Third-party peer data, reconciled; published 2026-09-14; Exchange-qualified MBIN, CUBI, TBBK, AX, QCRH and WAL prices and TTM fundamentals through September 14, 2026, reconciled to peer releases
• S17: Merchants Bancorp Announces Pricing of Common Stock Offering — Company capital-markets release; published 2024-05-13; Offering price, share count, proceeds and intended use
• S18: Michael Petrie Form 4 — SEC insider filing; published 2026-05-07; May 5-6, 2026 open-market common-share sales and post-transaction ownership
• S19: Michael Dury Form 4 — SEC insider filing; published 2025-07-31; July 30, 2025 code-P open-market purchases totaling 16,000 common shares
• S20: Sean Sievers Form 4 — SEC insider filing; published 2025-08-04; August 1, 2025 code-P purchase of 2,500 common shares at $29.75
• S21: Merchants Bancorp 2022 Form 10-K — SEC filing; published 2023-03-16; Business, mortgage-cycle, liquidity, capital and financial statements
• S22: MBA 2026 Commercial and Multifamily Mortgage Maturity Survey — Industry-association data; published 2026-02-09; Total maturity volume, multifamily share and capital-source maturity schedule
• S23: SEC EDGAR Company Filings — Merchants Bancorp — Regulatory filing index; publication date unavailable; Trailing 60-month filing index covering Forms 10-K, 10-Q, 8-K, proxy and Section 16 filings
• S24: Merchants Bancorp 2021 Form 10-K — SEC filing; published 2022-03-04; Business, financial statements, capital and share data
• S25: MBA January 2026 Commercial and Multifamily Forecast — Industry-association forecast; published 2026-01-22; January forecast table for 2025 estimates and 2026 originations
• S26: Customers Bancorp Reports Second Quarter 2026 Results — Peer company earnings release; published 2026-07-23; EPS, tangible book, return on common equity and asset quality
• S27: The Bancorp Reports Second Quarter 2026 Results — Peer company earnings release; published 2026-07-30; EPS, book value, ROA and ROE
• S28: Axos Financial Reports Fiscal 2026 Results — Peer company earnings release; published 2026-07-30; Fiscal-year EPS, tangible book definition and Q4 ROE
• S29: QCR Holdings Announces Second Quarter 2026 Results — Peer company earnings release; published 2026-07-22; EPS, tangible book, ROE, asset quality and LIHTC activity
• S30: Western Alliance Bancorporation Form 10-Q for the quarter ended June 30, 2026 — Peer SEC filing; published 2026-07-24; Common earnings, tangible book reconciliation, ROTCE and credit quality