CDFI Merger Analysis Must Count Donor Losses Alongside Cost Savings

Brett Simmons treats financial analysis as an early test of whether a potential CDFI merger is worth exploring—not as a substitute for mission alignment or a final decision. The assumption is that organizations considering a transaction have already asked what they could do together that they could not do separately. The next question is whether the combined organization could carry that mission further with each dollar available to it.

That question can cut both ways. A merger might improve the combined organization’s financial performance, or its ratios could be worse than either partner expects. Simmons’s method is designed to make that possibility visible quickly, before leaders spend substantial time building a detailed forecast. It is a screening tool, not a board-ready financial model: a final assessment would still require closer examination of accounts, portfolios, donor relationships, and other transaction specifics.

The framework responds to a shift in the financial concerns CDFIs had been raising with the Aspen Institute’s Microfinance Impact Collaborative, a group of six large CDFI microlenders that the Business Ownership Initiative has convened for about 15 years. Joyce Klein said that members’ immediate focus had moved from balance-sheet questions—such as net assets and raising capital—to income-statement pressures: generating revenue, managing expenses, and understanding what higher costs of capital could mean for financial performance.

The resulting framework connects six variables: earned revenue yield, cost of funds, provision expense, operating expense, contributed revenue, and total portfolio assets under management. The first five describe income, costs, and support; total portfolio assets under management supply the scale used to weight the percentage measures across organizations.

The calculation proceeds in stages. Earned revenue yield, less cost of funds, produces net margin, or the earning spread. Subtracting provision expense gives risk-adjusted net margin. Subtracting operating expense produces total portfolio return—the structural operating result before contributed revenue. Adding the rate of contributions produces total return on all activity, or net institutional margin. This makes explicit how portfolio economics and philanthropy combine in the institution’s overall result.

That sequence matters because it distinguishes the economics of lending from the contribution support that helps sustain the institution. A CDFI can have a structural deficit in its portfolio operations and still reach a positive total return once contributed revenue is included. Conversely, a merger that improves lending efficiency may not improve the institution’s overall position if philanthropic support falls at the same time.

Klein said the framework helped collaborative members interpret their financial performance and inform planning. The webinar series was designed to turn that analysis into practical tools for strategic choices, rather than a set of papers. The emphasis is on speed and usability: a model that requires too much data gathering or calculation may not get used. The first-pass method is intended to help leaders see whether a transaction merits further investigation and identify what would have to change to make it worthwhile.

That boundary also shapes what the framework cannot settle. Historical audited financials can establish a baseline, but they do not by themselves show how a particular portfolio will perform after integration, whether specific donors will continue giving, or what the combined organization’s detailed budget will look like. Those questions require deeper work. Simmons’s aim is to make the early conversation more concrete: which assumptions drive the result, what evidence is missing, and whether it is worth investing in a fuller analysis.

Financial benefit is only one reason to consider a merger, but mission alignment quickly becomes a margin question. Can the organizations do more with their available resources after joining? A positive answer cannot be assumed just because the missions fit.

Simmons cautioned against treating projected synergies as a likely outcome. He cited research figures indicating that about 20% of mergers achieve their stated strategic and financial targets, 60% fail to generate net tangible economic or mission value, and nearly half are eventually restructured, dismantled, or reversed. He did not identify the underlying studies in the presentation, so these are figures Simmons cited, rather than independently sourced estimates established here. He offered them as a reason to examine the odds carefully, not as a reason to rule out mergers. His point was that leaders can overestimate potential gains and underestimate the chance of failure.

“You are statistically more likely to destroy value than create it.”

The method adapts a public-company merger analysis to CDFIs. In the public-company version, analysts compare companies’ values before and after a merger announcement, account for the transaction terms, and examine how value changes for each shareholder group. Simmons used Comerica and Fifth Third as an example of market reaction after a merger announcement—not as an account of the completed results of a merger.

In his description of the announcement period, Comerica shareholders gained about $1.37 billion in value, while Fifth Third shareholders lost about $480 million. The combined market value increased by about $890 million. Those figures describe the change in the market’s valuation immediately after the announcement, before the merger had occurred.

The net increase was the market’s implied value created by the announced deal. Comerica’s gain amounted to 154% of that total, meaning the market’s assessment was that Comerica shareholders captured more than all of the expected synergy while Fifth Third shareholders absorbed dilution and integration risk. The example makes a distinction that matters for nonprofit transactions too: a combination can be expected to create value overall without distributing that value evenly.

CDFIs do not have traded shares to measure, Simmons noted. Instead, leaders can establish a baseline from audited financial performance, adjust for the relative sizes of the organizations, and model plausible changes in the combined organization’s ratios. The result is not a market verdict. It is a structured way to examine where possible gains or losses could come from—and how much confidence the case requires.

The first practical rule is to combine financial amounts before calculating ratios. A $100 million lender and a $20 million lender do not produce a 50/50 average of their ratios. Averaging the ratios directly would give the smaller organization the same weight as the larger one. Instead, add the relevant dollar amounts, average them over the period being analyzed, and then calculate the ratio against the combined average portfolio assets. This automatically weights the result by scale.

Simmons illustrated the method with two composite CDFIs. Their combined baseline earned revenue yield was 15.24%, compared with 17.57% for the first organization and 14.25% for the second. The combined cost of funds was 1.84%, between their respective rates of 2.10% and 1.75%. The resulting net margin was 13.40%. Their combined provision expense rate was 4.26%, between the individual rates of 3.66% and 4.47%; after provision expense, risk-adjusted net margin was 9.14%. The combined operating expense ratio was 20.43%, compared with 27.45% and 17.94%, respectively. Subtracting operating expense produced a total portfolio return of negative 11.28%. Contributed revenue yield was 14.59%, resulting in total return on all activity of 3.31%.

The figures show why the baseline is more than a bookkeeping exercise. One organization was more efficient in its operating costs, while the other brought more philanthropic revenue on average. Once the amounts are weighted, leaders can see how much each partner contributes to the combined starting point—and how much room a proposed improvement has to change it. The composite figures are illustrative; Simmons said he had combined characteristics from different CDFIs, and that the organizations represented were relatively large.

From that baseline, Simmons recommends constructing a “better” and a “worse” case for each variable. These are not claims about the best or worst imaginable outcome. They are estimates of plausible movement, based on what leaders know about the organizations and what they believe could change. The baseline itself is already weighted by dollars, so adjustments should be made from that combined ratio rather than by averaging the two organizations’ ratios again.

For earned revenue, relevant questions include whether one organization has a product that could be offered across a wider footprint, or whether lower rates are needed to preserve access to public funding. In the example, the combined baseline was 15.24%; the better case was 16.24%, and the worse case 14.24%. Simmons also warned that an organization charging higher rates might not be able to maintain them after combining with a partner whose public funders impose interest-rate caps. A product that works in one organization or geography may also be a niche that cannot simply be expanded across the combined footprint. The case for an increase should therefore identify what drives it: a product’s broader reach, a change in pricing, or some other specific mechanism.

Cost of funds may move in the less favorable direction. Combining balance sheets can increase exposure to particular lenders or use up available low-cost capital. Simmons said lenders may limit exposure to a larger combined entity, forcing it to seek more expensive debt; banks that provide low-cost capital for Community Reinvestment Act reasons may also have little incentive to increase lending once they have met their obligations. In the example, the combined cost of funds was 1.84%. The better case stayed at that level, while the worse case rose to 3.84%. Simmons’s broader rule of thumb was that an already-low cost of funds is unlikely to improve just because organizations merge. Leaders should test whether the merger would actually preserve existing facilities and relationships, rather than counting on scale alone to lower borrowing costs.

Provision expense requires a different kind of judgment. Existing loans have already been originated, so much of the portfolio risk is already present. A partner with stronger servicing and delinquency management might help improve performance over time. A messy portfolio, by contrast, could require a higher initial provision and intensive attention, potentially drawing effort away from other work. Simmons’s example put the combined provision rate at 4.26%, with a better case of 3.66% and a worse case of 4.47%. Leaders also need to consider whether different credit standards affect earned revenue: higher-risk lending may be associated with higher interest rates. Simmons described the scenario work as a way to reason about these comparative advantages at a high level, not a substitute for examining the actual portfolios.

Operating expenses are often where merger proponents expect the largest savings. Simmons focused on personnel and technology, while noting that technology savings may be more limited when spending is subscription- or user-based. In the example, the combined operating expense ratio was 20.43%; a better case brought it to 19%, while the worse case rose to 25.43%.

He translated the ratio into organizational consequences: a 1.4 percentage-point improvement could represent roughly $250,000 in technology consolidation and the equivalent of more than 20 full-time positions. The slide described the better case as saving $3.65 million annually in practice. Those figures are not an automatic staffing plan; they are a reminder that efficiency gains correspond to real costs, jobs, and organizational change. If savings depend on eliminating positions, morale and culture become part of the financial assessment, not separate afterthoughts.

The difference between the two partners’ expense ratios is also a warning against assuming that the more efficient organization’s practices can be adopted without friction. Simmons said a less efficient organization may not simply absorb its partner’s methods; integration can muddle processes and erode the expected performance. The better-case assumption should therefore include a plausible account of how practices transfer, who is responsible for making changes, and how long it may take. The worse case should reflect the possibility that the combined organization retains more of the higher-cost structure than leaders expect.

For CDFIs, a merger analysis that stops at operating costs misses a central part of the business model. Simmons said that even organizations generating substantial portfolio income often rely on philanthropy to cover the gap and fund new programs. But unlike portfolio yield or provision expense, contributed revenue cannot be projected from a simple ratio alone. It requires reviewing who gives to each organization and how those donors may respond to a combination.

He described four donor categories. Public entity-based awards, such as CDFI Fund awards, may not double when two recipients combine; the merged institution may receive one award rather than two. Public volume-based support, such as an SBA technical-assistance grant, may grow with portfolio volume, up to a program cap. Private donors who support only one partner may continue giving to the enlarged organization. Where the same donor supports both, leaders should not assume that two grants will simply become one grant twice as large. Simmons suggested testing a reduction—for example, combining two $50,000 gifts into an assumed $75,000 contribution.

This categorization is a prompt for donor-level investigation, not a substitute for it. For each public award, leaders need to determine whether it is tied to a single eligible entity, whether it scales with volume, and whether the program has a cap. For private support, they need to distinguish donors with a relationship to only one organization from donors already funding both, then test different assumptions about whether the donor will maintain, increase, or reduce support. A single across-the-board haircut can conceal important differences among grant types and relationships.

This potential reduction is what Simmons called philanthropic “capture”: donors may retain some of the financial value created by a merger by reducing their support. In a scenario where operating improvements produce $5 million in savings but annual philanthropic revenue falls by $5 million, the donors have effectively captured all of that synergy. Simmons suggested that organizations could make the case to funders that savings should be reinvested in the combined institution, but said leaders should not assume funders will agree.

The contributed-revenue slide and Simmons’s narration do not line up cleanly. The slide lists a combined baseline of 14.59%, a better case of 11.85%, and a worse case of 11.06%, while the narration describes a 24% reduction as a worse-case assumption and says that, alongside the better-case operating benefits, roughly 90% of the upside would go to philanthropic partners through lower giving. The figures and labels therefore do not establish a clear directional comparison between the slide’s “better” and “worse” cases. The spoken example is clearer on the analytical point: for organizations with substantial donor overlap, Simmons said a 10% to 25% reduction in contributed revenue would not be unexpected.

That change altered the bottom line. The example’s total return on all activity was 3.31% at the combined baseline and 3.59% in the better case; the worse case produced a deficit of 8.43%. These figures illustrate why donor analysis cannot be left until the end: it can materially change whether operational improvements translate into lasting institutional benefit.

The quick screen can flag donor overlap, but it cannot reliably resolve it. Simmons said a serious estimate requires going line by line through contributors, sorting them into the four categories, and considering how each donor might respond. The answer may depend on whether a grant is tied to a particular entity, grows with lending volume, or comes from a donor that supports both organizations. Those details are specific to the transaction and cannot be inferred from a general contributed-revenue ratio.

A plausible upside is not enough. Simmons’s method compares the expected outcome against both the transaction’s costs and the risk that the better case will not occur. Mergers require legal counsel, board and management time, audit expenses, transition work, and often severance or vendor-exit costs. They also divert leadership attention from day-to-day work.

As a rough guide, Simmons estimated transaction costs at 1% of combined operating expenses for a low-complexity deal, 2% for a typical consolidation, and 3% or more where there are major office changes, a headquarters relocation, or management-team restructuring. He advised assuming a floor of $500,000, even for a small acquisition. Management time, he emphasized, is a significant hidden cost. These are rules of thumb for an initial estimate, not fixed charges that apply to every transaction.

The categories are intended to make leaders account for complexity, not to imply that every expense can be predicted as a clean percentage. A low-complexity case might involve little staff impact, no offices to combine, and an easy technology transition. A more typical consolidation may involve staff changes, multiple systems, and technology work. Office closures, headquarters relocation, or management restructuring point to a more complex and potentially more expensive integration. Leaders should connect the cost estimate to the actual work the transaction entails, including the leadership time that may otherwise remain invisible in a spreadsheet.

For the composite example, Simmons estimated $5 million in direct and indirect merger costs. That $5 million is an illustrative assumption for the example, not a universal threshold. To demonstrate the probability calculation, he used an illustrative 80% chance of the worse case, at negative $21.5 million, and a 20% chance of the better case, at positive $9.5 million. On those assumptions, the expected outcome was negative $15 million, falling short of a $5 million target to cover the estimated friction cost.

That 80/20 calculation is distinct from the earlier research figures Simmons cited. It is an assumption applied to the two outcomes in this particular example; it is not presented as a measured probability for this CDFI transaction or as a direct statistical estimate derived from those cited figures. The calculation demonstrates how a decision changes when leaders weigh a downside and an upside by their assumed likelihoods.

Changing the assumptions changes the result. At a 10% chance of the worse case and a 90% chance of the better case, the expected outcome became positive $6.1 million, above the $5 million target. But Simmons described that reversal as highly unlikely. The point of the exercise is not to choose whichever odds make a deal look attractive. It is to test how optimistic leaders would have to be for the transaction to clear their chosen threshold.

Qualitative factors can inform that judgment: whether the leadership teams have worked well together, whether boards have merger experience, and whether the organizations already share technology or services. Simmons suggested that a move from 20% to 30% or 40% odds of success might be defensible with good reason. If a deal only works when success is assumed to be 90% certain, he argued, it is probably not worth pursuing. The upside should be strong enough to justify the risk without requiring everything to go right.

The probability test also helps focus the follow-up work. If an expected return is poor under reasonable assumptions, leaders can identify what would have to change: perhaps the organizations would need to reduce costs, preserve low-cost borrowing, or demonstrate that donors are likely to maintain support. If the case depends on a large shift in one variable, that assumption becomes a question to investigate rather than a benefit to take for granted.

The analysis is meant to be iterative. Once the baseline is in place, leaders can ask how many percentage points or basis points a variable would need to move for a transaction to become worthwhile. If the answer depends on operating efficiency, that may point to work the organizations should do before merging—or instead of merging.

Simmons noted that organizations can often address high operating costs or loan pricing that suppresses earned revenue without joining with another institution. He recommended considering those options first. A merger may become more attractive later if the partners improve their performance or develop practices that can be carried into a combined organization. The analysis can help identify which variable matters most: for example, whether preserving low-cost capital changes the result materially, or whether operating savings would have to do most of the work.

The same approach applies to acquisitions, though a much larger partner’s scale may wash out the smaller organization’s effect on the combined ratios. An acquisition might make sense because the smaller organization is more efficient, has strong fundraising relationships, or brings a local donor base; it might also bring a portfolio or capabilities the larger organization lacks. The weighted analysis can help show whether those differences are large enough to matter.

One-time philanthropic support for a transaction can reduce the cost target, Simmons said, but it does not remove the need to account for recurring donor changes. If a funder contributes $3 million toward a transaction estimated to cost $5 million, the organization could test whether the expected benefit clears a $2 million target. It would still need to assess whether annual support from existing donors might decline after the merger.

Simmons also described a pattern from the cases he had examined: organizations that overlap in geography or product type may face greater donor overlap and therefore larger philanthropic reductions. To make the numbers work, those combinations may require substantial operating savings—sometimes involving cuts of 20 or 30 staff. He said some potential mergers had been called off after organizations confronted the difficulty and organizational weight of those changes. Staff reductions may improve a financial ratio, but the disruption can weigh on an organization for years and make the expected integration harder to achieve.

That pattern connects the spreadsheet to the probability judgment. A scenario that assumes lower expenses may be arithmetically possible but difficult to realize if the savings depend on staff cuts, changes to management, or sustained process redesign. Likewise, a scenario that assumes donor support will hold may be weak if the organizations have overlapping funders or serve similar geographies and products. The point is not to convert every cultural or organizational question into a precise number. It is to make the assumptions visible and ask whether the proposed change is credible enough to support the probability assigned to it.

The first-pass analysis is therefore neither a certification nor a final forecast. If the financial case rests on implausible savings, near-perfect execution, or donor support that is unlikely to persist, leaders can stop early. If the case is close, the model can show which assumptions need more evidence before the organizations commit further time and resources. And if the merger does not make sense now, the analysis may still identify what would need to improve—or whether a narrower option, such as acquiring a partner’s portfolio rather than the whole organization, merits consideration.