코스탈 파이낸셜(CCB) 2026년 3분기 실적 발표회: 이익 회복, 2,000만 달러 절감 목표

코스탈 파이낸셜 코퍼레이션(NASDAQ: CCB)은 이전 분기의 파트너 관련 비용 발생 이후 2026년 3분기에 흑자로 돌아섰다. BaaS 프로그램 수수료 및 부외 거래 활동의 성장이 실적을 견인한 가운데, 경영진은 연간 2,000만 달러 규모의 비용 절감 추진 계획을 제시했다.

핵심 요약

• 3분기 순이익은 1,380만 달러(희석 주당순이익 0.89달러)로, 2026년 2분기 순손실 4,210만 달러 및 전년 동기 순이익 1,360만 달러와 대비된다.
• BaaS 프로그램 수수료 수익은 거래, 정산 및 서비스 수수료 증가에 힘입어 전분기 대비 17.6%, 전년 동기 대비 79.6% 증가한 1,410만 달러를 기록했다.
• 신용손실충당금 전입액은 2분기 9,220만 달러에서 4,600만 달러로 감소했다. 이번 충당금에는 BaaS 대출용 4,390만 달러가 포함됐으나, 비이자수익 내 BaaS 신용보강금 4,180만 달러로 대부분 상쇄됐다.
• 코스탈의 3분기 말 보통주 자본비율(CET1)은 6월 30일 기준 10.86%에서 상승한 11.47%를 기록했다. 현금은 총 8억 300만 달러이며, 8억 2,700만 달러의 비상 차입 한도를 추가로 확보하고 있다.
• 경영진은 일반적으로 환급받는 BaaS 대출 및 사기 관련 비용을 제외하고 연간 2,000만 달러 이상의 비이자비용 절감 기회를 발굴했다. 회사는 향후 4~6분기에 걸쳐 이러한 절감을 추진할 것으로 예상한다.
• 특정 파트너 포트폴리오는 여전히 주요 리스크로 남아 있다. 대출 잔액은 총 5억 1,100만 달러이나, 연환산 순상각률은 2분기 11.46%에서 9.81%로 개선되었으며 30일 이상 연체 대출 비율은 4.06%에서 3.41%로 하락했다.

주요 재무 데이터

사업 및 영업 실적

코스탈은 이번 분기 59억 9,000만 달러 상당의 BaaS 대출을 매각하여 2분기(45억 6,000만 달러) 대비 증가했다. 부외 신용카드 계좌 수는 130만 개에 달했다. 활성 CCBX 체크카드는 전분기 대비 180만 개(20%), 전년 동기 대비 600만 개(118.7%) 증가했다.

부외 예치 예금은 6월 30일 기준 42억 6,000만 달러에서 62억 9,000만 달러로 늘어났다. 해당 잔액에서 발생한 수수료 수익은 2분기 대비 16.4% 증가한 140만 달러를 기록했다. 또한 코스탈은 21개 활성 파트너 전반에 걸쳐 30억 4,000만 달러의 장내 BaaS 예금을 보유하고 있다.

CCBX 네트워크는 다양한 단계의 30개 파트너 관계(활성 파트너 21개, 테스트 1개, 온보딩 2개, 서명된 의향서 3개, 정리 중 4개)로 구성되었다. 8개 파트너 프로그램은 추가 상품으로 확장하고 있었다.

경영진은 심사 기준 강화와 상품 구성 변화가 전체 BaaS 대출 수익률 감소에 기여했다고 밝혔다. 현금 담보 할부 대출은 분기 동안 약 5,000만 달러 증가하여 약 2억 6,500만 달러를 기록했다. 동시에 BaaS 대출 비용 차감 후 대출 수익은 평균 대출금 대비 6.83%에서 6.86%로 상승했다.

커뮤니티 은행 부문은 1.54%의 예금 조달 비용으로 16억 4,000만 달러의 예금과 19억 8,000만 달러의 대출을 기록했다. 순상각액은 사실상 0이었다.

경영진 전망

경영진은 벤더, 외주업체, 인력 및 중복 지출 감소 추진을 통해 연간 2,000만 달러 이상의 세전 절감 가능액을 확인했다. 회사는 향후 4~6분기 동안 이러한 절감을 추진할 계획이지만 사업 니즈 변화에 따라 일부 절감액은 재투자될 수 있다.

코스탈은 목표 효율성 비율을 제공하지 않았다. 경영진은 회사가 레거시 시스템과 중복 비용을 제거하는 동안에도 기존 기술 및 컴플라이언스 플랫폼이 지속적인 성장을 지원할 수 있다고 설명했다.

밸리 내셔널 뱅코프의 블루바인 인수 계획과 관련해, 경영진은 거래가 종료된 후 해당 관계가 끝나더라도 그 변화가 코스탈의 유동성이나 자금 조달 구조에 중대한 부정적 영향을 미치지 않을 것으로 예상한다고 전했다. 코스탈은 대체 자금 조달원으로 부외 예치 예금을 다시 장내 재무제표로 가져올 수 있는 능력을 꼽았다.

리스크 및 관전 포인트

특정 파트너 포트폴리오 문제는 미결 상태로 남아 있다. 경영진은 회수 확률이 2분기와 대체로 동일하며 파트너 및 기타 당사자들과 논의를 계속하고 있다고 밝혔다.

3분기 추가 충당금 비용 중 약 310만 달러는 회수 가능성이 낮은 것으로 간주되어 신용보강 수익으로 상쇄되지 않았다. 관련 신용보강 자산에 대한 평가충당금은 4,600만 달러에서 4,560만 달러로 소폭 감소했다.

2분기에 설정된 2,280만 달러의 특정충당금은 변동 없이 유지되었다. 경영진은 해당 잔액의 필요한 상각 처리가 연말까지 인식될 것으로 예상하면서도 이러한 상각이 시정 조치나 회수 노력에는 영향을 미치지 않을 것임을 언급했다.

분기 말 BaaS 부실채권(NPL)은 총 5,460만 달러였으며, 이 중 5,280만 달러는 파트너 신용보강으로 커버되었다. 전체 BaaS 포트폴리오에 걸쳐 부실채권의 약 96.7%가 파트너 신용보강으로 커버되었다.

블루바인 관련 장내 예금은 약 4억 5,240만 달러였으며, 18억 5,000만 달러는 다른 은행으로 부외 예치되었다. 코스탈은 또한 블루바인이 취급한 대출 300만 달러를 보유하고 있었고, 커뮤니티 은행 부문은 블루바인에 대한 별도의 680만 달러 상업용 대출을 보유하고 있었다.

애널리스트 Q&A 하이라이트

• 특정 파트너 회수: 경영진은 회수 전망에 실질적인 변화가 없다고 밝혔다. 연체율과 상각률 개선은 금리 환경에 따른 CECL 영향과 포트폴리오의 가중평균 만기 연장으로 상쇄되었다.
• 광범위한 CCBX 신용 현황: 코스탈은 21개 활성 파트너를 점검한 결과 유사한 문제가 발견되지 않았다고 전했다. 경영진은 심사 기준 강화, 연체율 감소, 상각률 감소를 긍정적인 지표로 강조했다.
• 비용 절감: 경영진은 연간 2,000만 달러의 잠재적 절감액을 세전 런레이트 기회로 설명했다. 재투자가 이루어질 수도 있으나, 회사는 벤더, 외주업체 및 중복 제거 이니셔티브를 통해 4~6분기에 걸쳐 목표 절감을 달성할 것으로 예상한다.
• 블루바인 전환: 코스탈은 순이익에 미치는 영향에 대한 최종 추정치를 확정하지 않았다. 경영진은 대체 자금 조달이 가능하며 해당 파트너 관계에 국한되었던 감독 비용 일부도 감소할 수 있다고 언급했다.
• 파트너 리스크 관리: 회사는 손실보상 강도 및 포트폴리오 한도를 평가할 때 파트너의 유동성, 현금 담보, 수익 창출 능력 및 매출 다변화를 재평가하고 있다.
• BaaS 대출 잔액: 경영진은 평균 잔액이 3억 달러 증가했음을 지적하며 분기 말 잔액에만 의존하지 말 것을 당부했다. 예년보다 큰 규모의 신용카드 매각도 기말 장내 잔액을 감소시켰다.

실적 발표 컨퍼런스 콜 전문

전체 실적 발표 컨퍼런스 콜 녹취록

경영진 발표

Operator

Welcome to Coastal Financial Corporation's Third Quarter 2026 Earnings Conference Call.

[Operator Instructions]

Before we begin, I would like to point all of you to the disclosure near the end of the company's earnings release for information about any forward-looking statements that may be made or discussed on this call.

The earnings release is posted on Coastal's website. Please review that information along with our filings with the SEC for a disclosure of factors that may impact the subjects discussed in this morning's call.

The company will also be discussing one or more non-GAAP financial measures. Please look at the company's investor presentation and website for all of the disclosures required by the SEC, including reconciliation to the most comparable GAAP numbers.

I would like to now turn the call over to the company. Please go ahead.

Eric Sprink

Thank you, and good morning, everyone. With me today are Chris Adams, our Executive Chairman, and Joel Edwards, our Interim Chief Financial Officer.

Many of you are familiar with Joel from his 13 years as Coastal CFO. Joel has provided that same steady leadership over these last few months and has helped ensure continuity during a transition period.

Our search for a permanent CFO is in its final stages, and we look forward to providing an update in the near future. I also want to share a leadership update.

As part of the ongoing efficiency work Chris will discuss, our management team has worked together to realign roles and responsibilities.

As part of that realignment, Brian Hamilton has stepped down as President of Coastal and moved into a strategic advisory role. Brian has played an important role in modernizing CCBX's technology platform and leading investments in the CCBX infrastructure that have positioned our business for growth.

Mr. Hamilton will remain on the company's Board of Directors and continue to be a partner to Chris, me, and the rest of the Board and team. More broadly, we are focused on creating a more agile organization and aligning our cost structure with the needs of the business.

That effort starts with leadership. As part of this work, I am voluntarily reducing my compensation. I believe it is important that I participate directly in the expense discipline initiatives that we are applying across the business, and I am already encouraged by our progress to date on this front.

Consumers and businesses continue to move more of their financial lives to the digital channels. And the financial ecosystem continues to produce innovative products, business models, and operators.

These companies rely on a bank partner with the technology, compliance, and risk infrastructure to scale with them. Coastal has spent years building exactly that, and we have worked hard to become a partner of choice in this market.

We made a deliberate decision well before fintechs broadly turned to bank partners to gain exposure to this fast-growing landscape, and we did the early legwork to establish a strong foothold in the ecosystem.

Our systems allow us to onboard, monitor, and grow partner programs with discipline, and our model is structured so that the partners provide credit enhancements against losses on the programs they originate.

That combination is not easy to replicate. You can see that in our pipeline with 2 partners in onboarding and 3 signed letters of intent, and in the growth of our program fee income this quarter.

We are super excited about the opportunities in front of us.

Turning to the quarter. I want to begin where we left off in July. In the second quarter, we recorded $68.8 million in pretax charges related to Partners' consumer loan portfolio, resulting in a net loss of $42.1 million.

We know that this was a disappointing result for our shareholders and that it raised questions about our broader partner portfolio.

Over the past 3 months, we have remained focused on updating our assessment of the One Partners portfolio, reassessing the broader CCBX partner portfolio, and continuing to evolve our credit risk appetite as we always do.

Results for the third quarter better reflect the underlying earnings power of the business. We returned to profitability with net income of $13.8 million or $0.89 per diluted share.

BaaS program fee income grew to $14.1 million, up 17.6% from the second quarter and 79.6% from a year earlier, reflecting the strength and stability of our broader partner portfolio.

Importantly, we ended the quarter well capitalized with a common equity Tier 1 ratio of 11.5%. So let's get into the details.

I'll start with an update on one of the partners' consumer loan portfolios and the rest of our BaaS partner portfolio, including Bluevine, which in September agreed to be acquired by Valley National Bancorp, and then conclude with the performance of the underlying business.

Joel will then take you through the financials, and Chris will discuss our work on expense discipline and operating leverage. I want to spend a few moments discussing the specific partner-related portfolio that weighed heavily on our results for the second quarter.

As we have said, we do not see similar issues with the broader CCBX portfolio, but we recognize the significance of this matter and the questions it raised about our partner model.

Our BaaS model is structured so that the partners provide credit enhancement losses on the programs they originate. That structure continues to perform across our partner base. And today, approximately 96.7% of BaaS nonperforming loans are covered by partner credit enhancements.

What set last quarter apart were circumstances specific to one partner and its portfolio under that structure. We absorbed that impact, returned to profitability in the following quarter, and strengthened our capital ratios. That is what a resilient model looks like.

In the third quarter, we updated our assessment of the specific portfolio, including loan portfolio performance, expected credit losses, and our assessment of the collectibility of amounts due under the specific partner's indemnification agreement.

The results substantially support the estimate we established at June 30, although actual charge-offs were modestly above projections.

However, actual net charge-offs as a percent of average loans decreased from the second quarter, reflecting a positive directional trend.

Based on our assessment as of September 30, projected cash flows under the specific Partners program agreement continue to support our accounting treatment to record all of the corresponding incremental credit enhancement amounts substantially.

Approximately $3.1 million of the quarter's additional provision expense was not offset by credit enhancement income because collection of that portion was not considered probable.

The valuation allowance against the credit enhancement asset decreased modestly to $45.6 million as of September 30, compared to $46 million at June 30.

The $22.8 million specific allowance established in the second quarter remains unchanged. And we expect any required charge-off of that balance to be recognized by year-end.

Any such charge-off would not affect our ongoing remediation efforts or ability to pursue recovery. As of September 30, the specific partner portfolio had $511 million in loans outstanding compared with $507.5 million at quarter close for the second quarter on June 30.

Annualized net charge-offs were 9.81% compared with 11.46% for the second quarter, which has markedly improved since the third quarter of 2025.

Loans 30 or more days past due were 3.41% compared with 4.06% for the second quarter.

We continue to closely monitor the performance of the specific portfolio and the specific partner's ability to satisfy its obligations, and we'll update our estimates as additional information becomes available in future quarters.

We have not yet reached a complete resolution, though our discussions are ongoing and have many moving parts. Given the commercially sensitive nature of this matter for both parties, I won't go into more detail, but I can confirm that we continue to pursue and evaluate available pathways to maximize recovery.

Based on our review as of September 30, which covers credit performance, delinquencies, charge-off trends, partner liquidity, collateral funding, and reimbursement obligations across 21 active partners, we did not identify a comparable issue.

The data for the quarter is consistent with that assessment. Banking-as-a-Service nonperforming loans decreased from $14.3 million to $54.6 million, and $52.8 million of that balance is covered by the partner credit enhancements.

We also continue to tighten underwriting standards on new BaaS originations during the quarter, which partially explains why our gross BaaS loan yields declined this quarter.

Every partner operates under different portfolio limits we have set, and we test underwriting on a recurring basis for our partners with the largest exposures.

As of September 30, we had 30 CCBX relationships at varying stages, including 21 active partners, 1 in testing, 2 in onboarding, 3 signed letters of intent, and 4 winding down.

The caliber of the fintech business in our pipeline demonstrates that our approach to risk and partnership continues to resonate in the market.

At the same time, we expect to regularly move deposits on and off the balance sheet as our deposit base grows and our partner businesses evolve.

Valley Bank's recent announcement of the acquisition of one of our partners, BlueVine, represents one dynamic of the evolving fintech ecosystem.

We expect these companies to mature and, in some cases, change ownership over time, and our business model is built to manage those realities.

As of September 30, Bluevine-related deposits on our balance sheet were approximately $452.4 million, and another $1.85 billion in deposits was swept off the balance sheet to other banks.

We also held $3 million of Bluevine-originated loans at quarter end. Separately, the Community Bank held a $6.8 million commercial loan to Bluevine at quarter end.

As of September 30, we held $3.04 billion of BaaS deposits on the balance sheet across 21 active partners, with another $6.29 billion swept off the balance sheet.

Our funding alternatives include the ability to bring sweep deposits back onto our balance sheet, which gives us flexibility to replace Bluevine-related deposits.

We ended the quarter with $803 million of cash and $827 million of additional contingent borrowing capacity.

We do not expect the transaction or any resulting changes to the Bluevine deposit relationship to have a material or adverse effect on our liquidity or funding profile even if our relationship with Bluevine is terminated following the consummation of their transaction.

Now to the underlying business. BaaS program fee income of $14.1 million grew from higher transaction and interchange fees and servicing fees. 8 partner programs were in various stages of expanding into additional products with us, a sign that our partners see Coastal as a platform to grow.

Our capital-light activity continued to scale. We sold $5.99 billion of BaaS loans during the quarter, compared with $4.56 billion of loan sales in the second quarter.

Our off-balance sheet credit card accounts now stand at 1.3 million. CCBX active debit cards, a key source of interchange, increased by 1.8 million cards or 20% quarter-over-quarter and by 6 million cards or 118.7% year-over-year.

Deposits swept off the balance sheet rose to $6.29 billion from $4.26 billion at June 30 and generated $1.4 million of fee income, up 16.4% from the second quarter.

Net interest income was $88.3 million, down 1.2% from the second quarter and up 13.4% from a year ago. Net interest margin was 6.59% compared with 7.27% in the second quarter, primarily reflecting a shift in loan mix.

Net of BaaS loan expense, net interest margin was 3.93% compared with 3.98% in the second quarter, and loan income net of BaaS loan expense was 6.86% of average loans, up from 6.83%.

In the Community Bank, deposits grew $63.2 million for the second quarter, or 4% to $1.64 billion at a cost of deposits of 1.54%. Community Bank loans were $1.98 billion.

Net charge-offs were effectively 0. With that, I'll turn it over to Joel for the financial overview.

Joel Edwards

Thank you, and good morning. I'll keep my remarks brief and focus on a few key figures. More detail is in the release.

Third quarter net income was $13.8 million or $0.89 per diluted share compared with a net loss of $42.1 million or $2.76 per diluted share in the second quarter and net income of $13.6 million or $0.88 per diluted share a year ago.

The provision for credit losses was $46 million compared with $92.2 million in the second quarter. The $46 million provision expense included $43.9 million for BaaS loans, which is largely offset by $41.8 million of BaaS credit enhancements recorded in noninterest income.

$372,000 of the provision was for the Community Bank loans. $1.4 million was for unfunded commitments, and $319,000 related to accrued interest receivables.

The allowance for credit losses was $208.1 million, or 4.97% of loans, compared with $213 million or 5.07% in the second quarter.

Noninterest income included a $502,000 unrealized loss on equity securities, which included a $565,000 downward adjustment to the value of one of our strategic equity investments as of September 30.

Noninterest expense was $84.8 million compared with $141.1 million in the second quarter. This included the $46 million credit enhancement valuation adjustment.

Excluding BaaS loan and fraud expense, noninterest expense was $45 million compared with $96.4 million in the second quarter, which included the $46 million credit enhancement valuation adjustment.

We started to see positive movement on noninterest expense in the quarter, including a $1.6 million reduction in salaries and employee benefits and a $5.9 million reduction in data processing and software, partially reflecting the absence of $4.4 million of accelerated depreciation in the prior quarter.

These reductions in noninterest expense were partially offset by a $1.6 million increase in legal and professional expenses and a $440,000 increase in FDIC charges, both related to the second quarter net loss, while other expenses had smaller changes and effectively netted each other out.

Total deposits were $4.68 billion at September 30, and the cost of deposits was 2.61% compared with 2.57% in the second quarter.

Turning to capital at September 30. Our common equity Tier 1 ratio was 11.47%. Our Tier 1 leverage ratio was 8.81%, and our total risk-based capital was 13.2% compared with 10.86%, 9.11% and 13.3% at June 30.

We had no borrowings outstanding on our contingent funding lines at quarter end.

With that, I'll hand it over to Chris.

Christopher Adams

Thank you, Joel, and good morning. When I became Executive Chair in July, I said one of my priorities would be helping Coastal improve operating leverage and profitability.

Operating leverage means growing revenue faster than controllable expenses over the long term, and we recognize that a resilient CapEx structure is something that we need to establish and prove to our shareholders over time.

Over the past quarter, we've begun translating these priorities into initiatives and specific actions. Coastal has built a scalable, regulated platform through significant investment in people, technology, compliance, data, and risk management infrastructure.

Those investments are critical to the growth Eric described. One of our key objectives now is to run that platform more efficiently while continuing to grow revenue and convert that growth into more consistent profitability.

We're scrutinizing vendors, contractors, personnel, and discretionary spending, eliminating duplication across rules and processes, replacing legacy systems with more efficient technology, and holding every partner program, product, and service to a clear standard for risk-adjusted return.

We are also using AI to improve employee effectiveness and productivity and reduce future costs. Expense results from the quarter show early progress as well as where our work remains.

Data processing and software costs declined $1.5 million from the second quarter after removing the $4.4 million of accelerated depreciation recorded in the second quarter for software assets no longer in use, and salaries and benefits declined $1.6 million from the second quarter.

This work is not about indiscriminate cost cutting.

Our focus is on removing low ROI products or services and duplicative spending and redirecting our people, capital and technology toward our partners, products and services, and capabilities where Coastal is best positioned to win.

That focus is designed to support our customers and partners and produce appropriate risk-adjusted returns. We have identified over $20 million in potential annualized savings in noninterest expenses and expect to achieve these potential savings over the next 4 to 6 quarters.

These expenses exclude BaaS loans and fraud expenses, which are generally reimbursed by our CCBX partners and recognized in noninterest income in accordance with GAAP.

The Board also continues to receive rigorous reporting on the specific partners' performance, collections, collateral, remaining exposure and capital impacts, and counterparty conditions as management pursues recoveries and evaluates the strategic alternatives at its disposal.

The development of our operating efficiency plan during the third quarter was an important step, and we expect operating leverage to improve as the plan is implemented.

Coastal returned to profitability while preserving strong capital and liquidity, along with growing fee income from our off-balance sheet loan sales and deposit sweep activity.

Our responsibility now is to build on that performance with the balance sheet and operating model that supports durable earnings, lower volatility, and stronger shareholder returns.

With that, I'll turn it back to Eric for some closing remarks.

Eric Sprink

Thank you, Chris. The third quarter reflects the strength and earnings potential of our business.

We returned to profitability, group BaaS program fee income and capital-efficient off-balance sheet activity, maintained strong community bank credit performance, and ended the quarter with stronger capital ratios.

The specific partner matter we had in the second quarter was a challenging moment for Coastal, and we have treated it with the seriousness it required. We have made progress in each of the areas we discussed in July.

Our priorities moving forward are clear: maintain disciplined portfolio oversight, manage transitions across our vast partner portfolio, including Valley's anticipated acquisition of Bluevine while maintaining a strong funding profile, expand the products and partnerships that yield appropriate risk-adjusted returns, and convert revenue growth into greater operating leverage and more consistent profitability.

We recognize that 1 quarter does not encompass the breadth of our work ahead, but our capital, liquidity, and operating momentum provide a strong foundation, and we are focused on operating with discipline.

Open for questions.

Operator

Your first question comes from the line of Janet Lee with TD Cowen.

질의응답

Sun Young Lee

Without getting into the specifics of the relationship, based on what you've seen since the second quarter, has the probability of recovering some of the provision and/or credit enhancement valuation adjustment improved, deteriorated, or remained broadly unchanged?

Eric Sprink

Yes, thanks. I think it's materially stayed the same, as you say. We're monitoring it. We've got great open dialogue with the partner.

We still remain hopeful for their progress. We are monitoring our profile, our portfolio. I do want to call out our comments again. The past-due amounts are improving. The charge-offs are improving.

During the quarter, interest rates moved, and that can affect the CECL models on a fixed-rate portfolio. So, consequently, weighted average lives played a big part in needing to increase that allowance, and since we've deemed it unrecoverable, that shouldn't reflect on the underlying analysis of the model, which we think is very consistent with our second-quarter perspective.

Sun Young Lee

And overall CCBx credit, you said you're not seeing other trouble relationships. But given higher rates and a few recent consumer finance headlines, are there any areas you're watching more closely?

Eric Sprink

Yes. One, we're watching all of it very closely. You can imagine internally how much scrutiny all of the portfolios are receiving.

No, we are not seeing anything else stick out or cause any undue attention. We will draw your attention to the overall delinquency and charge-off rates in CCBX, which improved pretty nicely during the quarter and over the last year.

So internally, we also would draw attention to the gross yields going down. That's a product of both product mix changes, but also a higher class of underwriting credits resulting in lower gross yields, while we were able to basically maintain the net yields to Coastal.

So those are 2 indications I draw your attention to, to say, no, we feel confident right now that our oversight and our comments are accurate.

Sun Young Lee

And lastly, on the $20 million annualized savings. Should expenses start stepping down further in the fourth quarter? And is the benefit fairly linear from here?

As we think about '27, is there any efficiency ratio you believe the business can operate at once these actions are further along?

Christopher Adams

Janet, it's Chris Adams. We feel very confident about the potential $20 million that we have targeted. And the work has just started.

So I think that's the way to think about it in terms of on a go-forward basis. So we are digging into all vendors, all contractors, and duplication internally.

We'll continue that work in the fourth quarter and continue to give an update on how it flows through the P&L. On a go-forward basis, we're not giving any projections in terms of how different ratios will work out yet.

But the potential $20 million, we feel like, is a good number to share. We feel very confident in that.

Operator

Your next question comes from the line of Andrew Terrell with Stephens.

Andrew Terrell

Chris, if I could maybe tickle that last line of questioning just on the expenses and the announced $20 million. Just to level set expectations, is $20 million a gross or a net number?

So if we look at your operating expenses, excluding the BaaS expense and partner fraud, this quarter was $45 million, basically on the nose. 4 to 6 quarters from now, should you step down $5 million a quarter in the run rate?

Or is there a reinvestment we should be contemplating that you maybe don't see a true kind of $20 million drop in the run rate over the next 4 to 6 quarters?

Christopher Adams

Yes, that's a great question. So we certainly see that as a pretax number, and there will be potentially reinvestment that can happen at different stages as we evaluate the business and the needs and everything on a go forward.

But from a vendor/contractor and the duplication side of things, we do think right now, just on a go-forward basis, that we will see that run rate over the next 4 to 6.

So I think the way that you're thinking about it is probably correct.

Andrew Terrell

Just on the Bluevine relationship, I appreciate the color you guys provided. It looks like you can just pull separate deposits back on to replace what you'd be losing here.

Do you just have a net estimate on how losing this partner impacts your net revenue or net earnings? And then more broadly, just as you've, I'm sure, sifted through the partner base at the bank, do you feel like there's any risk that we see any more of this action going forward within your current partner base?

Eric Sprink

Yes. Good questions all the way around. So yes, well, first off, let me just complement Valley and Bluevine.

I think it's a good step for both of them. And I know both the CEOs extremely well, and we wish them nothing but the best. And we see this as not a surprising development.

In the maturing fintech ecosystem, et cetera, we all forget X1 was purchased by Robinhood. We forget that Bridget was bought by Upbound. Ownership changes happen.

So I can't sit here and say, hey, this isn't going to happen. Quite the opposite. We've been the direct benefactor of changes of ownership over time. I can tell you that specifically, we're not aware today of any other events, bank charters and/or changes in ownership.

What I can say is, on Bluevine specifically, we have plenty of funding, alternative sources to sweep back on the balance sheet. And our flexibility really plays into this structure.

I think it's often undercredited that having a diversified portfolio allows us to absorb these shocks. The second quarter event, unfortunate, is an example, but even change of ownership events.

I will say this to you, though, we haven't done the final calculations on a net income impact, but we're looking at everything. We've given you the numbers around it.

So I think what we're also trying to work out internally, Andrew, is our cost to oversee this. So there's puts and takes when you get to the net income number. And as there are some internal costs, what can actually be reduced due to losing the relationship or what will be reinvested. But thank you for the question.

Andrew Terrell

Now that 2Q is a little less fresh, you've had time to digest. Any shift in how you're thinking about managing partner credit?

I'm specifically asking around the credit enhancement asset versus a harder dollar cash pledge deposit guarantee. Is there any shift in how you're thinking about managing the credit indemnity structure with partners moving forward?

Eric Sprink

Yes. I mean we humbly admit we're a learning organization, and we're going to take all events and continually update and monitor how we look at partners, how we look at the indemnification ability capability.

You heard us talk about that, including liquidity, cash pledge accounts, how we look at the partners' earnings.

Some interesting context is the diversity of revenue streams of our partners has become an interesting topic internally to try to figure out the redundancy of the partners versus monoline fintechs.

We've started looking at a lot of different metrics to say, how can we improve this? And going forward, we want to be better today than we were yesterday, and we want to be better in a year than we are today.

Andrew Terrell

And if I could sneak one more in, just on the CCBX loan balances down a little bit this quarter.

I'm assuming some of that was impacted by -- it sounds like some portfolio mix change happened this quarter, maybe there's some higher quality credit.

Should we expect loan portfolio growth out of CCBX going forward? Or are we in a period of kind of just continued remix?

Eric Sprink

Well, Andrew, thank you for sneaking in our first pro forma question of the day. And I'm going to try my best to successfully not answer it.

Look at average balances versus end of period first off. Average balances were up $300 million. So there's quite a lot of variability in the day-to-day management of the balance sheet.

And I wouldn't read too much into quarter end. We have a robust pipeline. We have new products coming online. We have new partners coming online.

We're always going to be constantly managing the portfolio. And we're still a growth company. We did sell a larger block than normal of on-balance sheet credit cards, and you'll see that in the period-end balances by product mix.

So it's worth calling that out. You could see in one of the graphs we gave you that sold credit cards went up versus on-balance sheet credit cards, working with our partners and their risk tolerances, our risk tolerances played into that a little bit, but we enjoyed great discussions with our partners all quarter about risk appetites.

And it's just going to be a constantly fluctuating thing that Joel today and our new CFO will have to optimize the balance sheet going forward. But I do want to reiterate, we're still a growth company, and we have a robust pipeline, and we're going to manage that appropriately.

Operator

Your next question comes from the line of Tim Switzer with KBW.

Timothy Switzer

Putting the cost saves aside for a second, how has your efficiency improved over the last year or so to onboard partners and service current ones due to things like AI investments and stuff like that?

And the obvious follow-up to that is your ability to continue onboarding new partners and meeting the pipeline out there while still achieving these cost saves and keeping good expense discipline.

Eric Sprink

Yes. I'm going to answer it this way. The AI technology that's being put in place is definitely on the oversight. That was our main target over the last 12 months.

We're using it a lot in compliance, anti-money laundering as well as some other maintenance vehicles as compared to what you asked about AI and onboarding.

So I just want to draw a small distinction there. But I do think Tara Herger, our Chief Operating Officer, who helped us start CCBX many moons ago.

She was our first employee in that group is still running it today, and she has continued to refine her team techniques and I would say, if you look at the pipeline today versus some prior aspirational onboarding comments has been more rationalized for the attempt to be giving consistent earnings to the company.

So you've got a balance of all of that working forward that we think we're managing it appropriately, still a growth organization, keeping and maintaining the appropriate oversight to do it safely and soundly with great controls while getting those growth and income targets we're targeting.

Timothy Switzer

Got it. Okay. For the provision on the specific partner, are you able to maybe break down like what portion of that provision was due to loan growth versus macro changes in your CECL model and then also for like the charge-offs?

Eric Sprink

Yes. So the loan growth was nominal. I mean, 507 versus 511, that's a timing difference, if you ask me personally. We worked with the partner to maintain. So that's just a statement.

So I don't think loan growth played into the provisioning. You're right, we try to be as transparent as possible in our comments. And so that's why there's a little bit of a qualitative comment, my opinion, that the attorneys has allowed me to sneak in there saying, hey, charge-offs were a little bit different than what I was expecting, not materially, but modestly, I think, was the word I used.

But we did call it out to say, hey, we've got our targets inside that's built into the CECL models. But it was really close, but it wasn't better than what we were projecting.

So we want to just make that statement. But with that said, the charge-offs was down pretty dramatically quarter-over-quarter. And year-over-year, it's down from 15% down to 9% or really dramatically different and improving.

And we see the improving comment come through the delinquency status. That's why we added that in there, so you can get some color.

So I think what really, in my mind, and of course, we can dive into this more, but is the macro changes in the rate environment really, when you're talking about the CECL model, drives prepayment speeds and it extended our weighted average life or the wall in our warm method and CECL looks at lifetime and the lifetime extends and the losses extend.

And at this time, we deemed that we can't rely on the partner to reimburse us for those CECL expenses. So that $3.1 million is unfortunate. At the end of second quarter, we said, hey, we got to evaluate it every quarter.

We're going to continue to evaluate it. But we're trying to give you as much color as we can, Tim.

Timothy Switzer

Yes. No, that was really helpful. And I mean, is there any, I guess, aspect of this provision in terms of like roll forward where we should expect maybe smaller than this quarter, but like a modest amount of provision expense each quarter as the CECL model continues to move forward and you originate new loans?

Eric Sprink

Good backdoor to a pro forma question. What I can say is we're going to evaluate it every quarter. We've got our models in place. The partner is extremely cooperative.

We are talking with the investment bankers. We are talking with the potential investors on that side. We are talking with our accounting group, our finance group, our credit group.

All parties are at the table looking at everything constantly, and then the Board then comes over top and ask 50 new questions. So it's all eyes on everything right now.

We understand the sensitivity of it. We, of course, didn't want to have to come with this messaging today, but we think it's appropriate in the context to be as transparent as possible.

Timothy Switzer

And one last one for me. You talked about this a little bit, the higher core NIM and CCBX. It's moved higher the last 2 quarters, obviously, a pretty good trend to see with the gross loan yields also moving down.

How much of that is like loan mix and program mix versus better pricing on your lending programs?

Eric Sprink

Yes. I think there's a couple of slides that, in the earnings release, I think you're referencing, or I'm referencing in my comments, Page 29, Page 35, that kind of gives you the loan mix; cash secured installment loans go up $50 million during the quarter.

Well, cash secured consumer loans does get a lower gross yield than unsecured cards. So the mix did change, and it's just worth calling that out. Our cash secured installment loans now is about $265 million.

But if you look at Page 29 in the earnings release, you can see that the net yield to us was pretty moderated and controlled. But the gross yield is a mix question as well as an underwriting question. And we're seeing that our partners are very disciplined.

Our credit groups and their credit groups and management and management teams are talking all the time. We're tightening underwriting, and they are in a cooperative nature.

And we're starting to see that manifest itself in lower charge-offs, lower delinquencies, and lower gross yields while we're maintaining our yield.

Operator

Your last question comes from the line of Joe Yanchunis with Raymond James.

Joseph Yanchunis

Most of them might have been answered, but I do have a few I'd like to bounce off you. So you talked about the significant investments Coastal has made in technology compliance, or risk infrastructure.

And as you work through the expense initiatives, how much additional revenue do you think the existing platform can support without requiring another meaningful increase in spending?

Eric Sprink

No. I mean this is where Chris will talk a little bit about the inflection point. So I'll give it to Chris.

Christopher Adams

Thank you, Joe. Yes, we have been building the platform here for a number of years. And so we feel really good about where it is.

The technology build, the vendors that we put in place, the opportunity now to go back and turn off older vendors and look for duplication. I don't think there's any limitation in terms of how we've constructed it on our growth side of things.

I mean we remain a growth company. This is about building a leaner organization on a go-forward basis to really focus on efficiency and profitability.

So we feel really good about our growth path right now.

Joseph Yanchunis

And then just bouncing around here. So you talked about the big step-up in credit cards sold this quarter. Can you talk about how the market's appetite for that paper has changed over the past few months or quarters?

Eric Sprink

I'm probably not as close to answering that. Let me just say we were very successful. There was a very large transaction that I was involved with during the quarter that the team shepherded through with one of our partners.

It was larger than normal and very successful. So I'm just going to confirm by saying, hey, the market is liquid. The market is able to absorb this. We, of course, maintain liquidity and capital buffers if there is any type of slowdown.

That's part of our resiliency of our plan, but I haven't heard any pushback from the teams that the secondary markets aren't operating efficiently.

Joseph Yanchunis

And then last one for me. So obviously, part of the BaaS model involves potentially some of your largest and most successful partners eventually becoming acquisition targets or pursuing their own charters.

This isn't a new phenomenon, but has your thinking around the optimal size of any individual relationship changed?

Eric Sprink

No. I mean, over the last 3-plus years, we have skewed towards much larger companies. So that's evident in the list that we have there.

We still look for, in totality, the highest probability of success both for the partners' TAM as well as the bank's risk-adjusted return.

We do weigh, hey, when you sign some of these larger ones, the on-balance-sheet margin might be slightly less, but you get the off-balance-sheet upside potential with maybe a moderate-sized or middle-sized company that has slightly better operating margins, but maybe not as much upside because of their preexisting distribution network or what have you.

I think it's about diversification, Joe. When again, it comes back down to it, we believe in product diversification, program diversification, industry diversification and managing concentrations behind that.

So any type of Bluevine announcement shouldn't scare the markets too much because they know we're disciplined on how we manage the portfolios behind the scenes.

Operator

That concludes our Q&A session. I will now turn the call back over to Eric for closing remarks.

Eric Sprink

Well, thank you all, and thank you to the team members on the call and to all of our friends and supporters that woke up at 5:00 to hear this call today.

So we do appreciate the team and everything that they have done, and we are very excited about our future, and we appreciate everybody's attention and coming on the call today, and we'll talk to you all soon. Have a wonderful day.

Operator

This concludes our call for today. Thank you all for joining. You may now disconnect. Everyone, have a great day.