From Shoebox to Bank: Michelle Bagnall and the story of Bank First.

Michelle Bagnall is the CEO of Bank First, a mutual bank with 92,000 members and $4.7 billion in assets that exists to serve nurses and teachers in Australia. In this episode she tells Dexter the origin story of a bank that started with $480 in a shoebox, explains why nobody at BankFirst gets paid a bonus, and makes the case that mutuals were the original fintech startups.

Bank First. Forty-eight teachers and a shoebox.

I have spent almost two decades placing executives into fintech companies. I have watched founders raise millions on pitch decks that promised to give people access to fair, transparent, lending. Every investor presentation I saw between 2015 and 2022 had the same slide: banks are broken, we are the fix, we are changing the world. Some genuinely were, others were all about changing their world.

Last week I sat down with Michelle Bagnall, CEO of BankFirst, and she told me a story that made me feel passionate about Fintech again. In 1972, a group of 48 teachers in Victoria decided they were being ignored by the banks. So they pooled $10 each, put $480 in a shoebox, and started lending to each other. Their first loan was $250 to a single female teacher who needed help getting into a house.

Watch on Youtube

Watch on Spotify

Apple Podcasts

Today, BankFirst has 92,000 members and $4.7 billion in assets. Nobody at the bank gets paid a bonus. Not the CEO, not the board, not the branch staff. Every dollar of profit goes back to members or into the communities the bank serves.

About Michelle Bagnall

Michelle Bagnall is the CEO of BankFirst with more than 20 years in banking spanning NRMA, RBS in the UK, and multiple listed and mutual institutions. She grew up in Southwest Sydney, went back to university as a mature age student at 23, and describes herself as an accidental banker who found her way home in the mutual sector.

The Peer to Peer model and Fintech.

When SocietyOne launched in 2012 as Australia’s first peer-to-peer lending platform, the pitch was simple: cut out the banks, connect borrowers and lenders directly, give everyone a better deal. It raised $46 million in venture capital. It was backed by James Packer, News Corporation, Kerry Stokes and Westpac.

RateSetter quickly followed and Peer to Peer lending was the buzzword in Fintech right up until 2018 when NeoBanks and BNPL started to grab the attention of investors.

In 2021, MoneyMe acquired SocietyOne for $94 million. The peer-to-peer model was quietly shelved.

Zopa launched in the UK in 2005 as the world’s first P2P lending platform. By 2020 it had applied for a banking licence. By 2021 it had closed its P2P arm entirely and pivoted to become a regulated digital bank offering savings accounts and credit cards.

LendingClub, founded in 2006 in the US, acquired Radius Bank for US$185 million and shut down its retail P2P platform the same year. RateSetter, which had roughly £1 billion in loans on its books, was acquired by Metro Bank. The entire P2P lending sector converged on the same conclusion within a 12-month window.

Ravi Anand, managing director of ThinCats, a UK alternative lender that also shut its doors to personal investors, summarised it bluntly. P2P lending, he said, was a "moment in time response" to the global financial crisis. The model worked when trust in banks was at its lowest. Once trust recovered, the structural economics of crowdfunded lending could not compete with deposit-funded balance sheets.

Every major peer-to-peer lending platform in the world eventually decided it wanted to be a bank. And while they were spending a decade and hundreds of millions of dollars figuring that out, a model that already did exactly what P2P promised was sitting in a shoebox in Victoria. Community-based lending. Aligned incentives. Lower costs. Running since 1972.

50 years of patient capital.

The Australian mutual banking sector holds $178.4 billion in assets as of 2025, according to KPMG’s Mutuals Industry Review. That is 2.7% of all authorised deposit-taking institution assets in the country. It is not large relative to the big four, who control about 70% of the market. But it is not supposed to be.

Mutuals are not a niche. They are one of the oldest forms of organised lending on the planet. The credit union model traces back to 19th century Germany. Friedrich Raiffeisen built lending cooperatives in rural communities where formal banks refused to operate. The principle was identical to what 48 teachers did in Victoria a century later: people who share a common bond pool their capital and lend to each other on terms that a distant institution would never offer. It is the same principle that SocietyOne launched on. The difference is that the mutual version was never designed to generate an exit.

The structural difference is not scale. It is incentives. A mutual has no shareholders. There are no quarterly earnings calls. There is no investor pressure to prioritise short-term margin over long-term member outcomes. Capital allocation decisions are measured in decades, not funding rounds.

In the venture-backed fintech world, the dynamics are inverted. Australian startups raised $5.48 billion across 390 deals in 2025, a 31% increase on the prior year. Fintech was the second-highest funded sector at $868 million. But 46% of investors surveyed by Cut Through Venture saw at least one portfolio company shut down during 2025, and 77% reported layoffs across their portfolios. The capital is flowing, but the structural pressures of the VC model create a set of incentives that are fundamentally incompatible with building patient, community-first financial infrastructure. Grow fast, demonstrate unit economics, exit within seven to ten years. That timeline does not suit a bank built to last 100.

Michelle Bagnall has a phrase for the BankFirst model. It is not "not for profit." It is "for profit, for purpose." The distinction matters. The bank generates returns. It just does not distribute them to external shareholders. The more successful BankFirst becomes, the more it invests back into nurses and teachers. That feedback loop has been compounding for over 50 years.

Why understanding credit risk is essential for investors.

The advantage of the mutual model is not ideological. It is structural.

Consider credit risk. BankFirst lends heavily to nurses and teachers. These are professions with high rates of casualised employment. Shift work, agency contracts, no guaranteed hours. The standard credit risk frameworks used by the big four were not designed for this workforce. They were designed for full-time employees with fixed salaries and predictable income streams.

BankFirst can design products specifically for casualised workers because it has no pressure to maximise net interest margin across a diversified portfolio. It can take a longer view on credit performance. It can underwrite people the major banks would reject on automated scorecard alone. And it can do this not because it is more charitable, but because its capital structure allows it.

Then there is the no-bonus model. In listed banking, variable compensation drives behaviour. Risk appetite, product design, sales culture, hiring priorities: all of it flows downstream from incentive structures tied to short-term financial targets. BankFirst removed that lever entirely. Nobody gets a bonus. The CEO included. Michelle Bagnall does not frame this as sacrifice. She frames it as alignment. When nobody is being paid to optimise for the quarter, decisions default to what is right for the member over the long term. You can build a faster app. You can lower origination costs. You can use AI to automate credit decisions. But you cannot engineer away the structural tension between investor return timelines and the long-term, patient relationship that community lending requires. The mutuals solved that problem in 1972. They did it with governance, not technology.

Inventing a model that already existed.

The fintech industry has spent 15 years and billions of dollars trying to rebuild something that already existed. The narrative was that banks were broken and technology would fix them. What nobody factored in was that a different ownership structure, not a different technology stack, was the real disruption.

The mutual sector in Australia is consolidating. Fewer, larger mutuals are emerging through mergers. Bank Australia absorbed Qudos Bank and became just the fourth mutual to reach $20 billion in assets. The sector grew assets by 5.8% in 2025. This is not a sector in decline. It is a sector figuring out how to apply the structural advantages of member ownership at economies of scale.

The talent signal is hard to ignore. I place executives for a living. The candidates I speak to who have spent 10 or 15 years inside listed banks are asking different questions than they were five years ago. They are less interested in total comp and more interested in what happens after they leave. They want to point at something they built that still exists, that still serves the people it was designed for. That is not idealism. It is the natural career trajectory of someone who has already earned enough to care about meaning. Mutuals offer that. The difference now is that highly talented Fintech operators are making the move.

Tier One People is working exclusively with Bank First to hire their first ever Chief Product Officer. If you’ve built consumer lending products in a high growth Fintech reach out to me.