Australian retail banking is one of the most concentrated in the developed world: four institutions hold roughly three quarters of the market.
In 2017 I joined the board of Xinja, which became a licensed, mobile-only challenger bank built to give Australians a genuine alternative. An alternative that tens of thousands of Australians showed us they wanted, depositing hundreds of millions of dollars in a matter of months.
Xinja secured one of the country's first new banking licences in decades, during a rare window in time where that was possible. Building a bank isn't easy, but our passionate team comprised courageous visionaries, experienced finance veterans, and talented technologists. We rattled the cage of a powerful industry, captured the imagination of ordinary Australians, and made a difference.
So choosing to hand back the licence in the end was hard. But it was the right decision, and we devoted ourselves to caring for our customers in the same way we had throughout: communicating openly and returning every deposit in full.
That journey, from vision to licence to launch to voluntary wind-down, taught me more about markets, capital, regulation and leadership than any decade of success ever has.
I have spent thirty years in financial services, much of it building investment businesses inside one of the world's most successful institutions, Macquarie Group. Xinja was my first real education in what the world looks like from the other side of the moat. Here are some of the most valuable lessons I learned.
Market share is not the same as loyalty
Every challenger business has to answer the skeptic's question: does anyone actually want this? At Xinja the answer arrived faster than our own plans did. At its peak the bank held approximately A$480 to 500 million in deposits across around 47,000 accounts. Most was deposited in a matter of months, at acquisition costs just a fraction of the eye-watering amounts the incumbents spent.
In concentrated industries, incumbency and customer satisfaction are routinely mistaken for each other. Market share tells you who customers chose when they had four options that looked the same. It tells you nothing about what they will do the moment a genuinely different option appears. Concentration is not evidence of contentment. Sometimes it just means customers never had a choice.
Australians did not need to be convinced they wanted a better bank. They were waiting for one.
Trust transfers fast when you know your customer
The standing objection to challenger banks has always been that nobody will trust their savings to a startup. Forty-seven thousand Australians disproved that in under two years.
What I learned is that trust does not attach to a company's age or scale as much as the industry assumes. It attaches to signals: a banking licence, a government deposit guarantee, transparent communication, and a product that talks to a specific customer need and behaves the way it promises to.
Deliver the signals and the trust follows, at internet speed. That is exhilarating when you're the challenger. It should be sobering if you are the incumbent, because the same mechanism runs in reverse.
For a bank, capital is the product not startup fuel
Most startups treat capital as fuel: you raise it, you burn it, you build with it. A bank is different in a way I understood intellectually for decades and only truly learned at Xinja. A bank's capital is not what funds the machine. It is the machine. Prudential capital determines how many deposits you may hold, how fast you may grow, and whether you may operate at all.
This inverts the normal startup playbook. Growth, the thing every startup is built to maximise, consumes regulatory capital at a ferocious rate. Every customer who loves your product and deposits their savings increases your capital requirement before you have earned a dollar from them. Xinja's success at attracting deposits was, in prudential terms, a cost. We were, in a very real sense, punished by our own product-market fit until the lending side could catch up.
If you are building anything in a prudentially regulated industry, tattoo this somewhere visible: your growth plan is your capital plan.
The rulebook is written for the incumbents' scale
I want to say this carefully, because it is an observation about how the rules are designed, not a complaint about the people who enforce them.
Prudential rules exist to protect depositors, and they should be demanding. But rules calibrated for institutions with trillion-dollar balance sheets, decades of retained earnings, and permanent access to capital markets have a side effect nobody intended: they define the minimum size and speed a challenger must achieve to survive, and the hurdle is high. The capital required to operate a small bank safely is, proportionally, an enormous hurdle for a new one. And the timeline on which that capital must arrive does not flex for pandemics or market cycles.
None of that is anyone's fault. None of it excuses a challenger that fails to adequately plan for it. But founders and boards entering regulated industries should understand that the physics of the field were set with much heavier objects in mind. You are playing the same game as the giants, under rules sized for giants, without a giant's mass.
That's not unfair. It is simply the terrain, and pretending otherwise is a fatal error.
Timing is a balance-sheet item
Xinja's final capital raising efforts ran into 2020, with the arrival of COVID and the sharpest global capital markets seizure in more than a decade. Committed international capital slowed dramatically as the world locked down. For an ordinary startup, a delayed funding round means a painful bridge and some down-round dilution. For a bank, where capital is a regulatory precondition of operating, a delayed round is existential.
I had managed billions of dollars, including large-scale distressed assets through the global financial crisis, and thought I understood market timing risk. What Xinja taught me is that timing risk compounds with regulatory risk: the market decides when capital is available, the rulebook decides when you must have it, and neither consults the other. Any business plan whose survival depends on those two clocks staying synchronised is carrying more risk than its spreadsheet shows.
Governance at startup speed is its own discipline
I arrived at Xinja having sat on more than fifteen boards at Macquarie, together overseeing tens of billions of dollars in assets across fourteen international jurisdictions. Dealing with complexity was business as usual, and we did so with relative ease. So I thought I knew governance. Startup bank governance turned out to be a different sport played on the same field.
At an institution, the board's job is mostly to steward a machine that already works. At a challenger, the machine is being assembled mid-flight, by a small team, at maximum velocity, under full prudential supervision from day one. The information flows, committee structures and second lines of defense that large institutions take for granted must be built while they are already in use.
The honest lesson is that the governance load of a regulated startup is far closer to that of a major institution than its headcount will ever suggest. Boards must resource for it accordingly, earlier than feels natural or even possible sometimes given the balance sheet.
Knowing when to stop is not a failure of nerve
By late 2020 Xinja's capital pathway and the growth trajectory could no longer be reconciled, and the board faced the decision every founder dreads and few discuss. The gambler's option was available: keep operating, keep raising, hope the market turned. Plenty of institutions in history have chosen hope, and their depositors have paid a high price.
Xinja was built as a bank for the people. Their financial wellbeing was always our mission. When we could no longer deliver, we acted with integrity and chose to stop. At the board's initiative, Xinja closed voluntarily, while every customer obligation could still be met in full.
Stopping early looks like weakness until you compare it with the alternative. An orderly wind-up where you can still meet your obligations is a successful failure.
How you leave is the last product you ship
The wind-down became, in effect, Xinja's final product, and we built it with more care than anything else we shipped. In December 2020 the bank began Australia's first voluntary return of deposits by a licensed bank. Seven weeks later it was done: all of the remaining deposits, totalling close to A$252 million, was returned in full to around 37,900 customers, no depositor losing a cent, no call on the government guarantee, no contagion anywhere in the system.
At our lowest point, we held our heads high, and our hands remained steady as we seamlessly executed this complex process through a COVID lockdown, Christmas and New Year. The former chair of APRA, Australia's banking regulator, later described the outcome publicly as a "successful failure."
I am proud of many things I have built across thirty years. The way Xinja closed, despite the personal challenges involved, sits near the top of the list. Ventures fail; that is the price the economy pays for innovation. Whether the failure lands on customers is a choice made by the people in charge, usually years earlier, in decisions about culture and priorities that seem small at the time.
Build the ending into the company from the beginning, because you will ship it exactly once, and that's the product that will be remembered.
Conduct compounds
In 2021, APRA commenced an administrative review of Xinja's final capital raisings, which ran for more than four years. It was conducted and decided by the regulator itself, with no court or independent tribunal involved, and it ended in findings I have disputed publicly. I settled the matter for health reasons. I will not relitigate it here; my full personal account is linked.
The professional lesson I can offer belongs to anyone who takes a board seat or an executive role in a venture that might fail, which is to say every venture. Your protection in the aftermath should be the ledger you built along the way: what the customers experienced, what the record shows you did, how you conducted yourself when it was hardest. Titles wash off. Outcomes do not.
The years after a high-profile failure are their own chapter, and one you don't always get to write yourself. The aftermath was hard on me, and on colleagues I still respect deeply. Despite this, the line every one of us at Xinja can claim as our legacy is the one that matters most: every customer got every dollar back.
Lasting impact matters more than venture success
Xinja the venture did not survive. But our impact has. When I look at Australian banking now, I see the challenger banks' fingerprints everywhere: customer-centric app experiences, instant onboarding, savings features, and the seriousness with which the majors now treat digital experience. Some challengers were acquired, some folded, one or two endure. Regardless, the customers won.
That is the honest economics of disruption, and I wish more founders were told it plainly. Most challengers do not defeat the giants. But they do change them, at considerable cost to themselves, with the surplus landing with customers and, eventually, with the incumbents agile enough to learn. Would-be founders, if your definition of success is being one of the ones left standing, don't start. The risks are too high. But if moving an industry counts, then even our failures compound into something, and that's a worthy definition of success.
In closing, even though the personal cost has been high, I don't regret my time at Xinja. I have always been passionate about working to ensure Australia remains one of the best financial services environments in the world. That requires vision, courage and determination. And I will take the lessons learned forward with me.
As then APRA Chair Wayne Byres said, Xinja was a "successful failure." We dared to believe and we dared to try. We showed customers what they deserved and we delivered, and our challenge has ultimately made the industry better.