What Bankers Won’t Tell You About Lending Standards (and Why Reforms Keep Falling Short)

In the summer of 2007, a mortgage underwriter in suburban Phoenix sat staring at a file that was, for all practical purposes, empty. No pay stubs. No W-2s. Just a credit score of 620, a stated income of $14,000 a month, and a supervisor hovering nearby, gently reminding her about the quarterly volume bonus. She approved it. So did thousands of others, all over the country, until the whole machine seized up and took the global economy with it.

We all know how that story ends. But here’s the uncomfortable part: the reforms that followed were supposed to fix this mess once and for all. And they haven’t. It’s not that the reforms did nothing — it’s that they targeted the wrong layer of the problem.

What banking sector reforms keep ignoring about rational lending standards is hiding in plain sight, and as long as we miss it, we’re just setting ourselves up for the same movie. Again. Maybe with different actors.

The Problem: We Keep Building Forts After the War

After 2008, regulators swung the hammer with a vengeance. Capital requirements were jacked up. Stress tests were introduced. Banks were forced to hold more liquidity and map out their exposures in dizzying detail. Mortgage lending rules, particularly the ability-to-repay requirement under the Dodd-Frank Act, were rewritten to make sure no borrower could be granted credit without documented proof that they could actually pay it back.

Good. All of it, good.

But let’s be honest about what happened next. The pendulum didn’t settle in the middle. It just swung violently to the other side.

Lending to prime borrowers? That’s fine, the banks will happily do that all day. Lending to first-time homebuyers, small business owners, or anyone with a self-employed income that isn’t printed neatly on a W-2? Suddenly, that became nearly impossible. Credit scores that had been the beginning of a conversation became the entire conversation. Loan officers went from being judged on how many loans they originated to being judged on how few of them could ever possibly go bad. Which, by the way, is a wonderful way to guarantee that you only lend to people who don’t need the money in the first place.

Here’s the kicker. If you run a bank, the safest loan you can possibly make is the one you don’t approve at all. That’s the uncomfortable truth that drives the modern lending paradox. The system now rewards inaction. And inaction is not the same as prudence — it’s just a different kind of failure.

Remember what happened to those credit-crunched small businesses in 2010? The ones that couldn’t get a loan even with solid revenue, healthy margins, and a decade of consistent operations. They wound up either selling equity, tapping personal credit cards, or closing their doors entirely. The 2008 crisis has been widely studied, but the slow suffocation of legitimate borrowers in its aftermath has received comparatively little attention.

Reform without rational lending standards isn’t reform. It’s just exile — you’ve banished the bad loans, along with a whole cohort of perfectly acceptable ones.

What “Rational Lending Standards” Actually Mean

Let’s define our terms before we go further.

Banking sector reforms talk a lot about capital buffers, liquidity coverage ratios, and leverage ratios. Watch any conference panel on regulation and you’ll hear those phrases thrown around within the first 60 seconds. Those things matter, and their importance can’t be overstated. But they measure the bank’s ability to survive losses, not its ability to make sound credit decisions in the first place.

Rational lending standards are the rules, processes, and incentives that govern the individual credit decision itself. A loan is approved or declined based on real evidence of the borrower’s capacity and willingness to repay, with transparent and consistent criteria applied across applications. No corner-cutting in the pursuit of volume. No over-caution in the pursuit of safety. The borrower’s true ability to pay is assessed honestly, the loan structure is designed to be sustainable, and the whole process is documented in a way that can be scrutinized later.

That sounds so simple. Almost insultingly simple.

And yet, almost nobody in the reform world talks about it.

Here is where our story gets counterintuitive. The best performing loan portfolios in the world were not built on the loosest standards, nor on the strictest. They were built by institutions that figured out how to make rational decisions at scale — figuring out what data actually predicts repayment, rather than what data is easy to collect, and calibrating their appetite not to the regulator’s demands or the CEO’s quarterly targets, but to the underlying realities of their market.

That is the hypothesis, and so much of the evidence supports it.

The Cultural Problem That Reforms Can’t Solve

Walk into a commercial bank’s lending floor in almost any developed country and you’ll see the same structural tension playing out. The front-line lenders are incentivized by origination volume. The credit risk department is incentivized by portfolio performance metrics. These two forces are locked in a permanent tug-of-war that neither one will ever fully win.

Banking sector reforms act like this tension doesn’t exist. They just layer on more rules, assuming that every lender will follow them faithfully. But for decades now, the culture of banking has operated on a principle that can be summed up in a single line from the film Margin Call: “There are three ways to make a living in this business: be first, be smarter, or cheat.” Whenever leverage and compensation rewards short-term thinking, the culture bends around the rulebook. That isn’t cynicism. It’s realistically what happens when a loan officer’s annual bonus depends on one thing while a compliance manual says another.

Loans are made by humans, which means they’re made by creatures of motive. If reform doesn’t change the motives, it’s just rearranging the traps.

Passive structures, careful oversight, explicit gatekeepers — these have all been deployed in various combinations across the last two decades. And they’ve helped around the edges. But the core dilemma remains: human judgment, when aligned with bad incentives, is remarkably good at finding a way around any rule you can write.

What the Evidence Actually Shows

Now, I want to be fair to the regulators too. Some things have genuinely improved.

The adoption of IFRS 9 and CECL, which requires banks to recognize expected credit losses earlier rather than waiting for a default, was a meaningful leap forward. It forces lenders to think about what might go wrong, not just what has already happened. The Basel III framework also pushed banks to hold much higher quality capital, making the system as a whole more resilient to shocks. These are real, verifiable improvements. Nobody reasonable would argue for returning to the pre-2008 regulatory swamp.

Its also true that non-performing loan ratios across the developed world have been extraordinarily low in the past decade. But before we break out the champagne, let’s dig a little deeper. The low NPL ratios might have less to do with excellent lending standards and more to do with banks simply refusing to lend to anyone who wasn’t holding a golden ticket.

It’s like bragging about the health of your diet because you swore off all ice cream while also deciding that, actually, you’d rather stop eating all other foods too. You’re not healthier, you’re just starving yourself — and malnutrition has a long lag time before it shows up in the metrics.

The small business effect

Consider small business lending specifically. The data has consistently shown that small business owners, especially those from minority communities, have struggled to access credit post-crisis. The 2010 reforms created a genuine contraction in community bank lending. Community banks, the traditional lenders to small enterprises, faced disproportionately high compliance costs relative to their size. Many of them consolidated or exited lending lines entirely.

Was that the intended consequence of banking sector reforms?

No. But it was a predictable one.

And it points, again, to the blind spot: reforms were designed around the stability of the financial system, not around the quality of access to credit. The two are related but not identical. A system can be perfectly stable while also failing in its fundamental economic function, which is channeling capital to productive ventures. When credit allocation becomes purely defensive, you haven’t reformed finance, you’ve just made it timid.

What Rational Lending Standards Look Like in Practice

So, after all that hand-wringing, what do we actually do about it? What does a rational lending framework look like when it’s implemented properly, rather than just talked about in polite regulatory circles?

Let’s unpack this.

The anatomy of a good credit decision

A rational lending standard starts from a simple premise: credit decisions should be evidence-based, consistent, and documented. That means they should be:

• Clear definitions of acceptable debt-to-income ratios that genuinely reflect regional real estate taxes, insurance costs, and living expenses — not a single national blanket number
• Verification requirements that are proportional to the risk and complexity of the loan, rather than one-size-fits-all documentation rules that burden the self-employed
• An explicit weighting of character and history, including rent payment history, utility bills, and longer trend data, for borrowers with insufficient traditional credit files
• Portfolio-level diversification limits that force bankers to recognize concentration risk before it materializes, rather than after
• Compensation structures for loan originators that reward long-term portfolio performance, not just the number of loans closed this month
• A meaningful appeals process that lets a good loan application get a second chance if initial screening was too rigid

Notice something? None of those are exotic. None of them require a bank to throw caution to the wind. What they require is judgment. And judgment takes time, takes skill, and takes institutional commitment, which is why it’s mostly avoided by modern banking which runs on efficiency at scale.

Cash flow over collateral

Another shift that rational lending demands: emphasizing cash flow over collateral. This isn’t just a policy preference — it’s economic necessity, particularly when asset prices go through volatile cycles. In the years before 2008, lenders made the stupid assumption that a house’s value would bail them out. Falling prices turned that assumption into a catastrophe. In the years since, the opposite error still persists: banks lend against assets that are already liquid and valuable, which means the people who own those assets never really needed the bank in the first place.

A rational standard asks: will this business generate enough cash to service this debt over its full life? The most challenging, and hence most necessary, loan decisions to make rationally involve borrowers where collateral is thin but the cash flow story is compelling. That’s precisely where banking sector reforms have been silent.

The Oversight Question — What Changes Should We Make

At this point, one might reasonably ask: isn’t this the banker’s job, not the regulator’s?

And yes, a bank can implement all the rational lending standards in the world without being forced to. Some do. But most are caught between the same structural pressures that existed in 2006: shareholders demanding better returns, competitors underwriting aggressively to gain market share, and employees earning compensation based on annual metrics.

Banking sector reform that focuses only on macro-prudential measures will always be a step behind. What we need is reform that also looks inside the credit committee. To be blunt about it: supervisory examinations should start scrutinizing the quality of the approval process itself, not just the outcomes in the portfolio. Examiners should be asking questions like:

Was this borrower’s repayment capacity clearly established with real data or just a checklist exercise? Were exceptions properly escalated and documented? Did the loan officer have any conflict of interest in pushing this loan forward?

An entire generation of banking reforms has been devoted to questions of capital and liquidity. What if the next generation spent as much time on underwriting process quality and credit culture?

The Role of Technology: Promise and Peril

It wouldn’t be a modern article on lending standards without acknowledging the elephant in the room: machine learning and automated underwriting.

Much of the hope for rational lending now centers on technology. And yes, algorithmic credit scoring can remove human bias in certain contexts. It can process enormous volumes of non-traditional data — cash flow patterns from bank accounts, rent payment histories, even utility bills — to assess borrowers that traditional models deem invisible.

But there’s an irony here that deserves attention. The same automation that promises rational standards can also replicate the errors of the past at unprecedented scale. If the algorithm was trained on historical loan data that includes prohibited discrimination or informal and undocumented lending practices, it will learn that exactly the same discrimination as its human predecessors. The code doesn’t reason. It pattern-matches.

Rational lending, at its best, is a hybrid. Human judgment for anything complex or nonstandard. Algorithms for speed and consistency in the standardized world. Standards that treat models as tools, not as oracles.

What This Means for the Banking Industry

Let me leave you with a scenario to think about.

It’s 2030. A tailwind — maybe something that looks like widespread adaptation to new energy systems, or perhaps a demographic shift in a major economy — delivers a wave of entirely new lending demand. Tens of millions of new credit applications, arriving from business owners and households that have never been served by mainstream banking before.

Does the system have the ability to grant those loans rationally?

Thinking about this — I mean really thinking about it — is uncomfortable. Because the answer is probably no. The reforms of the post-2008 era made the banking system safer in a narrow, mechanical sense. More capital, more liquidity, more stress testing. But they did not build the muscle for making complex, judgment-based credit decisions with consistency and genuine evaluation. They built the opposite: an over-reliance on fortress balance sheets and blanket refusals.

If the next wave of opportunity arrives and the banking sector hasn’t relearned how to underwrite rationally — how to take measured risk based on evidence rather than blindly chasing it or shutting the window entirely — it will be the financial equivalent of arriving at a marathon with heavy armor and no legs.

The good news is that this isn’t inevitable.

Lending culture is not fixed. It’s built from individual decisions made every day under specific incentives. Change the incentives, change the training, and you change the outcomes. Bankers need permission to be rational again: to evaluate risk rather than avoid it, to know their customer’s actual business rather than just their balance sheet, and to make loans that serve the real economy rather than a spreadsheet.

Reform isn’t done. It never is.

But if we keep talking only about capital ratios and bank stress tests while ignoring the fundamental question of why and how loans are approved in the first place, we’ll get exactly what we’ve always gotten: a banking system that’s enormously well defended against the last war, and completely unprepared for the next.

The borrower in Phoenix, the one with the empty file, she was a symptom. Not the cause.

The cause was a system that lost its grip on rational standards, abandoned judgment in favor of volume, and imagined that rules written in hindsight could ever replace wisdom applied in advance.

Let’s not repeat that mistake. Either of them.

The next time a regulator bravely announces another round of banking sector reforms, ask the question. Not about their capital assumptions.

Ask them what they’re doing about the lending standards themselves.