Judo Banks 40% Crash|3 Bad Loans or the Start of an SME Credit Cycle?

A 40% Crash on Three Loans — Or Is It More?

Judo Capital shares fell 40 per cent on Thursday to a historic low of 91 cents, erasing $400 million in market value in a single session. The trigger was a profit downgrade driven by just three bad loans, spread across entirely different sectors and states. That combination — a violent price move explained by a number that sounds small — is the tension that the rest of this script resolves. The bottleneck is in what those three loans reveal about Judo's exposure to Australia's SME credit cycle, not in the loans themselves.

Judo's full-year profit before tax guidance was cut from $180–190 million to $163–169 million. FY27 guidance landed at $210–220 million, roughly 16 per cent below analyst consensus of $255 million. The stock had been buy-rated by 12 of 13 analysts as recently as six weeks ago. CEO Chris Bayliss arranged a hasty analyst call Thursday morning. By the time it ended, shares had fallen 46 per cent intraday to 82 cents.

The core claim from management is that these were borrower-specific failures — three customers who deteriorated after a recent customer-by-customer review, in sectors unconnected to each other. That framing matters because if accurate, the crash is a mispricing. Judo's new guidance still implies 30 per cent year-on-year earnings growth. At the session's lows, the bank was trading at 10.5 times trailing earnings — its cheapest valuation on record.

But the market is pricing something different. NAB dropped 3.4 per cent on the same session. Westpac, ANZ and Commonwealth Bank fell between 1 and 2 per cent. Investors are not reading three bad loans. They are reading a forward signal about what rate hikes do to small business borrowers.

The Provision Timeline Management Didn't Explain

The detail that most investors missed in the briefing is not the downgrade itself — it is the timing. In April 2026, Judo released a trading update that reaffirmed full-year guidance, raised collective provision coverage from 0.89 per cent to 0.94 per cent of loans, and flagged sensitivity in agriculture and transport. Twelve of thirteen analysts covering the stock called it a buying opportunity. The share price closed up 1.4 per cent that day.

Six weeks later, three loans have turned. The cost of risk has risen from $40.1 million in the first half to a full-year range of $116–122 million. That means second-half cost of risk is now running at $76–82 million — nearly double the first half on an annualised basis. Bayliss told analysts these deteriorations emerged "subsequent to" the April review, driven by "recent, borrower-specific developments."

The buried assumption in that framing is that the April review was comprehensive and that what followed was unforeseeable. But the review itself cited economic sensitivity in transport and agriculture. Those are rate-sensitive sectors. The RBA rate-hike probability for August now sits at 20 per cent, with 48 per cent pricing a hike by year-end — as of this session, after Australia's unemployment rate fell to 4.4 per cent with 40,300 jobs added in May, well above forecasts. The employment print, according to Bell Potter, "pretty much cements" the August discussion.

If the rate cycle is turning in this direction, the April collective provision — 0.94 per cent of a $13.8 billion book — is not a buffer. It is a floor that could move. Judo's 90-days-past-due and impaired ratio is now guided to reach 3 per cent of gross loans by 30 June, up from 2.65 per cent at March. That 35 basis point move in a single quarter, driven by three loans, is what the market is extrapolating forward.

The hidden assumption the consensus required — that the April review was a ceiling on impairments, not a floor — has now been broken. What remains unresolved is whether the break is structural.

What NAB's 3.4% Fall Actually Signals

The contagion read is the most consequential layer of today's session, and the pool is divided on it. The Australian ran a headline "Sting in the tail: Judo's warning for the big banks." AFR's close wrap noted NAB as the biggest big-four laggard, framing it as an SME exposure read. Both interpretations treat Judo's bad loans as a forward indicator for the broader business credit cycle.

The counter-evidence is structural. Judo is not a traditional bank. Its entire lending book — $14.4 billion at 24 June — is concentrated in SME relationship banking. Its customer-to-banker ratio is deliberately low, meaning each relationship manager holds fewer clients than a typical bank would. That concentration is what makes Judo's model differentiated in good times and brittle in bad ones. NAB's exposure to SME lending is one segment within a diversified book that includes housing, institutional and international operations.

The mechanism connecting Judo to NAB is sentiment, not credit book correlation. Investors holding NAB sold because Judo's shock raised the question of what other banks' SME books might carry. That is a legitimate question — it is not the same as confirming NAB has the same credit quality issue. The distinction matters for anyone deciding whether NAB's 3.4 per cent fall was a rational repricing or an oversell driven by association.

What the pool does not carry is evidence of NAB's own 90-DPD movements or any quantified SME stress in the big four's recent disclosures. That absence means the contagion thesis is an inference, not a confirmed read. It could prove correct if the SME credit cycle is indeed turning. It is not yet evidenced in the big four's own data.

The Variable That Resolves This by August

Two checkpoints will determine whether today's crash was the right read or a catastrophic overreaction. The first arrives in five days: Judo's 30 June year-end, at which the 90-DPD plus impaired ratio either stabilises at the guided 3 per cent or migrates above it. If three bad loans were genuinely the end of it, that ratio holds. If the macro environment is producing a fourth or fifth exposure in the final week, the April-to-June trajectory becomes a run rate, not a one-off.

The second checkpoint is Judo's FY26 full-year result in August. That will disclose the full cost-of-risk outcome, the final NIM (guided above 3.2 per cent for the second half, which is better than prior guidance), and crucially, whether FY27's $210–220 million PBT guidance is maintained. If Judo's 30 per cent earnings growth trajectory survives August, then a 10.5x earnings multiple for a profitable, well-capitalised SME bank with $14.6 billion in loans is, by any historical comparison, extremely cheap.

There is one genuine counter-evidence point the pool surfaces: Judo's NIM is improving, rising to above 3.2 per cent — actually outperforming the prior 3.15 per cent guidance — and its cost-to-income ratio is tracking below the first half's 48.5 per cent. The operating engine is not broken. The credit engine produced a shock. Whether those two things can be separated is what August reveals. The read survives this counter-evidence as long as the three bad loans do not compound into further disclosures before then.

For the holder who has now absorbed the 40 per cent drawdown: the action criterion is the 30 June loan book disclosure — if impaired ratio holds at 3 per cent, the FY27 earnings case remains intact and the valuation is historically compelling. For the watcher who did not hold: the entry trigger is the same date plus confirmation that no additional provisions emerge in the final week. A fourth bad loan before 30 June changes the thesis. Three contained does not.