Budget season almost always finds a simple lever. This year’s candidate is the Cash Isa cap: lower the allowance, the theory goes, and you coax money out of cash and into productive assets.
It sounds tidy in a speech, but it is messier in the plumbing. The UK mortgage market is not funded in the abstract. Building societies must raise at least half their lending from members’ retail deposits. Cash Isa inflows are the most visible, most seasonal and – crucially – stickiest of those deposits.
Kink that pipe and you don’t get a smooth equity rotation; you get a funding gap where mutual lenders live.
A smaller cash Isa allowance may not change how much households save, but it does change where that saving lands. For building societies – bound by the 50% retail-funding rule – less tax-sheltered inflow means less headroom to lend or higher marginal funding costs.
Using recent Bank of England flows, HMRC participation data and a transparent model, the central case suggests a £10k cap trims society funding by ~£2.5bn, while a £5k cap cuts ~£5.8bn. On conservative multipliers, that equates to ~24–31k and ~56–71k fewer mortgages a year respectively.
Societies can and do use wholesale markets, but only at the margin and at a cost that tightens pricing
If policy wants more equity ownership, fine – there are cleaner routes than starving mutuals of the cheapest, stickiest cash they can lawfully use.
Building societies are not banks with identical balance-sheet options. By statute, at least 50% of their funding must come from members’ retail deposits. That constraint isn’t a footnote; it is the business model.
Societies can and do use wholesale markets, but only at the margin and at a cost that tightens pricing. Cash Isas matter because they are tax-sheltered, high-trust and “stay put” longer than sight deposits – exactly what you want to support long-dated mortgage books.
Sector data show mutuals hold ~47% of all UK Cash Isa balances (about £191bn at March 2025) and around 29% of the mortgage market. Drain the pool and you expose the constraint. The flows aren’t subtle. Bank of England “Money & Credit” data show April 2025 brought +£14.0bn into Cash Isas (seasonally adjusted), up from +£11.7bn in April 2024.
Even outside “Isa season”, 2025 saw steady monthly inflows of ~£2–4bn. On the lending side, gross mortgage originations normalised in late summer at roughly £24–25bn a month, following a stamp-duty-driven March spike (£39.9bn) and April air-pocket (£16.9bn).
Retention is the key parameter: what proportion of the ‘lost Isa’ stays with the same society as a non-Isa deposit?
Short-term correlations between Isa flows and monthly lending are modest – rates and approvals drive volumes – but that misses the point. Funding is a constraint variable: you notice it most when it binds.
A cap does not necessarily make households save less. It moves a slice of saving out of the tax-sheltered channel societies rely on and into taxable deposits or other providers.
Retention is the key parameter: what proportion of the “lost Isa” stays with the same society as a non-Isa deposit? If retention is high, the impact softens; if retention is low, the constraint sharpens. The sector’s risk is not outflows per se; it is the loss of a privileged, sticky inflow aligned with their statutory model.
Last year, savers put £69.5bn into Cash Isas across about 9.94 million accounts – roughly £7,000 each on average. With almost half (≈47%) of all Cash Isa balances held by building societies, a lower cap means a portion of the money that would ordinarily arrive through that channel doesn’t.
Not all of it vanishes: many savers would still park the cash with the same society as an ordinary (taxable) deposit. We assume around 60% stays and 40% leaks elsewhere. That “leakage” is what bites.
Phil Wickenden: What advisers really want from a model portfolio
Why? Because societies must fund at least half of their lending from retail deposits. Lose £1 of that headroom and, at the margin, you can restrain about £2 of lending (today’s balance sheets suggest ~£2.55). Using those yardsticks – and a £210k average mortgage – the results are straightforward.
A £10,000 cap implies a ~£2.5bn shortfall, equating to ~£5.1–£6.5bn less lending capacity, or roughly 24,000–31,000 fewer mortgages. A £5,000 cap raises the shortfall to ~£5.8bn, or ~£11.6–£14.8bn less lending capacity – 56,000–71,000 fewer mortgages.
The effect depends heavily on retention. If 80% of “lost Isa” cash remains with the same society as ordinary deposits, the impact roughly halves (a £5k cap would mean ~28,000–35,000 fewer loans). If only 30% stays, the effect roughly doubles (~97,000–124,000). This is why the widely cited “~60,000 fewer mortgages” figure under a £5k cap sits neatly inside the central range.
Advisers: what to do now
- Pipeline discipline: Assume longer decisioning and tighter appetite at mutuals if the cap lands; secure rate locks early, especially for FTBs and remortgagers in society-heavy geographies.
- Lender mix: If a case relies on a society’s niche (self-employed, complex income, higher LTV), line up a bank fallback and document criteria trade-offs.
- Pricing optics: Small funding-cost moves (10–30bp) matter; help clients focus on total-cost-over-term, not headline rates alone.
- Cash strategy: A lower cap is not a guaranteed equities on-ramp; use laddered taxable deposits judiciously and explain the tax drag.
- Consumer Duty trail: Record how funding conditions shape product availability; explain constraints in plain English.
Be fair. High retention blunts the effect: if most savers keep cash at the same society outside the Isa wrapper, the funding shortfall shrinks. Banks can use wholesale funding more readily than mutuals, so the impact is uneven.
Not all societies sit near the 50% threshold; headroom varies. And month-to-month Isa inflows do not “cause” monthly lending prints – approvals and affordability dominate the short run. This is a structural story about capacity, not a day-trader signal.
If ministers kink the cheapest funding pipe for mutuals, the market won’t rebalance in theory; it will rebalance in price, product and patience
If the aim is more equity participation, there are cleaner tools. Streamline the Stocks & Shares Isa user journey. Permit targeted, rules-based nudges under the Advice/Guidance reforms to help cash-heavy clients overcome inertia. Use evidence-led comms around time-in-market rather than blunt fiscal steering.
Above all, don’t starve mutuals of the very funding the law insists they use. Plumbing, not poetry.
If ministers kink the cheapest funding pipe for mutuals, the market won’t rebalance in theory; it will rebalance in price, product and patience. Advisers should plan for that reality now – and watch April’s flows like hawks.
In Mighty Boosh terms (and I really should update my cultural references!): this crimp doesn’t slap; it stifles.
Phil Wickenden is founder of Ad Lucem
