The 2025 Budget Speech Cancellation and What It Means for Your Rand Savings

In February 2025, South Africa's Finance Minister withdrew the budget speech hours before delivery, a first in the country's democratic history. The episode was political, but it was also a signal about the structural pressures on the rand that rand-denominated savers cannot afford to ignore.

A budget speech had never been pulled in the democratic era until February 2025. It was pulled with hours to spare.

I watched it happen from my desk in George, refreshing the news feed like everyone else in the country. The thing that struck me was not the drama of it. It was how ordinary the underlying problem is once you strip the theatre away. A government wanted more money than its coalition would let it take. That is the whole story. It is also a story that ends, eventually, at the value of the rand sitting in your bank account.

What actually happened in February 2025

On the morning of 19 February 2025 the Finance Minister was due to deliver the national budget. The centrepiece was a proposed rise in VAT from its long standing 15%. The increase did not have the numbers inside the Government of National Unity, the parties could not agree, and the speech was withdrawn before it was read. The budget was eventually postponed to May and passed only after a compromise.

This was not a filing error. A budget is the one document where a government states, in rands, how much it will take from you and how much it will spend on your behalf. Pulling it at the last hour was an admission that the coalition could not agree on the most basic question a state ever answers.

The instinct afterwards was to file the episode under politics and move on. I would file it somewhere else. The reason a VAT rise was on the table at all is that the sums underneath the budget have stopped adding up quietly.

The fiscal position that forced it

South Africa's gross public debt has roughly tripled since 2008. By the end of March 2025 it had reached 77% of GDP, and the 2026 Budget Review expects it to peak near 78.9% in the current fiscal year before easing. Those are large numbers, but the number that decides your future is a smaller one. In the 2024/25 year, debt service costs, the plain interest on money already borrowed, absorbed 21.3% of all government revenue. More than one rand in five that the state collects now goes straight back out to bondholders before a single school, clinic or road is funded.

The tax base carrying that load is narrow. Unemployment above thirty percent means a large share of working age adults pay no income tax at all, so the weight falls on a thin layer of formal employees and profitable companies. When the money still does not stretch, a government has four moves and no more. It can raise taxes, which is what the withdrawn VAT increase was. It can cut spending, which costs votes. It can borrow more, which grows the interest bill you just read about. Or it can let the real value of its debt melt away through inflation, by expanding the money supply so that yesterday's borrowings are repaid in cheaper rands.

That fourth move is the one savers should watch, because it is the one that never sends you a bill.

The rescue nobody voted for

South Africa has already reached for that fourth tool. Twice.

In March 2020 the Reserve Bank stepped into a frozen bond market and, in its own words, created money to buy assets. It bought roughly R11.4 billion of government securities in April alone, more in May and June, and by October 2020 held about R38.8 billion of government bonds it had not held before. The Bank was adamant this was not quantitative easing. The mechanics said otherwise.

Then came the quieter one. The Gold and Foreign Exchange Contingency Reserve Account records the paper gains on the country's foreign reserves, and two decades of rand weakness had swollen it from minus R28 billion in 2003 to more than R500 billion by 2024. In 2024 the state agreed to draw R150 billion of that down to Treasury across three years, funded not by selling a single dollar of reserves but by the Reserve Bank creating new interest bearing liabilities. Economists who study the plumbing call the outcome de facto quantitative easing, whatever the label on the tin. By October 2025 broad money was growing at 7.5% a year, with R22.4 billion of a single month's expansion coming from fresh claims on the government sector. A revaluation gain that size is not a trading profit. It is twenty years of currency slide, written down neatly by the state's own accountants. I set out the full mechanics of this in what is actually happening to your savings.

What fiscal stress does to the rand

A currency does not weaken at random. It weakens because global capital, watching a country's books, decides it wants to hold less of what that country issues. South Africa's rand went from about R14 to the dollar at the end of 2019 to R18.70 by 2024, a slide I trace in full in what the rand's long decline means for Bitcoin.

The February 2025 withdrawal produced immediate volatility, and episodes like it usually do. When the path to a stable debt load looks unclear, the currency tends to soften. When it looks manageable, the currency can rally, as it did across 2025 for its best year since 2009, back near R16.50 by mid 2026. Give that rally its due. It was real. It was also driven mostly by a weaker US dollar and a credit rating upgrade rather than by anything that fixed the debt arithmetic underneath.

For a saver the consequence is simple and it does not depend on any single year. Money held in rands loses ground two ways at once. Domestic inflation eats the purchasing power at home while the exchange rate eats it abroad. Both losses are invisible on a month to month statement. You only notice them when you try to buy something priced in dollars, or when you sit down and add up a decade.

The exit is narrowing at the same time

Here is the part that gives the budget story its edge. The obvious answer to rand risk is to move some savings offshore, and the state is quietly making that harder even as it invites you to do it.

On paper the door has opened. The 2026 Budget lifted the single discretionary allowance from R1 million to R2 million a year, on top of the R10 million foreign capital allowance that still needs a tax clearance PIN from SARS. So a compliant individual can move R12 million a year, which sounds generous. Read the draft Capital Flow Management Regulations published on 17 April 2026, meant to replace the 1961 exchange control rules, and a different picture appears. The framework now treats foreign currency, gold and crypto assets alike, hangs much of its power on a threshold the Minister sets at will, requires residents to declare offshore holdings within thirty days, and reserves for Treasury the power to compel a sale of lawfully held foreign assets back to the state for rands. I go deeper into that shift in Bitcoin and exchange control after the 2026 Budget.

Read those two facts together. The allowance to leave went up. The surveillance of leaving went up further. That is not a contradiction. It is the same instinct that pulled a budget when the coalition would not vote for more tax, expressed in a different register.

Where Bitcoin fits, honestly

Bitcoin's relevance here is narrow and I would rather state it narrowly than oversell it. Its supply is fixed at 21 million coins and no budget speech, no monetary policy committee and no GFECRA agreement can expand it. A rand in a bond is a claim on an institution that is itself under fiscal stress. A bitcoin held in your own keys is a claim on nobody. That difference is the entire point, and it is structural rather than promotional.

What Bitcoin is not is a shelter from every storm. It is volatile, it carries its own regulatory weather, and holding it properly demands real custody. Anyone who tells you it removes risk is selling something. It swaps one category of risk, the slow certainty of currency debasement, for another, sharp price swings on a young monetary asset, and the swap only makes sense over a long horizon with money you have genuinely set aside.

The case, then, is not that Bitcoin is safe. The case is that for a saver whose every rand already lives inside one small emerging market currency, a modest allocation into an asset uncorrelated with South African fiscal outcomes is a rational form of diversification. There is a further wrinkle worth sitting with. The same draft rules that could one day compel the surrender of foreign assets are far easier to enforce against a coin held at an exchange than against one held in a properly structured self custody setup, a distinction I unpack in what South Africa's new crypto regulations mean if you hold Bitcoin.

In practice I do not tell clients how much. That is theirs to decide. My job is the other half of the sentence: you choose the size, I make sure you do not lose it. That means steady rand cost averaging rather than a lump bought on a brave morning, savings plans that start from R1,000 a month, guided self custody from R10,000, and the Vault for larger holdings, a multisig arrangement where no single key can move the coins alone. The budget can be pulled again next year. What sits inside your own keys cannot be pulled by anyone.