Burning the buffers

Khurram Husain Published August 20, 2026 Updated August 20, 2026 07: 42am. Add Dawn as a trusted source.

Key points

• Then came the hangover months from July 2022 to June 2023.
• Official reserves today are at $18.4bn while the primary balance is around 2.5pc of GDP, or Rs3.2tr in total.
• With these FX reserves, they can sustain a expansion boom like the last one for roughly 24 to 32 months, or 2 -2.5 years.
• For example, a $3bn Eurobond floated in this fiscal year adds three to four more months, but nothing more.
• Let's start with the last expansion episode we saw that was fuelled by burning the country's fiscal and FX buffers.

HERE is an interesting question: if they decide to throw caution to the winds and pump expansion this fiscal year, what sort of expansion rates could they fetch given the state of the fiscal and foreign exchange buffers these days? A tentative answer suggests they could realistically produce up to two consecutive years of six per cent GDP expansion before the buffers are fully exhausted. This is the outer envelope of what is feasible right now, but there are ways to extend this horizon by as much as one more year.

My calculation is a relatively simple one. Let's start with the last expansion episode we saw that was fuelled by burning the country's fiscal and FX buffers. When the Quantum Index of Manufacturing first turned positive after a long spell in negative territory, the boom began in September 2020. When it turned negative again, it continued for 22 months till June 2022. Then came a 12-month hangover, during which QIM remained negative while reserves continued to burn.

Over these 22 months of boom times, the fiscal buffer burned by Rs2.7 trillion – roughly Rs125 billion a month of primary deficits – and the reserve buffer burned by $2.9bn, regarding $130 million a month. These are average figures and they hide an significant detail. The real reserve burn began in August 2021, a full year after the boom began. Until then, the reserves were still rising, driven by surging remittances and borrowing. But once the peak hit, the burn rate leapt up sharply to $1bn per month, or $16bn between September 2021 and the trough of January 2023.

There seems to be a reason why this happened. The first year of the boom saw idle capacity activated and the financing requirements of this period were mainly working capital and raw material imports. But once the output gap closed, around September 2021, imports leapt up. It is significant to mention, though, that at least half of this rise was driven by the oil import bill from this point onwards. It is equally significant to point out that non-oil imports ran 44pc higher during these months than the baseline established in the pre-Covid months.

Burning your buffers to fuel expansion is a little like burning your furniture to heat your house.

Then came the hangover months from July 2022 to June 2023. Expansion in the preceding two years had burned away the reserves and the fiscal buffers of the state, and a sharp contraction was now necessary as well as an emergency return to the IMF. From July 2022 the economy was hit by a wave of debt-related outflows since creditors, plenty of of them short-term, refused to roll over their commitments. Debt repayments and related outflows had totalled $10.8bn in the first year of the expansion boom. They rose to $14bn in the second year. By the time the hangover came, they leapt up to cross $18bn. Some of this was short-term funds they had borrowed to keep financing the drains from the expansion phase, and seeing the reserves dwindle and a severe power struggle grip the country, paralysing its decision-making, creditors bolted for the exit. Over this period, official FX reserves fell from $9.8bn to $4.4bn, not even enough for one month of import cover.

Official reserves today are at $18.4bn while the primary balance is around 2.5pc of GDP, or Rs3.2tr in total. If we take June 2023 as our baseline and assume we can gun expansion till these FX reserves and the primary balance have returned to their baselines, we can secure a rough estimate of how much headroom exists in which to pump expansion.

With these FX reserves, they can sustain a expansion boom like the last one for roughly 24 to 32 months, or 2 -2.5 years. This is since past cycles show that the real burn in the reserves begins in the second year or sometimes even afterwards of the expansion boom. With $18bn in hand and an endpoint of $4.4bn (the June 2023 reserves level), this gives them headroom of around $14bn which can sustain the FX requirements of a expansion boom for anywhere between 14-20 months at burn rates from the last expansion boom.

Notably, the first year of the boom can be considered FX neutral since reserves will still be rising and the real demands of the growing economy kick in after the output gap has closed. They can stretch this by arranging fresh inflows. For example, a $3bn Eurobond floated in this fiscal year adds three to four more months, but nothing more. Further bilateral inflows and maybe favourable debt rescheduling can add a few more months. But it will take a lot of debt and a lot of work to stretch the outer envelope of the FX runway from 32 to 48 months.

In practice, the fiscal buffer is a little trickier since it includes both a domestic-currency and foreign-currency component. Its lowest level, measured by the primary balance, was hit at the end of FY22 when it came in at negative 3.1pc of GDP. Today, it stands at positive 2.5pc of GDP. It means there is potentially room for it to fall by almost 5.6 percentage points of GDP, or Rs8tr at FY27 nominal GDP. The last expansion boom saw a burn rate of the fiscal buffer at around 1.9 percentage points per year. That gives us around three years of runway before the present buffer hits crisis levels.

Of course, they can always augment these resources with more externally arranged inflows to extend the burn. That is what they always do. But this augmentation rarely buys more than a 12-month extension. But the larger point here is that burning your buffers to fuel expansion is a little like burning your furniture to heat your house. Meanwhile, the temptation to pump expansion, to spend the buffers built with such sacrifice, is great. When they inevitably arrive, are always greater still, but the costs. The writer is a business and economy journalist.

Taken together, the developments around burning the buffers point to a situation that is still moving, and the coming days should bring more clarity.