<p>Your columnist has a disconcerting feeling that something repulsively odorous is gathering around the corner. Wars, expensive money and governments living beyond their means are hardly reassuring cohorts. Add an artificial-intelligence boom financed by borrowing and the global economy begins to look too dependent on everything going right. Consider government debt. America’s gross federal debt exceeded USD 40 trillion in October. The narrower measure, debt held by the public, was projected by the Congressional Budget Office to reach 101% of GDP this year. France’s public debt reached 119% of GDP in June. Britain’s net public debt passed GBP 3 trillion in September. Japan’s ten-year government bond yield is now around 2.8%, a world away from its years near zero. These countries have different financial strengths, but all face a very awkward adjustment as cheap borrowing matures and must be replaced with dearer money.</p><p>Higher yields do not simply announce that governments might default. They also reflect inflation, central-bank policy and competition for savings. Nevertheless, mounting interest bills leave less for public services and investment. Politics makes the sums harder as voters generally favour fiscal discipline until they discover that the spending to be cut affects them. Alongside this sits the enormous punt on AI. 2025 estimates place worldwide data-centre funding needs through 2030 at USD 6.7 trillion, including USD 5.2 trillion for AI. This is projected spending, not money already borrowed.</p><p>Whilst the technology may justify the excitement, the financial timetable is less forgiving. Companies need time to reorganise work and turn experiments into profits. Even widespread adoption does not guarantee that providers can charge enough to recover their costs. Interest payments, however, do not wait for the learning curve. Nor is every participant equally exposed. Some leading chipmakers generate ample cash. Elsewhere, borrowing and long-term payment commitments are growing. Financing structures link technology companies, private-credit funds, insurers and banks. A loan can disappear from one company’s balance sheet without disappearing from the financial system. Banks therefore need not lend directly to a struggling data centre to suffer. They may finance the funds or vehicles behind it. If expected revenues disappoint, losses and forced asset sales could spread well beyond technology shares. Regulators have already warned about rising borrowing and concentrated AI exposures, but nobody seems to listen.</p><p>There is, however, a twist. An AI bust could push investors into government bonds, lowering yields and easing some borrowing costs. A simultaneous debt and technology crisis is therefore not inevitable. The uglier possibility is a deep downturn that damages tax receipts while persistent inflation or distrust of governments prevents borrowing costs from falling sufficiently. America’s central role would make any such shock international. Global portfolios own American assets; businesses depend on dollar funding. Investors seeking cash could sell holdings elsewhere, squeeze credit and unsettle currencies. The dollar’s reserve status can cushion America while transmitting financial pressure abroad. Indian companies could feel it through weaker export orders, volatile capital flows and more expensive finance. The consequences would reach investment, household spending and employment. Governments already stretched could find themselves supporting weaker economies while struggling to protect services. This is a risk scenario, not a forecast that the financial system is about to collapse.</p><p>For Indian businesses, the fallout could be quite practical. Weaker orders and tighter credit would encourage manufacturers to postpone investment and trim output. Hotels and airlines could face fewer business trips, conferences and discretionary holidays. Uncertain jobs and incomes, coupled with harder-to-get loans, could persuade households to delay buying cars and washing machines. Purchases would not necessarily be abandoned, merely postponed. But when enough customers decide to wait, businesses cannot escape the bill. Warnings from regulators and analysts are clearly visible. Your columnist’s worry is that much of the business world may still treat them as someone else’s problem.</p><p>So what to do? Boards should test cash flows against falling sales, delayed payments and restricted borrowing. They should secure funding early, examine vulnerable customers and suppliers and decide which investments can be postponed. A contingency plan costs less before everyone needs one</p>
<p>Your columnist has a disconcerting feeling that something repulsively odorous is gathering around the corner. Wars, expensive money and governments living beyond their means are hardly reassuring cohorts. Add an artificial-intelligence boom financed by borrowing and the global economy begins to look too dependent on everything going right. Consider government debt. America’s gross federal debt exceeded USD 40 trillion in October. The narrower measure, debt held by the public, was projected by the Congressional Budget Office to reach 101% of GDP this year. France’s public debt reached 119% of GDP in June. Britain’s net public debt passed GBP 3 trillion in September. Japan’s ten-year government bond yield is now around 2.8%, a world away from its years near zero. These countries have different financial strengths, but all face a very awkward adjustment as cheap borrowing matures and must be replaced with dearer money.</p><p>Higher yields do not simply announce that governments might default. They also reflect inflation, central-bank policy and competition for savings. Nevertheless, mounting interest bills leave less for public services and investment. Politics makes the sums harder as voters generally favour fiscal discipline until they discover that the spending to be cut affects them. Alongside this sits the enormous punt on AI. 2025 estimates place worldwide data-centre funding needs through 2030 at USD 6.7 trillion, including USD 5.2 trillion for AI. This is projected spending, not money already borrowed.</p><p>Whilst the technology may justify the excitement, the financial timetable is less forgiving. Companies need time to reorganise work and turn experiments into profits. Even widespread adoption does not guarantee that providers can charge enough to recover their costs. Interest payments, however, do not wait for the learning curve. Nor is every participant equally exposed. Some leading chipmakers generate ample cash. Elsewhere, borrowing and long-term payment commitments are growing. Financing structures link technology companies, private-credit funds, insurers and banks. A loan can disappear from one company’s balance sheet without disappearing from the financial system. Banks therefore need not lend directly to a struggling data centre to suffer. They may finance the funds or vehicles behind it. If expected revenues disappoint, losses and forced asset sales could spread well beyond technology shares. Regulators have already warned about rising borrowing and concentrated AI exposures, but nobody seems to listen.</p><p>There is, however, a twist. An AI bust could push investors into government bonds, lowering yields and easing some borrowing costs. A simultaneous debt and technology crisis is therefore not inevitable. The uglier possibility is a deep downturn that damages tax receipts while persistent inflation or distrust of governments prevents borrowing costs from falling sufficiently. America’s central role would make any such shock international. Global portfolios own American assets; businesses depend on dollar funding. Investors seeking cash could sell holdings elsewhere, squeeze credit and unsettle currencies. The dollar’s reserve status can cushion America while transmitting financial pressure abroad. Indian companies could feel it through weaker export orders, volatile capital flows and more expensive finance. The consequences would reach investment, household spending and employment. Governments already stretched could find themselves supporting weaker economies while struggling to protect services. This is a risk scenario, not a forecast that the financial system is about to collapse.</p><p>For Indian businesses, the fallout could be quite practical. Weaker orders and tighter credit would encourage manufacturers to postpone investment and trim output. Hotels and airlines could face fewer business trips, conferences and discretionary holidays. Uncertain jobs and incomes, coupled with harder-to-get loans, could persuade households to delay buying cars and washing machines. Purchases would not necessarily be abandoned, merely postponed. But when enough customers decide to wait, businesses cannot escape the bill. Warnings from regulators and analysts are clearly visible. Your columnist’s worry is that much of the business world may still treat them as someone else’s problem.</p><p>So what to do? Boards should test cash flows against falling sales, delayed payments and restricted borrowing. They should secure funding early, examine vulnerable customers and suppliers and decide which investments can be postponed. A contingency plan costs less before everyone needs one</p>
