Stagflation Part 18: Philippine Banks as Shock Absorber — Record Lending, Record Bonds, Record Hiding Places

  

But the boom cannot continue indefinitely. There are two alternatives. Either the banks continue the credit expansion without restriction and thus cause constantly mounting price increases and an ever-growing orgy of speculation, which, as in all other cases of unlimited inflation, ends in a “crack-up boom” and in a collapse of the money and credit system. Or the banks stop before this point is reached, voluntarily renounce further credit expansion and thus bring about the crisis. The depression follows in both instances—Ludwig von Mises 

In this issue: 

Stagflation Part 18: Philippine Banks as Shock Absorber — Record Lending, Record Bonds, Record Hiding Places

I. Introduction: Three Shocks, One Balance Sheet

II. The Seen: Record Bank Lending, Bailout Edition

III. The Unseen: Financial Wile E. Coyote Keeps Running

IV. Who Funds the Sovereign? Borrowed Money, Held to Maturity

V. The Wall: Corporate Maturities Join the Queue

VI. Whac-a-Mole: Every Fix Has a Forwarding Address

VII. The Bank–Sovereign Doom Loop: Who Rescues the Rescuers?

VIII. Conclusion: When the Coyote Looks Down 

---

 

Stagflation Part 18: Philippine Banks as Shock Absorber — Record Lending, Record Bonds, Record Hiding Places 

How policy relief, sovereign borrowing, and hidden balance-sheet losses turn banks into shock absorbers—and shift the cost of each rescue to the next balance sheet. 

I. Introduction: Three Shocks, One Balance Sheet 

Every shock since 2020 has had the same first responder: the banking system. Every bailout has had the same last bearer: the banking system. In between sits the sovereign, whose paper the banks keep buying. 

Underneath runs the same operating framework: free-lunch politics rewards spending now and sends the bill later. Spending outruns real savings; the gap is financed through domestic and external credit, central-bank accommodation, and the printing press. Monetary aggregates duly celebrate. The underlying scarcity of real resources, meanwhile, receives no such memo. 

Following the Keynesian dogma that spending drives GDP, the political class gets its preferred scorecard: expenditure goes up, and measured output may follow. 

But debt-financed spending is not synonymous with sustainable production. Borrowing can pull expenditure forward; it cannot, by itself, manufacture the real savings and complementary capital needed to sustain production. GDP records the expenditure. The liabilities remain on the balance sheet. 

Emergencies amplify this dynamic. 

A banking system cannot manufacture real savings simply by creating money. It can create deposits and extend credit, generating claims on future income. Those claims then accumulate across the economy, and the banks become warehouses for a growing web of financial promises. 

The magnification began with the pandemic rescue: historic BSP injections of Php 2.3 trillion, an unprecedented combination of interest-rate and reserve-requirement ratio (RRR) cuts, regulatory relief, fiscal deficits, and a USD/PHP regime operating under an implicit Php 59 (soft peg) ceiling. 

The 2022 oil shock arrived on top of that liquidity. 

Now comes the second oil shock of 2026, catalyzed by the Iran/Middle East conflict, but into an economy carrying a very different set of vulnerabilities: record public debt, eight months of twin deficits, weaker GDP growth, a return to 7.2% inflation, sharply lower gross international reserves (GIR), and a bond market beginning to reprice sovereign risk. 

The rescue capacity that absorbed the earlier shocks has itself become part of the balance sheet. 

This is the operating environment behind EO 110, BSP relief, and the current foreign-exchange regime. The Php 61.75 soft ceiling has given way to a wider operating range of Php 62.465–Php 62.86, before the peso breached Php 63. 

In summary:

• The first rescue created the liquidity.
• The second shock tested it.
• The third arrives after the balance sheets have accumulated the consequences.

The bill now travels through a system whose capacity to absorb it can no longer be taken for granted. And the banks remain central to that transmission. 

Consider EO 110. When a price is held below the level required to clear a market under prevailing conditions, the gap does not vanish. Someone must absorb it. The power producer or transport operator may initially bear the shortfall; bank credit can finance the resulting cash-flow gap; subsidies or other government support may shift part of the burden to the sovereign. Depending on the contractual and institutional arrangements, losses can migrate from one balance sheet to another. 

That is balance-sheet transfer. Bank credit is one of its shadows in the ledger. 

This week's market tape put a price on the accumulated pressures.

Figure 1

In a follow-up to our latest outlook, Whac-a-Mole Economics: The Week the Philippine Financial Markets Cracked, the USD/PHP printed its 27th record low of the year, breaking through 63 to close at 63.06 on October 15, before ending the week slightly lower at 62.985—still a record on a weekly basis. 

The 10-year BVAL yield, after reaching 8.1115% on October 8, eased to 8.0184% at the week's close. That reprieve did little to erase the broader repricing: the yield remained above the May 2026 and November 2022 highs cited in our market comparison. The 20- and 25-year yields closed at multiyear highs of 8.3856% and 8.3846%, respectively. (Figure 1, topmost pane) 

Equities offered a different surface impression. The PSEi skidded toward 5,600, hit by ICTSI's 6.89% plunge on October 14. But gains concentrated in the top 5 heavyweight stocks, led by ICT during the final two sessions, subsequently lifted the index by 1.43% for the week, recovering earlier losses. 

For a short follow-up to last week's article, see this week's account of the splenetic Philippine markets in my Substack note, Another Whac-a-Mole Week: Three Markets, One Signal—Escalating Financial Stress. 

Still, the peso, bonds, and PSEi remain under stress. Three markets, one signal. 

They were not merely reacting to three separate stories. They were repricing claims and risks across three interconnected balance-sheet domains: the government's, the banks', and the conglomerates'. 

• The peso prices the external constraint.
• Equities price claims on private businesses.
• Sovereign yields price the terms on which the government can finance itself—and provide a benchmark that feeds into financing conditions across the economy.
• The banks sit in the middle, holding claims on the other two while depending on their own funding conditions, asset valuations, and borrowers' capacity to pay. 

This installment reads the middle balance sheet. We turn to the August banking data to examine how the transfer manifests itself in lending, asset allocation, and the accumulation of financial claims. 

II. The Seen: Record Bank Lending, Bailout Edition 

The news headline writes itself: total loan portfolio (TLP) hit a record Php 17.783 trillion in August, up 13.21% year-on-year. Production loans accelerated for a third consecutive month, reaching 10.57%. 

Now look at who borrowed more. (Figure 1, middle image) 

• Electricity, gas, and steam: +27.03% YoY, accelerating from 24.36% in July. Share of the loan book: 14.1%.
• Administrative and support services: +20.91%, up from 15.14% in July.
• Financials: +9.04%, up from 5.36%.
• Mining: +54.54%, up from 51.4%.
• Trade: +10.52%, up from 8.76%. 

And those whose borrowing growth slowed: 

• Manufacturing: +6.07%, decelerating from 7.2%—ironically, amid a chip-export boom.
• Real estate: +4.52%, down from 5.55%.
• Consumer loans: +16.19%, down from 17.07%. Credit-card growth was 23.34%, slowing for the fifth consecutive month; auto loans grew 5.83% from 6.97% in July. 

The headline is record lending. The more revealing story is its composition. 

The surge is concentrated in sectors where intervention, regulatory arrangements, or financing pressures has been shaping capital allocation. 

Start with electricity. 

The SMC-AEV-MER and Prime Infra-FGEN consolidations, the suspension of real-property taxes (RPT), the GEA-ALL transition layered on top of FIT-ALL, and an ERC that returns what it takes in one ruling and recovers it in the next. For instance, it orders Meralco to refund consumers in one ruling while approving increases in generation charges in another 

These forms part of the institutional landscape we've examined in previous installments in 2025 (see references below). 

The contradiction is hard to miss: regulate prices, redistribute costs, alter the terms of recovery, and then observe the financing demands generated by the resulting structure. 

The 27.03% lending growth’s mechanism is the Averch-Johnson effect: When allowable returns are tied to a regulated capital base, firms can have an incentive to expand that base beyond what an unregulated profit-maximizing firm would choose. 

The bank finances the enlargement. The regulator helps determine the returns. 

Bluntly, the Averch-Johnson effect needs no ideology:  the regulator's formula pays you to borrow. 

Now consider financials. 

Financial-sector lending accelerated from 5.36% to 9.04%. Essentially, financials are the BSP relief regime finding its own counterparty. The system lends to the system, providing liquidity and bridge financing within the very structure the interventions are intended to stabilize. Credit circulates; the question is how much new productive activity it creates and how much of the existing financial structure it keeps afloat. 

Then comes trade. 

Trade is where the money illusion becomes relevant. Retailers see nominal sales growth against Q2 inflation of 6.5%, interpret the plus signs as stronger demand, and borrow to replenish inventories. But nominal turnover is not real volume, and rising sales values do not necessarily mean consumers are buying more goods. 

The goods arrive. The inventory has to move. 

Yet consumer-credit growth is decelerating. Household purchasing power is under pressure. If inventories were accumulated on the assumption that nominal sales growth represented stronger real demand, retailers could find themselves financing goods that turn over more slowly than expected. 

That is how a credit expansion becomes an inventory problem, then a cash-flow problem, and eventually a delinquency problem. It is a causal sequence to test, not an automatic consequence of the lending figures. 

The employment data add another contradiction. 

Trade employment increased by 227,000 in August even as trade lending growth reached 10.52%. Construction employment rose by 389,500, while construction lending contracted 11.73% year-on-year. (Figure 1, lowest graph) 

Employment, credit, and output need not move together in the short run. But when lending expands rapidly without a corresponding improvement in real activity—or contracts while employment rises—we need to look beneath the headline indicators to understand what is financing production, what is sustaining existing operations, and where adjustment may be accumulating. 

Record lending is not record value-added. 

Our Part 14 already documented the broader disconnect: Q2 GDP growth of 2.3%, first-half growth of 2.6%, a 13.9% contraction in construction, and a 9.2% decline in gross capital formation. Universal and commercial (U/KB) bank credit grew 10.3% against an economy expanding at roughly 2%. 

The difference was not productive expansion. Credit was financing refinancing, liquidity support, and the cost of keeping an intervention-distorted structure viable. 

Credit expanding while investment contracts is not a boom. It is a financial system extending claims while the real economy's capacity to generate returns deteriorates. The lending figures celebrate expansion; the underlying economic signals reveal its failure to translate into productive growth. 

III. The Unseen: Financial Wile E. Coyote Keeps Running 

Chased by the Road Runner, Wile E. Coyote runs off a cliff and hangs in midair. Everything appears fine until he looks down and realizes there is no ground beneath him. 

The gross NPL ratio is that coyote. 

Figure 2 

It is a fraction: nonperforming loans over total loans. While the denominator sprints, the ratio can remain perfectly composed, even as the numerator climbs. The trick is not that bad loans disappear. It is that the loan book grows faster than they do. 

August's gross NPL ratio held at 3.35%, unchanged from July. Look beneath the calm. 

Gross NPLs reached Php 596 billion, up 8.4% year-on-year and the second-highest level on record, just below May's Php 601 billion. Meanwhile, the total loan portfolio hit a record Php 17.783 trillion, expanding 13.21%—all under relief measures. (Figure 2, topmost window) 

The denominator is running faster than the numerator. 

Then the other financial coyotes. 

• Allowances for credit losses reached a record Php 555 billion, up 6.99% year-on-year and equivalent to 3.12% of loans—roughly 93% of gross NPLs. 
• Distressed assets climbed to Php 1.008 trillion, up 10.66%. (Figure 2, middle diagram) 

The reported NPL ratio looks stable because the loan book is expanding faster than the stock of recognized bad loans. That is the arithmetic. It is not evidence that underlying credit quality is improving. 

The ratio does not tell us how quickly the system is accumulating risk. It tells us how much recognized bad debt sits inside an expanding pool of loans. When the pool grows fast enough, the ratio can hold steady while the underlying burden rises. 

The coyote keeps running. The ground has not returned. 

Now go one level down, where the aggregate cannot hide the pattern. 

The consumer is already slowing 

In Q2 2026 consumer lending, loan growth decelerated across every category, while NPL growth outran loan growth in nearly all of them: (Figure 2, lowest chart) 

• Consumer loans: lending grew 14.42%, against NPL growth of 19.64%.
• Motor vehicle loans: lending grew 6.2%, while NPLs surged 16.6%. Vehicle sales slid by about 11.4% in August YTD; risibly, the weather is the alibi.
• Credit-card receivables: lending grew 25.28%, but NPLs jumped 46.53%—nearly twice as fast.
• Salary loans: NPLs grew 24.56%, even after the maturity extension under Circular 1239. 

The pattern is straightforward: recognized distress is growing faster than the underlying credit categories in which it is accumulating. Credit-card NPLs are expanding at nearly twice the pace of receivables. Motor-vehicle NPLs are rising while sales contract. Salary-loan distress continues despite extended repayment periods. 

Extending maturities can ease immediate payment pressure. It can also push the recognition of distress further down the road. The calendar changes; the obligation does not vanish. 

Real estate: the conspicuous exception 

Real-estate NPLs rose by only 2.74%, despite the known vacancy overhang. 

That is the outlier. While consumer-credit deterioration shows up clearly in the numbers, real-estate distress appears far more contained. Sustained refinancing, relief-enabled reporting, inaccurate data and delayed recognition offer possible explanations. The available figures do not establish which mechanism dominates. 

But the contrast is worth watching. A subdued NPL figure is not the same thing as a healthy underlying asset market, particularly when vacancies and refinancing needs remain part of the picture. 

The sectoral bailout 

Now return to the system-wide ratio. 

The aggregate remains stable while consumer-credit quality deteriorates and lending expands rapidly in electricity, financials, and administrative and support services. These are precisely the sectors in which regulatory arrangements, policy relief, and institutional financing needs shape the demand for credit. 

The point being: asset quality is being held up by the bailout sectors. 

The composition of lending matters because rapid expansion in one part of the loan book can dilute the reported deterioration elsewhere. Remove the denominator growth contributed by the expanding sectors, and that cushioning effect weakens—or the rot gets exposed. 

The aggregate ratio conceals the divergence: consumer distress is accelerating, while policy-sensitive sectors continue to expand the pool of outstanding credit. 

This is why record lending and a stable NPL ratio can coexist with deteriorating underlying economic conditions. The banking system can continue extending claims even as the capacity of households and businesses to service those claims comes under pressure. 

When the measurement rules move 

And the measurement institutions have been moved. 

NPL grace periods under Monetary Board Resolution 296, intragroup credit reforms, the release of previously required capital buffers (BSP Circular No. 1235, Series of 2026), and mark-to-market relief (BSP Memorandum No. 2026-027) alter the terms under which financial stress is recognized, absorbed, or reported. 

A reclassified loan is not a repaid loan. It is a loan given more time before it is called what it is. 

A maturity extension can postpone a missed payment. A reporting accommodation can change when a loan is classified as distressed. A capital-buffer release can create additional room to absorb losses. A mark-to-market waiver can defer the recognition of valuation changes. 

These measures are not identical, but they share a consequence: they can buy time between the emergence of financial pressure and its full recognition in reported indicators. 

And time changes incentives. 

When institutions are under pressure to preserve capital, maintain reported solvency, and avoid recognizing losses, the temptation to postpone bad news grows. Relief can provide breathing room for a genuine recovery. It can also prolong the appearance of stability when the underlying problem remains unresolved. 

Charles Kindleberger understood the darker side of financial distress. In Manias, Panics, and Crashes, he describes the sauve qui peut mentality—the scramble for individual survival that takes hold during crashes and panics. As institutions and individuals attempt to escape mounting losses, swindling, fraud, embezzlement, and other forms of misconduct can become part of the unraveling. 

The connection is institutional, not merely moral. When losses threaten survival and the incentives favor preserving appearances, desperation can turn deferred recognition into deliberate concealment. 

First, the numbers are softened. Then the incentives deteriorate. Eventually, someone has to absorb the loss. 

The coyote keeps running because the denominator is still growing, relief is still buying time, and the accounting has not yet forced the adjustment into view. 

But the balance sheet is accumulating the bill. 

A stable ratio is not a stable system. It may simply be a system that has not yet looked down. 

IV. Who Funds the Sovereign? Borrowed Money, Held to Maturity 

The loan book is half the balance sheet. The other half is paper. Yet another Coyote.

Figure 3 

• Net claims on the central government (NCoCG): Php 6.345 trillion, +16.6% YoY, a seventh consecutive month of double-digit growth, equivalent to roughly a fifth of bank assets. 
• Debt securities at amortized cost, formerly held-to-maturity (HTM): Php 4.493 trillion, +11.5%, a record. Growth accelerated from 3.17% in March to 11.5% in August. (Figure 3, topmost image) 
• Available-for-sale (AFS) securities: growth of 14.19%, with accumulated losses of Php 145.65 billion, the third-highest on record. 

AFS shows the losses recognized in the books. HTM is an accounting choice: hold the bond to maturity, and market-price fluctuations generally do not flow through its carrying value. The price still exists; it simply does not appear in the accounts in the same way. Add a mark-to-market waiver, and more of the sovereign's repricing stays outside reported valuations. 

When a bank's loan portfolio grows 13% while deposits rise 9.44%, the funding gap has to come from somewhere. It came increasingly from borrowing: 

• Bills and bonds payable: Php 2.312 trillion, the second-highest level, +51.84%.
• Interbank loans: Php 484 billion, +20.44%.
• Repos: Php 648 billion, +54.55%, following June's record Php 983 billion. (Figure 3, middle and lowest charts) 

Borrow short, hold long, hope the rollover keeps rolling. 

That is transitional Ponzi finance in the Minskyan sense: a position increasingly dependent on continued refinancing. Here, the position is the sovereign's own paper.

Figure 4

The cushion is thinning from the other side. Cash to deposits fell to 9.24%, tying its all-time low; liquid assets to deposits stood at 47.64%, trending downward. (Figure 4, upper window) 

But what counts as liquid? A substantial share is government paper that may have to be sold into the same market, during the same squeeze, as other holders seek liquidity. A liquid asset is one you can sell without taking a damaging loss—not merely one for which a market exists. PDS data already show the strain: September trading volume was the lowest of 2026 as yields rose. 

The fourth balance sheet: captive savings 

The Commission on Audit recently flagged Php 40 billion in SSS equity-paper losses. That is what is seen. 

Beneath the other three balance sheets sits one whose contributors cannot simply leave. SSS coverage is compulsory, the state shapes the institutional framework, and members have no direct say over individual investment decisions. Contributions reached Php 373.6 billion in 2025, up 14.8%. (Figure 4, lower SSS table)

About Php 572 billion—46% of its Php 1.254 trillion in assets—is invested in government paper. Another parallel with the banks: Php 531.8 billion in bonds, more than 80% of non-current financial assets and up 45% in 2025, sits at amortized cost. The books record carrying value, not what those bonds would fetch in the market today. 

When yields rise, bond prices fall. The carrying value need not follow. Temporary is a forecast, not an accounting category. 

The sovereign is also the borrower, so rising yields pressure government financing and the market value of SSS's bond holdings simultaneously: correlation, not diversification. 

Foreign money can exit, and the peso reflects part of that repricing. Captive savings cannot exit as freely; the consequences emerge through investment returns, fund sustainability, and the capacity to meet future pension obligations. 

The channel into stagflation runs through the paycheck. The compulsory contribution rate rose to 15% as September 2026 CPI reached 7.2%, including 9.0% for the bottom 30%. 

Take-home pay is squeezed to fund a pool that finances the sovereign. Pension disbursements rose 12.3% to Php 237.7 billion; retiree pensions are scheduled to rise 10% annually through 2027, while contribution rates remain frozen. 

In 2025, contributions exceeded benefit payments by Php 68 billion in cash. But that is an ex-post result, not proof of long-term sustainability. There was no forced selling in 2025. 

The trigger is net cash outflow that forces the sale of cost-carried bonds below carrying value. The investment portfolio carries market risk; cash flows carry demographic and economic risk. They can deteriorate together. 

The money supply: credit finds another route

Figure 5

Meanwhile, M3 grew 11.2% in August, its seventh consecutive month of double-digit growth. Deposit substitutes surged 75.9% YoY and have remained above 50% growth since March. M1 & M2 growth also accelerated from 7.4% and 9% in July to 8.6% and 9.8% in August. (Figure 5, upper diagram) 

The banks are funding the state partly by issuing their own paper. The sovereign borrows; banks acquire the claims; financial institutions expand their funding. Credit expansion feeds the sovereign-financing loop. 

This is where the Cantillon effect enters. New money does not reach everyone simultaneously. The first recipients—the sovereign and the sectors receiving the initial financing—gain purchasing power before the resulting price adjustments spread through the economy. Those further down the chain encounter the adjustment after relative prices have already moved. It is no coincidence that the oil-shock CPI surge (third wave of the cycle) has been accompanied by pandemic-era money-supply growth. September CPI hit 7.2%—returning to April 2026 high. (Figure 5, lower image) 

The sovereign gets its financing. Banks accumulate the claims. Captive savers supply funds they cannot freely redirect. Households absorb the pressure through reduced purchasing power and compulsory contributions. 

The same system that finances the sovereign transmits the cost of financing it. 

V. The Wall: Corporate Maturities Join the Queue 

The BSP has already identified the next claimant. Its Financial Stability Report (2025) p.30, warns of a sizeable maturity wall among large conglomerates: approximately Php 1.6 trillion, or 22.7% of conglomerate debt, falls due between 2027 and 2029 . US-dollar debt averages 37.6% of conglomerate debt over the next five years. Refinancing has so far relied on bonds, bank funding, and internal liquidity. The BSP's prescription: “close monitoring.” 

That warning predates the second oil shock. Since then, the peso has broken through its previous soft ceiling, reserves have fallen to a three-year low, and the 10-year BVAL yield has climbed above 8%. The wall did not move. The ground in front of it did. 

The stock market repriced the risk first. On October 15, the PSEi fell 2.2%, with index heavyweights doing the damage. Equity markets price conglomerate balance sheets daily; the maturity schedule brings the reckoning more slowly, then all at once. 

Refinancing is not repayment. It is a fresh claim on loanable funds. Banks can create nominal credit; they cannot manufacture the real savings, labor, materials, and capital goods needed to sustain every project that credit finances or it cannot mint the real savings behind it. 

The arithmetic puts the wall in perspective. The BSP's figures imply roughly Php 7 trillion in conglomerate debt. The Php 1.6 trillion due over 2027–2029 equals about 9% of today's bank loan book. Spread evenly across three years, it amounts to approximately Php 533 billion annually—about a quarter of last year's Php 2.08 trillion increase in bank loans. 

And conglomerates are not refinancing in a vacuum. 

The sovereign is borrowing, with banks' net claims on government at Php 6.345 trillion, up 16.6%, as the 10-year BVAL yield exceeds 8%. 

Utilities and other policy-supported sectors are absorbing credit: electricity lending grew 27%, while financial and administrative lending accelerated. 

Banks themselves are increasingly reliant on bills and bonds payable, interbank loans, and repos. 

Now the conglomerates arrive at the refinancing window with Php 1.6 trillion in maturities due over 2027–2029, 37.6% of their debt in dollars. 

Meanwhile, consumers face slowing loan growth, and MSMEs are already being squeezed out, as discussed in our Magna Carta installment. 

Everyone wants the same pool of savings and liquidity. Not everyone gets access on equal terms. 

Four stagflation channels follow. 

1. Crowding out. Funds are rationed by access, not by productivity. Those with sovereign standing or a conglomerate crest get financed; others face higher rates or a rejected application. Credit that rolls over old projects does not build new capacity. That is how credit can grow 13% while gross capital formation contracts 9.2%. Rolled-over debt keeps yesterday's projects from liquidating. That is not capital allocation. It is capital preservation for the already-allocated. 

2. Gearing. Roll over at 8% what was borrowed at lower coupons, and the interest burden rises even if the principal does not. Rolling over is not deleveraging; it is gearing with extra steps. These groups are simultaneously borrowers, bond issuers and, in many cases, owners of the banks that lend to them. Their exposures are not independent bets. They are tied to the same policy regime and financing conditions. 

3. Dollars. With 37.6% of conglomerate debt denominated in dollars, the maturity wall competes for foreign currency with sovereign external debt service and the oil import bill. All three draw on a reserve stock that has just fallen by $4.85 billion. A wider peso trading range does not create dollars. It reprices their scarcity. 

4. Money. If real savings cannot meet the queue, adjustment must come through some combination of higher rates, asset sales, defaults, project cancellations, or credit rationing. 

But if the political system refuses to accept that adjustment, monetary accommodation becomes the escape route. Section IV has already shown the direction: seven consecutive months of double-digit M3 growth, alongside CPI inflation of 7.2%. When financing sustains claims without expanding productive capacity, the economy gets more nominal demand competing for constrained real resources. That is how the same process can feed both sides of stagflation. 

The BSP calls it a wall. Walls are what you hit when you keep driving. The 2027 maturities begin in a few months. The 2026 data already show the system's preferred response: relief, rollover, and more of the same paper. 

VI. Whac-a-Mole: Every Fix Has a Forwarding Address 

The September–October data are not a new story. They are the old story at new addresses. 

Years of spending ahead of real savings (or the record savings-investment gap) created the underlying imbalances; emergency interventions postponed adjustment by drawing down reserves, bank capital, and fiscal room. The oil shock was the trigger, not the cause. Each fix suppresses pressure in one place, only for it to surface in the next least-controlled price.

Figure 6 

The pressure surfaces in goods, dollars, and yields because the underlying constraint is real savings. What Austrian economist Eugen von Böhm-Bawerk called the subsistence fund or real savings is the stock of goods and resources that sustains workers and projects until production yields output. Credit can create competing claims on that fund; it cannot conjure the real resources themselves. When claims outrun available resources, the adjustment surfaces in prices: consumer goods, foreign exchange, and the price of time—interest rates. 

In each case, the fix moved the problem; none removed it. The hammer shrinks with every swing: defending the peso spent reserves, the peso broke anyway, and the bond market is the next mole. A currency can be managed. A yield curve can only be managed so far. (Figure/Table 6) 

Figure 7

The rate channel shows the same blockage. The BSP cut its policy rate nine times, by a cumulative 225 basis points, from August 2024. Average bank lending rates peaked at 8.234% that month and bottomed at 7.125% in February 2026—a decline of just 110.9 bps, less than half the policy-rate reduction. By July, lending rates had climbed back to 7.883%, up 75.8 bps from the low. (Figure 7, upper graph) 

Banks price loans according to their own funding costs and risk, not simply the BSP's policy corridor. Both are being reshaped by the imbalances already discussed: wholesale funding costs are rising, with bills and bonds payable up 51.8% and repos up 54.6% (August 2026); sovereign yields near 8% raise the opportunity cost of private lending; and deteriorating asset quality requires provisions against potential losses. 

The policy rate is a nudge. The balance sheet is the answer. 

Interest is supposed to coordinate saving and investment. Here, there are two prices: the BSP's policy rate and the rate banks actually charge. The first can be cut by decree. The second must contend with funding costs, credit risk, and the government's demand for financing. 

A supply shock makes the contradiction worse. Higher rates cannot produce more rice or oil. They work through credit conditions instead, tightening financing for consumers and MSMEs while policy-supported borrowers continue to absorb credit. 

That is the asymmetry: the sectors bearing the shock face tighter constraints, while the sectors sustained by intervention retain access to financing. 

Whac-a-mole economics is the diagnosis. The bank is the mallet's table. Each strike lands on the same surface. 

VII. The Bank–Sovereign Doom Loop: Who Rescues the Rescuers? 

Name the loop plainly:

1. The sovereign runs deficits larger than revenue—Php 1.054 trillion in eight months—and needs banks to absorb its paper.
2. Banks hold that paper, fund it short, and park much of it at amortized cost. Regulatory relief can obscure market losses without eliminating them.
3. Rising yields, driven by the oil shock, peso weakness, and global bond repricing, reduce the market value of those holdings. Accounting rules determine how much of the loss is reported, not whether it occurred.
4. Capital relief and forbearance soften the immediate consequences, prolonging the system's dependence on sovereign paper.
5. The sovereign is a major asset of the banking system and cannot reprice without transmitting losses back to its creditors. The BSP is pulled toward the familiar response: suppress, subsidize, transfer, borrow.
6. A weaker peso raises the peso value of external debt—about Php 90 billion in August alone (Part 17)—widening the next deficit and returning the process to step one.

The sovereign and the banks are one balance sheet in two names. The crisis is not simply that either is weak. It is that each can be rescued only at the risk of weakening the other. 

Who rescues the rescuer? 

The last time developing-world sovereigns and their banks hit this wall, the Baker and Brady Plans offered successive responses. The Baker Plan (1985) sought to sustain debtors through structural adjustment, multilateral support, and fresh commercial-bank lending. It failed to resolve the debt overhang. The Brady Plan (1989) shifted toward debt reduction, exchanging commercial bank loans for bonds, some backed by US Treasury collateral, with IMF and World Bank support. The mechanism changed from extending credit to restructuring claims. 

Both approaches depended on coordinated creditors and access to external financing. 

This time the donors have their own queue: record public debt in the advanced economies, Fed hike bets after hot US inflation, and an AI financing trade that imploded this week (Nasdaq −4.5%, the worst week since June; the Semiconductor (SOX) index −11.4%; more than $1T off the top 20 AI and chip names). South Korea’s KOSPI plunged 14.5% this week and 18.3% over seven sessions, leaving it nearly 40% below its June 19 high. Samsung fell 17.7% this week and 43% from its June high, while SK Hynix dropped 24.3% and 57%, respectively. Japan’s Nikkei bushwhacked, plunging 5.4% (Figure 7, lowest chart) 

Bond vigilantes, investors who sell a government's bonds when they distrust its finances, are already repricing here and abroad. 

This matters to the Philippines. The same global repricing that threatens AI financing also competes with sovereign borrowing and corporate refinancing for investor capital. A sustained unwind would tighten credit, weaken collateral, and compete with sovereigns and conglomerates for a shrinking pool of risk-bearing capital. The Philippines' refinancing problem is therefore arriving as the capacity of external markets to absorb risk is itself under pressure. 

The central bank faces two broad paths: let yields, the peso, and asset valuations adjust, forcing losses into the open; or extend liquidity and relief, postponing recognition while deepening dependence on continued accommodation. The first is politically painful. The second feeds the loop. 

The system rewards the second until it breaks. 

One reserve stock, three claimants: the currency, the banks, and the sovereign. 

The textbook version is the Mundell–Fleming trilemma or the Impossible Trinity: a country cannot simultaneously maintain a managed exchange rate, free capital flows, and independent monetary policy. Something has to give. The BSP has sought to manage the peso while domestic monetary conditions lag behind inflation. The exchange rate has absorbed much of the resulting pressure. 

The structural problem is the deepening fiscal dominance. 

When the deficit path remains fixed, the central bank becomes the residual financier. Tight money today can simply mean more money creation tomorrow—the unpleasant monetarist arithmetic of Thomas Sargent and Neil Wallace. Their 1981 analysis explains how persistent fiscal financing needs can ultimately constrain monetary policy and shift the burden into future monetary expansion. 

That closes the loop. Deficits create the need for financing; banks absorb the sovereign's paper; repricing threatens bank balance sheets; relief postpones recognition; and monetary accommodation sustains the financing structure. The cost returns through the economy then the peso, yields, and prices. 

The AI trade adds an external vulnerability: the very global capital markets expected to absorb more borrowing are reassessing the returns on one of their largest investment booms. 

The sovereign borrows from the banks, the banks depend on the sovereign, and the central bank is expected to keep both standing. The rescuers are inside the loop—and the outside rescuer is busy putting out fires of its own. 

VIII. Conclusion: When the Coyote Looks Down 

The banking system has become the shock absorber for a political economy that spends beyond its real savings. Credit expands, sovereign paper accumulates, losses are deferred, and liquidity is consumed to keep the structure running. Each intervention transfers the pressure without removing its cause. 

This is the bank–sovereign doom loop: the sovereign depends on banks to finance its deficits; banks depend on the sovereign's paper to sustain their balance sheets; and the BSP is expected to keep both afloat. The same rescue that postpones adjustment deepens the system's dependence on the next rescue. 

The tipping point comes when the chain breaks—when refinancing fails, sovereign repricing forces losses into bank capital, or pressure on the peso exhausts the room for intervention. These are not independent risks. Each can trigger the others. 

The 2027–2029 corporate maturity wall adds another claimant to a financing system already stretched between the sovereign, policy-supported borrowers, and banks' own funding needs. 

The Coyote does not fall because he looks down. He falls when he realizes there is no ground beneath him. Looking down is the recognition; the fall is the adjustment that can no longer be deferred. 

The balance sheets are still running. The ground beneath them is already giving way. 

---

References: 

Stagflation Part 17: Eight Months of Twin Deficits, Record Debt, and the Bill for Deferred Adjustment, October 4, 2026 

Stagflation Part 14: Q2 GDP at 2.3% and Falling — A GDP Built on Suppression and Debt, August 9, 2026 

USD-PHP at Record Highs: The Three Philippine Fault Lines—Energy Fragility, Fiscal Bailouts, Bank Stress, December 21,2025 

The Oligarchic Bailout Everyone Missed: How the Energy Fragility Now Threatens the Philippine Peso and the Economy, December 7, 2025 

Inside the SMC–Meralco–AEV Energy Deal: Asset Transfers That Mask a Systemic Fragility Loop, November 23, 2025

 

 

Stagflation Part 14: Q2 GDP at 2.3% and Falling — A GDP Built on Suppression and Debt

  

Public choice theory predicts exactly this: concentrated benefits and dispersed costs produce political pressure for expansion. Sovereign credit makes the expansion financially viable. The opacity makes it politically sustainable—Michael Dioguardi

In this issue

Stagflation Part 14: Q2 GDP at 2.3% and Falling — A GDP Built on Suppression and Debt

I. Introduction: The GDP Number Is Not Neutral: When Policy Intervention Becomes the GDP Story

II. From Countercyclical Buffer to Debt Dependence

III. The Credit Boom That Households Aren't Feeling

IV. Net Primary Income: When the External Cushion Starts to Fail

V. Real estate and tourism: the visible cracks behind a still-solid labor market

VI. Electricity's Engineered Strength Versus Transport's Engineered Weakness

VII. Construction: a government-led downward spiral

VIII. Trade: exports without manufacturing depth, and a historic deficit

IX. The external financing loop closes on itself

X. The Two Precarious Trends Beneath the Headline

XI. Confusing Stagflation with an Event Rather Than a Process

XII. Conclusion: GDP Is the Symptom, Not the Diagnosis 

Stagflation Part 14: Q2 GDP at 2.3% and Falling — A GDP Built on Suppression and Debt 

How debt, credit, price suppression and external financing are sustaining growth while weakening adaptive capacity

I. Introduction: The GDP Number Is Not Neutral: When Policy Intervention Becomes the GDP Story 

The Philippine economy grew 2.3% in the second quarter of 2026, down from 2.8% in Q1, bringing first-half growth to 2.6% — the weakest non-pandemic print since Q4 2009. 

The consensus reception treated this as a simple story of "as is, where is": inflation is high, investment is weak, ergo growth slows. What that framing consistently misses is that the 2.3% figure is not a passive reading of an economy left alone. It is the visible residue of a set of policy choices that concentrated benefits on a narrow set of interests while dispersing the costs across consumers, savers, and the fiscal balance sheet — Bastiat's seen and unseen, playing out in real time in the GDP release itself. 

Two suppression mechanisms did the heaviest lifting in keeping the headline number in positive territory at all. 

One. EO 110's emergency interventions in electricity and energy markets, layered on the earlier round of rice and fuel price interventions, suppressed some of the immediate price pass-through directly. 

Two. BSP's regulatory relief cascade — the NPL grace periods, the mark-to-market waiver, the capital reliefs — combined with a soft-peg regime and NDF restrictions, limited the extent to which the exchange rate and bond yields could reprice the oil shock into the real economy. 

These measures did not eliminate the shock; they altered its transmission, absorbing some of its immediate effects and shifting others onto consumers, savers, bank balance sheets, and the fiscal balance sheet. 

The 2.3% print therefore cannot be read as evidence that the underlying economy absorbed the Iran-oil-shock and flood-control-scandal disruptions well. It is the measured outcome after policy intervention had already changed the channels through which those shocks propagated. 

II. From Countercyclical Buffer to Debt Dependence 

The PSA data confirms what our Stagflation Parts 12 and 13 flagged as an emerging pattern: government spending is no longer a countercyclical buffer but a permanent, entrenching share of output.

Figure 1

Government final consumption expenditure grew 8.3% year-on-year in Q2 2026 — decelerating only slightly from 8.7% in Q2 2025 — and its share of real GDP rose to 18.6%, up from 17.5% a year earlier and 16.3% in Q1 2026. 

For the first half, GFCE's constant-price share climbed to 17.4% from 16.8% in 1H 2025—an all time high. (Figure 1, topmost visual) 

The significance is not merely that government consumption is rising, but that an increasing share of economic activity is being allocated through the state rather than through decentralized private demand. And GFCE captures only the direct component: it excludes the broader private-sector activity sustained by government procurement, construction, PPPs, contractors, and supply chains serving government agencies nationwide. The state's effective economic footprint is therefore larger than the GFCE ratio alone suggests—a sign of deepening centralization. 

That expansion in the numerator was financed the way it always is now: through debt. 

The national government's first-half fiscal deficit hit Php 786.8 billion, the largest January–June shortfall on record. Net public borrowing for the half reached Php 1.357 trillion — the second-highest first-half figure on record, trailing only 2021's Php 1.371 trillion, a year defined by pandemic emergency financing. 

The parallel is not comforting: what was once crisis-contingent borrowing has become the ordinary operating mode of the fiscal accounts. National government debt closed June at Php19.07 trillion, and the debt-to-GDP ratio breached 66% — a 22-year high, last seen in the Arroyo-era aftermath of the early-2000s fiscal crisis. (Figure 1, middle graph) 

The nominal debt-growth-versus-GDP-growth gap is the cleaner way to see the mechanism. Nominal debt has grown faster than nominal GDP in every year since 2019; the 1H 2026 gap between nominal debt growth and NGDP/rGDP growth is now at its widest since 2020! 

Public debt is not tracking the economy's expansion — it is outrunning it! (Figure 1, lowest chart) 

Sustained divergence of this kind changes the character of sovereign finance ex ante: as the debt stock grows faster than the income base from which it is ultimately serviced, continued refinancing becomes increasingly central to meeting existing obligations. 

The government may continue to roll over that debt without immediate distress, but the system becomes more exposed to a ‘sudden stop’ in financing or a sharp repricing of risk. In Minskyan terms, that is the sovereign analogue of a shift away from hedge finance and toward a speculative posture — not because default has already occurred, but because continued solvency increasingly depends on the uninterrupted availability of new credit.

III. The Credit Boom That Households Aren't Feeling 

Despite EO 110 subsidies, sector-specific relief, and what the earlier parts of this series have already documented as record consumer and electricity-sector bank lending, household final consumption expenditure grew only 2.8% in Q2 2026 — down from 5.2% a year earlier — while per-capita HFCE growth in real terms slowed to 1.9% for the half, a rate not far from the pandemic-adjacent trough. The deceleration is not evenly spread.

Figure 2

Transport spending within the household basket contracted outright, falling 7.5% year-on-year in Q2, the single largest drag on HFCE growth, a direct product of fare structures that remain administratively restricted even as fuel and operating costs were not similarly controlled for operators. 

Restaurants and hotels (-0.2%) and recreation and culture (-0.8%) also contracted — consumption categories that track discretionary income most closely, and the ones collapsing first. (Figure 2, topmost pane) 

That households are cutting discretionary and mobility spending while credit to the household and electricity sectors keeps expanding at record pace is the seen/unseen split of the credit channel: the lending shows up in bank balance sheets and, through the electricity sector's credit-financed "recovery," in industrial GDP; the squeeze shows up in what households actually do with their own cash flow. 

The aggregate consumption data make that squeeze visible: credit is expanding, but the purchasing power and discretionary spending of households are not keeping pace. (Figure 2, middle chart) 

IV. Net Primary Income: When the External Cushion Starts to Fail 

A quieter but structurally important number in the release: Net Primary Income from the Rest of the World grew just 1.0% year-on-year in Q2 2026, against 31.7% in Q2 2025 — but the more important signal is the deterioration underneath the quarterly number. (Figure 2, lowest image) 

Since its 2023 peak, the growth of the external-income stream has been in a sustained waterfall, with both compensation income and property-income flows seeing their growth rates materially weaken through Q2 2026. This line is the GNI-side counterpart to the remittance-shield thesis developed earlier in this series (Part 7.0): OFW compensation and other primary-income flows have functioned as a standing subsidy that allowed vested domestic interests to defer structural reform. The income stream remains positive, but its growth impulse is rapidly disappearing. 

With that external shield now materially weaker, GNI growth (2.2%) fell below already-weak GDP growth (2.3%), while the external cushion that historically absorbed part of the consequences of domestic policy failures is thinning at precisely the moment domestic demand is weakening. The significance is therefore not that external income has already disappeared, but that a once-reliable source of support is no longer expanding fast enough to offset the deterioration elsewhere in the economy. 

V. Real estate and tourism: the visible cracks behind a still-solid labor market 

Real estate and ownership of dwellings grew only 1.3% in Q2 2026, down sharply from 5.9% a year earlier — the weakest print since Q4 2009 outside the pandemic. 

Accommodation and food service activities similarly decelerated to 1.7% from 6.8%.

Figure 3

Neither figure is disaggregated regionally in the national accounts release, but the Cebu office market offers a live, granular preview of what a real-estate demand air-pocket looks like on the ground: CBRE reported first-half 2026 office demand in Cebu crashed 68.2% year-on-year to 20,200 sq.m., a reversal from 2025's "bull run," with vacancy climbing to 13.9% and expected to reach 18–22% by year-end as AI-driven BPO consolidation and a wave of new supply collide. 

The accommodation-food deceleration is consistent with, and reinforces, the tourism slowdown already noted across Baguio, Boracay, the Hundred Islands, and Eastern Visayas — destinations where softer discretionary household spending (recreation, restaurants and hotels both contracting per the HFCE breakdown above) is now visible in occupancy and footfall. (Figure 3, topmost diagram) 

What makes this genuinely puzzling rather than simply confirmatory is that it sits alongside labor force data that has not (yet) cracked in the same way. 

The dissonance between a resilient headline employment picture and visibly weakening real estate, hospitality, and discretionary consumption sub-sectors is itself a data point: it suggests the labor market is a lagging rather than a leading indicator here, or that "benchmarkism" — embellishing a stable unemployment rate as evidence the economy is fine — risks missing where the stress is actually accumulating. (Figure 3, middle image) 

VI. Electricity's Engineered Strength Versus Transport's Engineered Weakness 

Electricity, steam, water and waste management was the one industry sub-segment that meaningfully accelerated: 4.0% in Q2 2026, up from 0.7% a year earlier, with electricity itself growing 4.5%. This is not simply organic demand recovery. It is the GDP-side signature of the redistribution machinery this series has tracked since Q4 2025: the tacitly officiated SMC-AEV-MER and Prime Infra-FGEN bilateral consolidations, the suspension of real property taxes (RPT) on generation assets, the FIT-ALL-to-GEA-ALL transition, and record bank lending concentrated in the electricity sector, all of which function as implicit and direct bailouts routed through regulated utility balance sheets. (Figure 3, lowest chart) 

Averch-Johnson dynamics apply directly here: regulated firms with an assured allowable return on capital have an incentive to expand the regulated asset base, particularly where the regulatory framework permits those investments to earn an allowed return regardless of whether underlying demand is strong enough to justify them on an unregulated-market basis. That expansion has partly supported measured GDP even as the households ultimately paying for the system see no corresponding improvement in affordability. 

The same investment bias is reinforced by the (Department of Energy) DOE's broader supply-side architecture: the lifting of foreign-ownership restrictions for renewable energy, successive rounds of the Green Energy Auction Program (GEA), fast-tracking mechanisms for priority projects, and planned expansion of transmission and energy-storage infrastructure, all aimed at accelerating renewable capacity toward the government's 35% generation-mix target by 2030. These measures deliberately lower barriers to entry, accelerate project development, and create investable opportunities in the electricity sector. Combined with regulated returns, sector-specific relief, tax concessions, and concentrated credit, they help explain why electricity-related capital formation can remain a source of measured GDP growth even while the household affordability constraint remains unresolved. 

The mirror image is the transport sector, where fare adjustments remain administratively suppressed even as input costs were not. Transport equipment capital formation collapsed 27.2% in nominal and 32.2% in real terms year-on-year in Q2 2026 — the largest single component drag on gross fixed capital formation for durable equipment, alongside HFCE transport's outright contraction. 

One regulated sector was bailed into growth; the adjacent sector, denied the same price-adjustment mechanism, is disinvesting. Both outcomes are administrative rather than market-determined, which is the point: the "growth" and the "decline" are two faces of the same suppression architecture, not independent market signals. 

VII. Construction: a government-led downward spiral

Figure 4

Construction contracted 13.9% year-on-year in Q2 2026 (constant prices, production side) and gross fixed capital formation in construction fell 14.8%, the single largest driver of industry's overall 2.4% decline. General government construction collapsed 32.4% — the flood-control-scandal hangover working through the capital formation accounts a full year after the scandal broke, as officials remain reluctant to greenlight infrastructure disbursement amid ongoing accountability proceedings. Private construction did not step into the gap: financial and non-financial corporations grew a modest 3.8% and households/NPISH just 0.8%, both far too small to offset the public-sector collapse. This is not a diversified construction sector experiencing a public-led correction while private activity compensates; it is a sector where the public sector was effectively the only source of growth, and where withdrawing it exposes how little organic private capital formation exists underneath. (Figure 4, topmost window) 

VIII. Trade: exports without manufacturing depth, and a historic deficit 

The headline expenditure-side bright spot was net exports: exports of goods and services grew 12.2% (goods +17.0%, services +6.9%), comfortably outpacing 5.5% import growth and contributing 1.2 percentage points to GDP. (Figure 4, middle graph) 

But the composition matters. Export growth was overwhelmingly a semiconductor and AI-hardware story — consumer electronics up 230.3%, components/devices up 13.4%, office equipment up 77.6% — while broad-based manufacturing growth (2.6% for the sector overall) remains muted relative to that electronics surge. This is a narrow, AI-cycle-dependent export engine, not a diversified manufacturing recovery. 

Should the AI capex cycle slow — a real possibility given how concentrated the growth in a handful of product lines already is — the one clean bright spot in this release loses its main support. 

Meanwhile, the trade-in-goods deficit for the first half hit $30.81 billion, the widest since PSA's series began in 1991, even as both exports (+13.1%) and imports (+17.8%) posted record first-half nominal levels. (Figure 4, lowest diagram) 

A widening deficit funded by strong headline trade volumes is still a widening deficit: it means the economy's dollar liabilities from imports are growing faster than its dollar receipts from exports, precisely the imbalance that eventually forces itself onto the external accounts. 

IX. The external financing loop closes on itself

Figure 5

That imbalance, plus slowing organic dollar revenue from OFW compensation (per the Net Primary Income data above), means BSP's soft-peg regime and its effort to rebuild gross international reserves via Net Foreign Assets (NFA) accumulation increasingly runs through borrowing rather than organic inflow. (Figure 5, topmost window) 

The July GIR print, released the same week as the GDP data, showed reserves falling to $103.4 billion — an 18-month low — down from $104.74 billion in June, driven by BSP's own FX operations and the national government's drawdowns on foreign-currency deposits to service external debt. The reserve buffer built earlier this year via eurobond and World Bank inflows (documented in Part 13) is now being spent down to meet obligations those same inflows were meant to be seen as covering. (Figure 5, middle graph) 

And as government borrowing accelerates to fund both the fiscal deficit and the electricity-sector and BSP-relief bailouts, the crowding-out is not confined to private investment. It extends into savings. 

CMEPA-assisted flows are channeling household and institutional savings into government securities; banks and elite conglomerates are competing alongside the government itself for a shrinking pool of savings, rather than the government crowding out only private borrowers. 

It is not that bank lending is contracting — this series has already documented that lending continues at a record pace, even as signs of peaking emerge — but that banks are simultaneously amassing government securities as an ever-larger share of their balance sheets, reinforcing the sovereign-bank doom loop already flagged in Parts 11 through 13: banks funding the sovereign, the sovereign's creditworthiness increasingly resting on banks that are themselves increasingly exposed to the sovereign. 

X. The Two Precarious Trends Beneath the Headline 

First, on timing: headline year-on-year GDP growth has been decelerating in trend since Q2 2021 — the quarter immediately following BSP's historic pandemic-era bank rescue measures — with that deceleration visibly accelerating from Q2 2025 onward, when the flood-control scandal surfaced, and again through 2026 as the Iran-oil shock compounded. This is not a one-quarter air pocket; it is a five-year decay curve with two discrete accelerant events layered onto it. (Figure 5, lowest visual) 

Second, on the trend itself: both nominal and real GDP now sit at what should be read as precarious trend support. If either the year-on-year growth trend or the nominal-GDP trend breaks decisively from here, a technical recession moves from a tail risk to a live scenario — not because of a single bad quarter, but because the growth that has been recorded through 2025–2026 has been substantially manufactured through price suppression, debt-financed government consumption, and administratively engineered sectoral wins (electricity) offsetting administratively engineered sectoral losses (transport, construction). Remove the suppression and the debt financing, and the underlying trend has already been decelerating for five years. 

XI. Confusing Stagflation with an Event Rather Than a Process 

The recurring objection to this series is that “stagflation” has a technical definition—a threshold combination of low growth and high inflation, sometimes with high unemployment—and that 2.3% growth with 6.2% inflation may or may not clear that bar depending on which textbook is consulted. This misunderstands what the term is doing analytically.

Figure 6

As I put it recently: stagflation isn't a one-off event or merely a set of statistics. It's a cumulative process. GDP, CPI and employment are symptoms, not causes. The 1970s oil shocks exposed and intensified underlying imbalances that had already been building. (Figure 6, topmost window) 

Applied today, the economy could continue posting positive GDP growth even as shocks generate severe price pressures and distortions, with debt accumulation and policy accommodation allowing the underlying imbalances to persist rather than forcing immediate adjustment. 

The fact that the statistics did not necessarily satisfy the later textbook definition of stagflation at every point does not mean the underlying process was absent. 

By 1983, the accumulated imbalances had produced the combination of recession, inflation and unemployment that made the diagnosis technically unambiguous. 

That is the link between this quarter's headline GDP number and the debt-growth-outpacing-GDP-growth gap documented above. See previous discussion in Part 7 and Part 4. 

Leveraged GDP is fragile in a specific, mechanical sense: it depends on the state's ability to keep borrowing at a pace that outstrips nominal output and on the central bank's ability to keep suppressing the price signals through which the economy would otherwise adapt. The Philippine response today is not simply monetary easing. It is a combination of balance-sheet transfers, administrative controls, and BSP easing and relief measures that suppress or redistribute the signals of stress across the financial system and the real economy. 

Those interventions can buy time, but they do not create adaptive capacity. Market adjustment may be difficult and disruptive, but it forces prices, capital and balance sheets to adjust to underlying conditions. 

Suppression does the opposite: it delays adjustment, redistributes the resulting imbalances and uses borrowed time to keep the existing structure operating. The longer that process continues, the more deeply the economy becomes dependent on the interventions themselves. 

Growth built this way can appear stable until it fails abruptly. 

It can hold—as it has, barely, for several quarters now—until financing conditions tighten or a ‘sudden stop’ occurs, at which point the accumulated imbalance can compress quickly. The current Iran oil shock is only five months old: it is the third wave of the inflation cycle, following the Russia-Ukraine oil shock of 2022 as the second wave. (Figure 6, middle graph) 

The important point is therefore not the latest shock itself, but the structure it has hit. As in the 1970s, an oil shock has been layered onto pre-existing imbalances and met with political responses that suppress adjustment and buy time. 

The result is visible in the record first-half fiscal deficit, the record first-half trade deficit, the second-highest first-half debt level on record, and a strained GIR-BOP position—all against a GDP growth trend that has not merely weakened but has been decelerating for five years, with that deterioration visibly accelerating through 2025 and 2026. (Figure 6, lowest chart) 

That is the significance of the 1983 episode: the crisis did not begin when the statistics finally satisfied every technical criterion. The crisis was the CULMINATION of a process that had been building for years. 

The 2.3% print is not evidence that the process is absent; it is what that process looks like while the economy is still being financed and the underlying adjustment is still being suppressed. 

XII. Conclusion: GDP Is the Symptom, Not the Diagnosis 

None of the individual figures in this release are, by themselves, damning. A quarter of soft growth, the Iran war oil shock, a construction contraction tied to a corruption scandal, a temporary dip in remittance-linked income — any one of these could be read as noise. 

What makes the Q2 print diagnostic rather than incidental is that the mechanisms keeping the headline number positive is the same mechanism this series has been tracking since Part 11: administrative price suppression flattering the deflator, debt-financed government consumption substituting for private demand, and a handful of politically favored sectors (electricity, exports concentrated in AI-linked electronics) carrying industries that are otherwise contracting or stagnant. 

Stagflation is not a reading you take off a single quarter's GDP-and-CPI print. It is what you see when you trace how that print was produced — and 2.3% growth built this way is not evidence the process has stalled. It is evidence the process is still running, and that the bill for running it is still being deferred rather than paid. 

___

Last four stagflation series

-Stagflation Part 13: Record Twin Deficits Signal the Cost of Deferred Adjustment

-Stagflation, Part 12: The Philippines' Balance-Sheet Origins of Inflation

-Stagflation Part 11: The Intervention Ecosystem Behind Moody's and Fitch's Banking Warnings

-Stagflation Part 10: The Politics of Contradiction—Rate Hikes, Liquidity Addiction, and External Constraint Under Balance-Sheet Stress