Psychologist James Reason’s ‘Swiss cheese model’ holds that major disasters in complex systems are the result of the accumulation and alignment of multiple failures—the ‘eyes’ or holes in some kinds of cheese. Today, several risks—overvaluation, high debt, cash flow shortfalls, volatility, rising interest rates, interconnections, financial games, poor safeguards—are creating the perfect conditions for a financial crash—Satyajit Das
In this
issue
Three Markets, One Signal: The PSEi Unravels, the Peso
Weakens, and T-Bills Surge
I. Whac-a-Mole 1: The Floor Gives Way
II. The Global Mallet: Equity De-Risking and Deleveraging
Reaches the Core
III. Whac-a-Mole 2: The Peso Borrows from the Reserves
IV. Whac-a-Mole 3: The Mallet Hits the T-Bills
V. Three Markets, One Signal: The Mallet Moves the Pressure, Not the Problem
Three Markets, One Signal: The PSEi Unravels, the Peso
Weakens, and T-Bills Surge
Financial stress keeps migrating. The PSEi's support
stock cracks, the peso borrows from the reserves, and T-bills price a liquidity
squeeze while the AI unwind reaches Asia.
I. Whac-a-Mole 1: The Floor Gives Way
This week, the mallet found the floor.
Figure 1
The PSEi 30 lost 286.51 points (-5.12%) to 5,306.98, its biggest weekly drop since the week of January 31, 2025 and a 2020 low. The 5,600 "Maginot line" is gone. The March 19, 2020 closing low of 4,623.42 now sits 12.9% below. (Figure 1, topmost pane)
"ICTSI selloff drags PSEi to 6-year low," ran the headline. Wow. This is a rarity. The Overton window has almost always treated the PSEi as synonymous with the Philippine stock market.
For once, it was accurate. Partly though.
Friday. ICTSI's 7.22% crash delivered 89.52 gross points, or 52.37% of the index's 170.93-point loss (-3.12%). SM (-4.38%, 23.31 points) lifted the pair's contribution to 66%. The remaining 34% was reflected in negative PSEi breadth: just 5 advancers against 23 decliners. The broader PSE market was equally negative, with 41 advancers against 131 decliners.
Yet ICTSI alone was 49.5% of the Php 9.11B main board
volume (MBV)! (Figure 1, middle graph) Foreigners
sold Php 993.6M of its roughly Php 4.51B turnover (22%), so nothing bid
underneath it. The top 10 brokers handled 66% of MBV.
The week. ICTSI
was 136.01 gross points, or 47.5% of the 286.51-point plunge. With SM, BDO, BPI
and SMPH, the share was 64.54%. Breadth was 283-593. That -310 spread is the
widest since September 30, 2022's -326, and it is the eleventh straight
negative week (Figure 1, lowest image)
Figure 2
The average PSEi member fell 4.45%, and only two held up. ICTSI fell 10.58%, while PLUS (-16.92%) and SCC (-19.8%) kept collapsing. (Figure 2, topmost pane)
ICTSI's index weight slid to about 21.7%, yet the top 5 still carry 53.84%. (Figure 2, middle diagram)
The three big banks (BDO, BPI, MBT) rose from 20.34% to 21.12% simply by falling less.
Foreigners were net sellers for a second week (-1.128B). ICTSI's own net foreign selling (-1.271B) exceeded the entire market's net outflow. Daily main board volume rose 14.8% to Php 6.771B, with cross trades at 25.84% and ICTSI at 38.73%.
Why does one stock carry half the tape? Because a managed index changes who participates.
When the benchmark is defended, with floors under the heavyweights, rescue bids post-recess (afternoon delight), into the close (pre-closing pumps) and/or 'rotation’ from one name to the next, prices stop telling participants where the risk is. A measure managed as a target stops measuring. Those who need true prices leave, and a thin ring of brokers and cross trades works a few names. The PSE posted its lowest retail and institutional participation in 2025, and a thinner market attracts more managing, which thins it further.
What I call benchmarkism is the whole exercise: managing market outcomes through an unofficial "national team," and statistical output to project a panorama of “macro stability.” ICTSI was its primary instrument. Its rising price, valuation and volume share from July to September were engineered to put a floor under the PSEi and power a rotation-led bull market, as our three-part series laid out.
- PSEi 30: The ICTSI Show, June 07, 2026
- The PSEi-ICTSI Show, Part II: When One Stock Becomes the Market, July 19, 2026
- The PSEi-ICTSI Show, Part III: The Fulcrum Cracks—Benchmark-ism Unravels, August 31, 2026
The index was made to look stronger than the market underneath it. Concentration isn't an oversight. It is what benchmarkism leaves behind, and the establishment has been as silent on WHY it happens as on how concentrated trading has become.
Each rescue must also be bigger than the last to buy the same calm. ICTSI peaked at Php 1,027 on July 20 and was celebrated for its Php 2T market cap in early September. Then its decline accelerated. The stock engineered to be the floor was holding the index up, and when it broke, the index went with it.
Benchmarkism backfired. Price controls buy time but deepen the maladjustment, and the consequences run opposite to what was intended. In a savings-starved capital market, the floor had to be built from one stock's price, and that is the unforeseen consequence.
This week’s collapse essentially validates the arguments advanced in our three-part series.
ICTSI is now technically oversold, making a rebound possible at any time. But an oversold condition is not a reversal signal. A bounce may provide temporary relief without changing the underlying trend, restoring market breadth or repairing the concentration that benchmarkism left behind. A recovery in price should not be mistaken for a recovery in market structure.
II. The Global Mallet: Equity De-Risking and Deleveraging
Reaches the Core
Of 19 national bourses, 16 fell, with an average change
of -3.49%. (Figure 2, lowest chart)
Figure 3
Taiwan's Taiex crashed 17.5%, its worst week since 1991. The Nikkei fell 7.7% and the KOSPI 7.95%, ending the week 48% below its June record. (Figure 3, topmost diagram)
The Dow, S&P 500 and Nasdaq lost 8.02%, 8.61% and 10.02%. Gold, silver and copper fell 8.92%, 15.44% and 9.78%, and Bitcoin 14.36% (StockCharts). (Figure 3, middle and lower visuals)
The dollar rose 0.4% while the 10-year TNX slid from 5.154% to 4.912%.
The order matters. Financial market stress first appeared at the periphery, where funding is thinnest: Indonesia's crash in October, Korea, then Brazil, whose real has lost about 14–15% since the re-election and whose central bank intervened for the first time since 2024. Now it has reached the core: the Nasdaq, US banks (KBW -9.5%), the Magnificent 7, high-yield credit. Weak balance sheets lose dollar funding and risk capital first. When the core starts selling, no one is left to fund the periphery.
The Philippines sits on the periphery side of that line, so this week's PSEi selloff is the same signal on a smaller screen.
Gold falling with equities matters too. Leveraged holders
are selling what they can sell, not what they want to sell. That is
deleveraging, not rotation.
Reflexivity
turns market stress into a self-reinforcing process. Falling prices
do not merely reflect deteriorating confidence; they help produce it. Asset-price
declines impair collateral, tighten financing conditions, trigger margin calls
and force liquidation. Forced selling then pushes prices lower, further
weakening balance sheets and restricting access to funding. What begins as a repricing becomes a
deleveraging cycle. The market is no longer simply discounting economic
weakness; it is transmitting and amplifying it.
The AI unwind, in
causal order:
Falling AI
equities are the symptom. Nvidia
lost 22.9% on the week, semiconductors 17.1% and the Magnificent 7 index 10.3%.
Collateral falls
and funding tightens. Oracle's
CDS hit a record 341 bps, CoreWeave's yields jumped 84 bps to 15.11%, and
high-yield spreads widened 51 bps, the most since April 2025. Banks are
reportedly selling data-center loans at the 80s and 90s. Falling valuations
weaken collateral, while widening credit spreads raise refinancing costs. The
financing conditions that amplified the boom now accelerate the unwind.
Stress reaches
circular financing. Chipmakers
fund neoclouds, which buy chips, which are pledged against loans. The ECB
counts over $4 trillion of off-balance-sheet commitments among big tech.
The structure depends on continuously rising valuations, continued financing
and demand. A weak link therefore threatens the entire chain. What appeared as
independent sources of demand was partly circular capital formation.
The unwind reaches the supply chain and geopolitics. The US is moving to require licenses for AI chip exports to Singapore, Malaysia, Thailand, Indonesia and Brazil. Strategic initiatives such as Pax Silica and national semiconductor policies make commercial adjustment more politically costly. Governments increasingly treat AI and semiconductor capacity as strategic assets rather than ordinary commercial investments. This creates another layer of rigidity: capital, infrastructure and supply chains become harder to redirect even when their economic returns deteriorate.
But strategic
importance cannot repeal profit and loss.
The policy
response risks transferring rather than resolving the adjustment. Guarantees, liquidity injections and
monetary accommodation may interrupt forced selling and postpone balance-sheet
recognition. They cannot, by themselves, restore the economic viability of
misallocated capital. If intervention sustains excess capacity, debt and demand
against constrained productive supply, the adjustment can migrate from
financial markets into the currency and price level. Losses that cannot be
absorbed through asset repricing eventually be transferred to currency holders
through monetary accommodation and inflation.
This time the policy dilemma is sharper. The Fed is
debating hikes even as financial markets demand relief. The traditional
central-bank put collides with inflation. Easing may stabilize asset prices but weaken the currency and reignite
price pressures; tightening may contain inflation but intensify deleveraging.
Policymakers face a conflict between financial stability and monetary stability
that cannot be resolved simply by supplying more liquidity.
Figure 4
The Philippine link. Semiconductor exports soared 86% in September to a record, lifting their share of exports to a record 57.5%, but imports surged too (see Stagflation 19). (Figure 4, topmost window)
September imports rose 25.7% to $15.03B against exports up 35.6% to $9.87B, a $5.15B deficit. 9-month deficit was $47.158B—a record. The export boom is smaller than the import bill it creates. A bust cuts the export leg quickly while the oil and debt-service bills stay put.
A note on the week's outages (Microsoft, the FAA ground stop, four major US banks, United, the NSA network and Apple): coincidence cannot be ruled out, but neither should the events be viewed in isolation. In a war economy, communications, transportation, financial infrastructure and digital networks are strategic assets and potential points of vulnerability, therefore potential targets. Whether these disruptions were ordinary technical failures, stress-induced operational fragility, strategic probing or something more remains unresolved. The important point is that systemic complexity has created vulnerabilities with geopolitical consequences. The distinction between civilian infrastructure and strategic infrastructure is increasingly difficult to maintain.
The current “periphery to the core” market stress increasingly resembles a crisis dynamic in progression.
III. Whac-a-Mole 2: The Peso Borrows from the Reserves
The dollar rose against 7 of 10 Asian currencies, yet the regional average was +0.02%. India's rupee and Japan's yen rallied, offsetting weakness in the Korean won and ASEAN currencies. USDPHP rose 0.14% to 63.81, close to the 64.17 record of October 27.
That is a quiet print in a week when Brazil's real fell 4.7%, South Africa's rand 3.7% and Mexico's peso 3.3%. A stable price in a stressed week is information only if you ask who paid for it. The reserves did.
The BSP reported a $1.5B October BOP deficit, $10.6B for January to October, and GIR down to about $98B. The 10-month deficit is 15% above the BSP's $9.2B full-year forecast, and the BSP's year-end reserves forecast of $104B is $6B above where reserves stand with two months left. The forecast has already failed. (Figure 4, middle graph)
The Treasury also awarded an initial $453.4M from its third Retail Onshore Dollar Bond issuance at a 5.65% coupon, against $839.5M in tenders—4.2 times the initial $200M offer. The offer remains open until November 24.
The onshore issuance mobilizes dollar savings for sovereign financing rather than directly generating new external borrowing from international capital markets. It converts privately held dollar claims into a sovereign dollar liability, which the government must eventually service in foreign currency. Residents queuing more than four times over for dollar paper at 5.65% is also a statement about how they expect the peso to behave.
The mainstream awaits a Q4 spending recovery, but Q3 GDP grew 2.1% and nine-month growth is 2.45%. Importantly, Q4 and household GDP has been materially slowing since 2021, printing 3.8% and 3% in 2025. (Figure 4, lowest image)
Year-end government spending without real savings behind it is financed by money creation or foreign borrowing.
Expect the record
fiscal gap to widen, alongside another record trade and current account
deficit (twin deficits). That widens the saving-investment gap, which only
foreign funding or reserve depletion can bridge. The BOP is where the bridge
shows up.
IV. Whac-a-Mole 3: The Mallet Hits the T-Bills
Figure 5
The Philippine (BVAL) yield curve is flattening from both sides Bills are being pulled up while the 3- to 10-year belly is pulled down, closing the spread from both sides. The 1-, 3- and 6-month bills rose 44.4, 22.58 and 10.54 bps on the week, reaching multi-year highs.
Mid-curve moves were +7.79 (1y), -4.04 (2y), -10.13 (3y), -14.16 (4y), -15.91 (5y) and -15.9 bps (7y), while the 10-year fell 8.2 bps. Against October 30, bills are up 36 to 73 bps and 7- to 10-year yields down 10 to 50 bps, while the 20- to 25-year sector is up only about 4 bps, essentially unchanged. (Figure 5, topmost pane)
Against October 30, bills are up 36 to 73 bps and 7- to 10-year yields down 10 to 50, while the 20- to 25-year is up only about 4 bps, essentially unchanged. (Figure 5, middle image)
The front end is being pulled up, the belly pulled down, and the long end barely moves. The front-end repricing raises the possibility that markets are anticipating further BSP tightening. (Figure 5, lower graph)
Persistent peso weakness, reserve depletion and inflation risks constrain the BSP's policy marking ability to either ease, or force it to raise rates despite weakening economic growth. The multi-year highs in short-term bills reflect this tightening pressure, although market yields do not establish that a policy hike is inevitable.
There are three candidates for the belly's rally:
- Weak growth. Materially slowing Philippine GDP growth
points to weaker economic activity and lower future credit demand, putting
downward pressure on intermediate-term yields.
- Imported duration
demand. US 10-year yields fell
about 20 bps and 2-year yields 22 bps this week. Falling US yields provide an
external tailwind for Philippine intermediate bonds, even as domestic
short-term funding costs rise.
- Domestic liquidity. This is likely the more consequential factor. Diminishing bank liquidity buffers and tighter funding conditions coexist with growing demand for government financing. Banks and other institutions face competing demands for liquidity, while short-term government securities must offer increasingly higher yields to attract funding. This helps explain why bills are repricing upward even as intermediate yields decline.
There is another complication. The local bond market is
losing both turnover and foreign participation.
Figure 6
PHP Treasury volume keeps falling and now sits at 2024 lows. (Figure 6, upper graph) The market is clearing with less turnover, making price movements more dependent on the marginal transaction. Lower volume does not establish the absence of demand, but it raises the question of how much depth remains to absorb a larger adjustment.
The BSP reports four consecutive quarters of foreign outflows from Philippine bonds, bringing cumulative withdrawals to $4.477 billion. Q3 alone accounted for $748 million, while year-to-date outflows reached $2.941 billion. (Figure 6, lower chart)
This sustained retreat came as the 10-year yield hit a multi-year high. The foreign bid built around the announced JPMorgan index inclusion, starting next year, is retreating even before the index arrives.
Higher yields are also a fiscal fact. Every bill that rolls over at a multi-year high raises the cost of refinancing a record deficit, and the interest bill competes with everything else the budget has promised.
The adjustment has moved from the stock market and the
peso into the government's own cost of funds.
V. Three Markets, One Signal: The Mallet Moves the Pressure, Not the Problem
- In stocks, the support stock became the fault line.
- In the peso, stability is being sustained through reserve depletion against persistent external deficits.
- In bills, the price of liquidity is rising while market volume shrinks.
All three expose the same shortage of real savings, while the AI unwind removes the external subsidy that had been papering over it.
Engineering support for the heavyweights does not create market breadth; it suppresses it. Smoothing the peso through intervention and dollar borrowing does not create savings. Institutional demand for government securities at higher yields does not eliminate the cost of debt. Each intervention moves the pressure to the next price that cannot be controlled.
The mole was never the problem. The mallet is.
Prices are information. Strategic importance cannot repeal the laws of scarcity.



















