This piece is the middle portion of Chapter 19 (2) of The Declaration of the Age of Physical Economics (Yoon Jong-won, Yoon So-ri, Yoon Jun). It is an academic exposition presenting the authors' physical economics hypothesis, and the body, figures, and citations follow the manuscript as written.
The 2008 Global Financial Crisis: The Proposition That the Five DTDMC Stages Proceed in the Same Order
The DTDMC Five-Stage Collapse Pathway Law is the strongest proposition of the theory established earlier. A crisis begins at the factor stage, moves on to the onset stage, enters the blockade stage, passes through the manifestation stage, and finally concludes in annihilation or recovery. The statement that these five stages proceed in the same order is the essence of the Pathway Law. If, in a given crisis, the blockade were to ignite first and the onset were to occur afterward, or if the manifestation were to appear before the blockade, the Pathway Law would collapse. The best material for verifying whether these five stages truly follow the same order in the data is the 2008 global financial crisis.
There are reasons for choosing this crisis. First, the events are all recorded with exact dates, so chronological tracing is possible. Second, the stages proceeded slowly over fourteen months, so each stage is identified separately. In a fast crisis the stages are compressed and hard to distinguish, but in the global financial crisis the time intervals between the stages are sufficient. Third, the marker events that signal entry into each stage all remain in the official record, so after-the-fact verification is possible. From the collapse of the Bear Stearns hedge funds, the BNP Paribas redemption freeze, the sale of Bear Stearns, the bankruptcy of Lehman Brothers, the NBER recession declaration, the S&P 500 trough, to the KOSPI recovering to 1,000, every event remains together with its date.
The factor stage had been accumulating for several years before the crisis ignited. In the United States, the household debt principal-and-interest repayment ratio reached a peak of 13.0 percent in 2005 and 13.2 percent in 2006, and the house-price-to-income ratio rose as high as 5.1 times in 2006. In Korea too, the interest burden on households was accumulating, and as the Gini coefficient rose to 0.340, income distribution was worsening. In both countries, microcalcification had thickly built up in the microvasculature of households. At this stage no sign of crisis appears on the surface. It is the same state as a patient in whom cholesterol is building up on the vessel walls but whose chest does not hurt. In the chronological structure of the primary trigger (root), the secondary trigger (lifestyle), and the tertiary trigger (acute event) established by the funnel theory, this was the period when the primary and secondary triggers were slowly being filled.
The first event of the onset stage occurred in June 2007. Two hedge funds under Bear Stearns went bankrupt due to subprime mortgage losses. This was the event in which the emergency vault was shaken for the first time. It revealed that the equity capital that financial institutions had accumulated in preparation for a crisis was exposed to the fall in real estate prices. Two months later, on August 9, 2007, BNP Paribas suspended redemptions on three asset-backed securities funds. This was the point at which the market recognized for the first time that there were insufficient funds to be withdrawn from the emergency vault. From this time, trust in the short-term funding market began to collapse, and the phenomenon of financial institutions not lending funds to one another appeared. This corresponds to the point in the human body when parathyroid hormone is activated and calcium begins to leach out of the bone. The first event in which the tertiary trigger of the funnel theory, the acute-event trigger, ignited was the BNP Paribas redemption freeze.
Entry into the blockade stage began in March 2008. On the 16th of that month, America's Bear Stearns was announced as being sold to JPMorgan Chase at 2 dollars per share. This was the point at which the tertiary trigger of the funnel theory ignited strongly once again, and it was the month in which, in the Korean data, the simultaneous-warning axes of the blockade diagnosis crossed the threshold for the first time. Abnormal signals began to appear simultaneously across several of the 9 axes, and the blockade diagnosis entered the caution stage. In April it broke through the alert threshold, and with simultaneous warnings across multiple axes, the CAM blockade ignited. The state in which policy signals fail to reach households and small business owners had begun. At the same time, the official position of the Korean government was that "the impact of the U.S. subprime problem on Korea is limited," but the diagnostic engine had already ignited the blockade. The blockade state was maintained from May through July, then there was a temporary easing in July, and after it re-ignited in August, the dual blockade was established at the point when Lehman Brothers filed for bankruptcy on September 15, 2008 (the peak ignition of the tertiary trigger).
The manifestation stage began in earnest from September 2008, when the dual blockade was established. Just as microcalcification surfaces in seven forms in the human body, seven manifestation patterns appeared in the economy as well. In the capital market, 1M (obstruction, acute infarction) was manifested first. With the Lehman bankruptcy in September the KOSPI plunged, and in October Korea's won/dollar exchange rate broke through 1,400 won and the KOSPI fell 22 percent for the month. In November, 6M (disconnection) appeared. Korea's exports fell 18 percent year on year, and in the global supply chain Korean component companies suddenly became unable to receive orders. In the same period, 2M (slowdown) also proceeded. Industrial production declined in tandem, the employment statistics worsened, and in December the simultaneous-warning axes expanded further. 3M (dilapidation) was also observed alongside. Even though the Bank of Korea cut the base rate by a cumulative 225 basis points from October through December, that signal did not reach households, and the appearance of policy non-response emerged. The way the four manifestation patterns 1M, 2M, 3M, and 6M appeared simultaneously or sequentially is the core characteristic of the global financial crisis, and the definition of "single or compound manifestation" predicted by 7M theory operated exactly as such in the data.
GFC 2008: The 14-Month DTDMC Five-Stage Time Series (Including the Tertiary Trigger and Manifestation Patterns)
| Month | DTDMC Stage | Score Grade | Simultaneous Warning Axes | Manifestation Pattern | Event at the Time |
|---|---|---|---|---|---|
| 2007-06 | Onset (outflow begins) | Below threshold | · | : | Two hedge funds under Bear Stearns collapse. First shaking of the emergency vault |
| 2007-08 | Onset (tertiary trigger) | Below threshold | · | : | 8.9 BNP Paribas redemption freeze. Collapse of trust in the short-term funding market |
| 2008-03 | Blockade imminent | Caution | Several axes simultaneously | · | 3.16 Sale of Bear Stearns. First breach of the simultaneous-warning threshold |
| 2008-04 | Blockade entered | Alert | Multiple axes simultaneously | · | CAM blockade ignites. Government: "impact limited" |
| 2008-07 | Single blockade temporarily eases | Caution | Multiple axes simultaneously | · | Oil price reverses after peaking at 147 dollars. CAM temporarily lifted |
| 2008-08 | Blockade re-ignites | Alert | Multiple axes simultaneously | · | CAM re-ignites. U.S. housing distress begins to spread |
| 2008-09 | Dual blockade established | Alert | Most axes simultaneously | 1M (acute infarction) | 9.15 Lehman bankruptcy. CAM + DLT established simultaneously |
| 2008-10 | Manifestation proceeds | Crisis | Most axes simultaneously | 1M (capital market) | Won/dollar breaks 1,400 won. KOSPI -22% for the month |
| 2008-11 | Manifestation (disconnection) | Crisis | Most axes simultaneously | 6M (supply chain) | Exports -18.3%. Global supply chain disconnected |
| 2008-12 | Manifestation peak | Crisis | Most axes simultaneously | 2M+3M (slowdown + dilapidation) | BOK cumulative 225bp cut, policy non-response |
| 2009-02 | Recovery entered | Crisis | Most axes simultaneously | Recovery signals begin | First supplementary budget of 28.4 trillion. Direct reach to households begins |
| 2009-03 | Recovery proceeds | Crisis | Most axes simultaneously | Recovery proceeds | 3.9 S&P 500 global trough. KOSPI recovers to 1,000 |
| 2009-04 | Blockade lifted | Below threshold | · | : | Dual blockade lifted. Two months ahead of the NBER end declaration |
In this 14-month time series there is not a single case in which the five stages were reversed. The factor accumulated first, then the onset ignited, the blockade entered following the onset, and after the blockade was established the manifestation unfolded. Recovery proceeded from the point within the manifestation stage at which the policy prescription began to take effect. The point that the five stages precisely followed the order the theory predicted is the conclusion proven within the 14 months of a single crisis. The chronological order of the funnel theory's primary, secondary, and tertiary triggers was also clearly traced. The primary trigger (the root accumulation of household debt and income polarization) was filled over 2005 to 2006, the secondary trigger (the lifestyle of subprime issuance practices) was filled over 2006 to 2007, and the tertiary trigger (the acute event) ignited in stages, beginning with the collapse of the Bear Stearns hedge funds in June 2007 and continuing to the Lehman bankruptcy in September 2008.
The same order is confirmed in the 1997 foreign exchange crisis as well. The accumulation of foreign debt in the mid-1990s was the factor stage, the crash of the Thai baht in July 1997 was the external-event ignition of the tertiary trigger, the CAM blockade proceeding from May to November 1997 was the blockade stage, the chain of large-corporation bankruptcies after November 1997 was the manifestation stage, and the 1998 IMF program and the recovery of foreign exchange reserves were the recovery stage. The 2003 card crisis followed the same order as well. The excess of household credit in 2002 was the factor, the rise in card-company delinquency rates in early 2003 was the onset, and after the CAM blockade ignited from May, the bankruptcy of LG Card in November was the manifestation stage. The types of crisis differ, but the order of the stages was the same in all of them.
There is one more point to note. It is that the advance-detection lead time was almost the same in every crisis. About 6 months in 1997, about 6 months in 2003, about 6 months in 2008. In the three crises, the time from the ignition of the blockade to the trigger event was almost identical. This is not a coincidence. It is because the physical time over which the simultaneous warnings accumulate until the blockade is established is similar regardless of the type of crisis (the ignitions of 1997 and 2003 were single blockades of CAM alone, while 2008 was a dual blockade). This differs from the way existing models produce outputs like "a 47 percent probability of recession in 6 months." Because those models estimate a probability, the lead time differs each time. Because the diagnostic engine measures the physical time of the blocking process itself, a regularity appears in the lead time. This regularity is also direct evidence for the statement that the diagnosis is not a forecast but a measurement.
The verification of the DTDMC Pathway Law is completed in two respects. One is the sameness of the order of the stages. In the four crises the five stages all proceeded in the same order, and cases of reversal numbered zero. The other is the regularity of the lead time. In the three micro crises, the same advance detection of about 6 months occurred. The conclusion to which these two facts together point is that the temporal structure the theory predicted actually exists in the real data.
The Multiplicative Structure of CAM and DLT: Verification of the Self-Recovery of the Single Blockade
The mathematical core of the dual blockade theory is expressed in a single line. It is the statement that the intensity of the blockade is determined by the multiplication of CAM and DLT. The essence of this proposition established earlier is that it is multiplication, not addition. If the two blockades were added, then even if one side were 0, the blockade would be established to the extent of the value of the other side. But if they are multiplied, then when one side is 0 the product also becomes 0, and the blockade is not established. The difference between addition and multiplication is, mathematically, a difference of a single character, but the results they produce in the data are exact opposites.
If we recall the nonlinearity of multiplication in everyday life, this difference becomes clear. Suppose there is a task in a company that can only be completed when two employees collaborate. If one person works at half efficiency and the other works at half efficiency, the two efficiencies multiply and the whole proceeds only one quarter of the way. If one person has 0 efficiency (does no work at all), then no matter how hard the other works, the whole becomes 0. The same principle operates in blockade. If the signal channel and the physical channel each narrow by half, the total flow plunges to one quarter, and if one side is normal, then even if the other side is blocked, the entire system is not paralyzed.
The best material for verifying whether this multiplicative structure truly exists in the data is the three cases of self-recovery of the single blockade in Korea. During the 351-month verification period, out of the 344 months excluding the 7 months in which the dual blockade ignited, there were three occasions on which one side of the blockade was partially activated but the other side was not established, so the system recovered on its own. The aftermath of the collapse of the U.S. information technology bubble in 2001, the downgrade of the U.S. Treasury credit rating in 2011, and the period of the sharp rise in U.S. interest rates in 2022 are those three periods. In all three periods a considerable external shock was applied to the Korean economy, but it recovered on its own without igniting into a dual blockade.
In 2001, DLT (channel blockade) was partially activated but CAM (signal blockade) was not established. In the aftermath of the collapse of the U.S. information technology bubble, the KOSPI fell to the 500 line, and Korea's information technology exports and venture companies were shocked. The figures of the blockade diagnosis rose to the caution stage in some months but did not break through the alert threshold. Among the 9 axes, stress was detected in part of A3 (digestive system) and A8 (demographic and vulnerable groups), but A1 (circulatory system) maintained the stable stage, and the number of simultaneous-warning axes did not cross the threshold either. The only trigger activated was cost explosion (I), and no more. The signal channel by which the government's interest rate policy is transmitted to households was alive, and the Bank of Korea's rate cuts eased the interest burden on households to a certain degree. In the multiplication, CAM was close to 0, and as a result the product was also close to 0. The system did not collapse, and in 2002 it entered a recovery phase.
In 2011 it was the reverse. The downgrade of the U.S. Treasury credit rating shook the global financial market, and CAM was partially activated. The figures of the blockade diagnosis rose to near the alert threshold in some months, and stress was detected simultaneously in A1 (circulatory system) and A5 (immune system). The only trigger activated was monetary sclerosis (H). This means that the signal of the international funding market was blocked. But DLT was not established. The scores of A3 (digestive system) and A8 (demographic and vulnerable groups) maintained the stable stage, the income and consumption paths of Korean households were being maintained, and the accumulation of household microcalcification was at a level able to absorb the shock. The signal was cut off for a moment, but the channel was alive. This time too the product was close to 0, and the system recovered on its own.
In the period of the sharp rise in U.S. interest rates in 2022, DLT was again partially activated. As the interest burden on Korean households surged, the consumption channel began to narrow. The figures of the blockade diagnosis rose to the caution stage in some months, and stress was detected simultaneously in A3 (digestive system) and A8 (demographic and vulnerable groups). Two triggers were activated: cost explosion (I) and price surge (A). But CAM was not established. The Bank of Korea was maintaining the path by which policy signals are transmitted to households, and the government's household support policies were operating. In a state where only one channel was blocked and the other was alive, the system once again recovered on its own. In all three cases the periods and the types of shock differed, but the principle of the multiplicative structure operated identically.
The Self-Recovery of Korea's Three Single Blockades: Quantitative Verification of the Multiplicative Structure
| Period | External Shock | Active Blockade and Stress Axes | Active Trigger | Result |
|---|---|---|---|---|
| 2001 | U.S. IT bubble collapse, KOSPI 500 line | DLT (channel blockade) partial (caution stage), A3 (digestive system) and A8 (demographic and vulnerable groups) stress | I alone | CAM (signal blockade) not established, product close to 0. Self-recovery |
| 2011 | U.S. credit rating downgrade, global financial turmoil | CAM (signal blockade) partial (near the alert threshold), A1 (circulatory system) and A5 (immune system) stress | H alone | DLT (channel blockade) not established, product close to 0. Self-recovery |
| 2022 | U.S. interest rate surge, sharp rise in household interest burden | DLT (channel blockade) partial (caution stage), A3 (digestive system) and A8 (demographic and vulnerable groups) stress | I+A | CAM (signal blockade) not established, product close to 0. Self-recovery |
The three cases directly prove the proposition of the multiplicative structure. If the two blockades had a structure in which they were added, then a partial crisis should have ignited from the partial activation of one side alone. But in all three periods it did not progress into a crisis and recovered on its own. The mathematical property of multiplication, that when one side is close to 0 the product becomes close to 0, operated exactly as such in the economic data. In the contrast between the four dual blockade crises (IMF, card, GFC, COVID) and the three self-recoveries of the single blockade, the proposition that the two blockades are multiplied is proven most vividly.
The same multiplicative structure is confirmed within a single crisis as well. Looking again at the 14-month time series of the 2008 global financial crisis, there was one self-recovery in July. The CAM blockade was maintained through June, was temporarily lifted in July, re-ignited in August, and then progressed into a dual blockade in September. The temporary lifting in July was the result of the price pressure being temporarily eased as the international oil price reversed after hitting a peak of 147 dollars per barrel. When the pressure on one channel was released for a moment, the other channel was released together, and this oscillation itself shows the fact that in the multiplicative structure the product is sensitive to the two variables simultaneously. The moment a single blockade is briefly released suggests the possibility of self-recovery, and the moment a single blockade re-ignites and meets the blockade on the other side signals the imminence of a dual blockade.
The 344 months of silence carry yet another meaning. It is the contrast with existing alert systems. The credit-to-GDP gap of the Bank for International Settlements (BIS) exceeded its threshold several times in Korea during the 2010s, but there was not a single case in which it led to an actual crisis. Bloomberg Economics' U.S. recession probability model output 100 percent in October 2022, but the U.S. economy went on to continue growing for the following two years. This is the result produced by the method of issuing an alert immediately as soon as a single indicator merely crosses a threshold. Because the multiplicative structure of the diagnostic engine requires the condition that two blockades cross the threshold simultaneously, in the case where only one variable is shaking, an alert is not triggered. The point that the 344 months of silence are not mere silence but the result of the operation of the multiplicative structure is confirmed in the cases of self-recovery of the single blockade.
COVID in 2020: Verification of the Macro Crisis Classification and the Prescription of Direct Injection into the Microvasculature
The COVID shock of 2020 is a case that simultaneously verified two propositions within the 351 months of Korean data. The first proposition is the fact that the macro crisis category actually exists, namely the statement that not all crises share the same mechanism but that external acute-shock-type crises and internal structural-accumulation-type crises exist as separate categories. The second proposition is the fact that the prescription of direct injection into the microvasculature operates in the data, namely the statement that when policy funds reach households directly without passing through large financial institutions, the speed of the lifting of the blockade becomes dramatically faster. The one-month recovery of COVID proves these two propositions from both sides.
Let us begin from the fact that V-Series output "no detection." If a diagnostic tool fails to detect a crisis, two interpretations are possible. Either the tool is broken, or the crisis itself lies outside the measurement range of the tool. As examined in Chapter 17, V-Series is a tool that measures structural vulnerability, a tool that asks how deeply a system would collapse when the same shock strikes. The result that this tool did not detect COVID proves not a failure of the tool but the accuracy of the crisis-type classification. This is because COVID was a macro crisis, that is, an acute shock applied from the outside.
Transposed to the human body, it is like a traffic accident. Even a person who is usually healthy can suffer a fracture if they are in a traffic accident, and even a person with an underlying condition will not suffer a fracture if they are not in an accident. A health checkup report cannot predict a traffic accident. Because what the checkup measures is the underlying condition and not an external event, one cannot call that a failure of the checkup. If V-Series had triggered a COVID alert in 2019, that would rather have been a false detection and a defect of the system. This is because it would have meant that it responded to an internal signal that did not exist.
Let us examine the data-based grounds on which V-Series output no detection. In 2019, Korea's growth rate of the working-age population was maintaining a positive value of +0.20 percent, and the Gini coefficient, at 0.345, was at a level similar to that just before the 2008 global financial crisis (0.340). The household debt principal-and-interest repayment ratio was also not greatly different from its level at the time of the global financial crisis. The domestic demand base had not collapsed either. None of the 9 gauges had entered the serious stage, and the external shock was met in a state where structural vulnerability had not accumulated. V-Series "no detection" was the accurate classification that "this crisis did not come from an underlying condition."
The output pattern of the blockade diagnosis was in a direction consistent with the V-Series non-detection. In November 2019 the simultaneous-warning axes crossed the threshold for the first time, and in January 2020, as the number of simultaneous-warning axes expanded further, the CAM blockade ignited. In February the D (income polarization) and I (cost explosion) triggers were activated, and the CAM blockade was maintained through March as well. On March 11, 2020, the World Health Organization (WHO) declared COVID a pandemic, and lockdowns and social distancing went into full swing in both Korea and the United States. In April the dual blockade was established. The largest number of axes among the 351 months entered simultaneous warning, and the triple trigger of D+I+H was activated. In that same April, Korea's exports plunged 24.3 percent year on year, and first-quarter GDP recorded negative growth of -1.3 percent.
There is one point to note here. In a macro crisis, advance detection is impossible in principle. Because a micro crisis requires about 6 months of time in the process by which the accumulation of internal structure progresses into a blockade, the system catches the signal during that accumulation process. In a macro crisis, the shock arrives suddenly at the moment the external event occurs, so there is no accumulation process itself. One cannot catch the signal of accumulation in a state where there is no accumulation. The result that in COVID the blockade diagnosis caught the dual blockade in the same month (April 2020) means that the tool operated precisely at the earliest point at which it could catch it. To demand 6-month advance detection for a macro crisis is like demanding that a blood pressure monitor predict a traffic accident.
The strongest evidence proving the accuracy of the diagnosis was the speed of recovery. On March 27, 2020, the U.S. government passed the COVID response bill (the CARES Act) of 2.2 trillion dollars, and at the same time the Korean government too directly paid emergency disaster relief funds of the first supplementary budget of 14.3 trillion won to households. What the two prescriptions had in common was that the funds reached households directly. Rather than funds that pass through large financial institutions and, released into the market through quantitative easing, no one knows where they will flow, these were funds deposited directly into households' bank accounts. Once these funds reached the microvasculature (households), the blockade was lifted in just one month, in May 2020.
This conclusion is precisely the result in which the effect of the prescription of direct injection into the microvasculature was proven with data. The prescription principle established in the canonical text, namely the statement that "even if you transfuse into the aorta (large corporations), it does not reach the microvasculature (households), so a direct injection into the microvasculature is needed," was confirmed to have operated in the data rather than remaining a theoretical claim. If, in that same period, the government had put funds only into support for large corporations, there is a strong possibility that the funds released through quantitative easing would have flowed into the asset markets and driven up real estate and stock prices. Funds that do not reach the microvasculature cannot contribute to the lifting of the blockade. The one-month recovery of COVID was a result made possible because the funds reached the microvasculature directly.
The fact that all three diagnostic tools drew the same picture strengthens the diagnostic weight of this case. V-Series output "no detection" and classified the crisis type as macro. The blockade diagnosis output "same-month detection" and showed an accurate output within the design range of a macro crisis. The satellite output "a temporary plunge followed by a V-shaped rebound" and confirmed the physical recovery. The result that the three tools, measuring in different ways, pointed to the same conclusion (a macro crisis, rapid recovery possible) proves the internal consistency of the entire system. It is not that the output of one tool coincidentally matches the output of another tool, but that the three tools were aligned because they measured the same physical fact from different angles.