In the previous part 2, we saw the measurement of the four crises (1997, 1998, 2003, 2008, 2020). The three endogenous crises were warned six months ahead, and the one exogenous shock came simultaneously without warning. Now we examine what this result means, how it complements existing early-warning systems, why it is six months of all things, and the possibility of a preventive economics that this grammar opens.
The Meaning of This Result: The Domain-Independent Verification of Gradient Theory
Compressed into a single sentence, the result this chapter has shown up to here is this. The fact that the grammar of gradient that came from the human body operates in the same form in the completely different domain of the economy has been confirmed in 690 months of data from Korea and the United States. The meaning of this result does not lie at the level of updating an individual theory of economics. It lies in raising the strength of the claim that the grammar of gradient this book has presented is a domain-independent physical law.
The grammar of gradient in the medical field overlaps in many parts with the accumulated results of adjacent fields such as the vascular hypothesis, tumor microenvironment research, and microvascular dysfunction research. This overlap is positive evidence supporting the validity of the grammar of gradient, while at the same time leaving open the possibility that this grammar is confined to phenomena peculiar to medicine. The confirmation of the fact that the same grammar operates in the completely different system of the economy narrows this possibility. Whether the medium is calcium or cash, whether the storehouse is bone or foreign exchange reserves, whether the blockade is microcalcification or loan interest, the higher grammar of gradient operates without being subordinate to the concrete medium beneath it. This is the meaning of domain-independence.
In the history of science, domain-independent laws are discovered rarely, and when discovered they have structurally changed the understanding of the domain in question. The second law of thermodynamics was discovered in the steam engine but extended all the way to the phenomena of life and information theory, and the power-law distribution of network science was discovered in the structure of the Internet but repeatedly confirmed all the way to protein interactions and social relationship networks. Whether the grammar of gradient this book argues for possesses domain-independence at that level is a matter that must be verified over the coming decades. The result shown in this chapter is the first step of that verification, and it can only be said to have met the conditions necessary as a first step. Any claim beyond that will be told by the data to come.
The Relationship with Existing Early-Warning Systems
The International Monetary Fund (IMF) has developed and operated an early-warning system called the Vulnerability Exercise (VE) since the Asian foreign exchange crisis of the late 1990s. This system is a structure that comprehensively evaluates each country's foreign exchange reserves, short-term foreign debt, current account, fiscal state, and financial system soundness and scores its crisis vulnerability. The credit-to-GDP gap indicator developed by Claudio Borio and Mathias Drehmann of the Bank for International Settlements (BIS) is also widely used in a similar context, and this indicator has been officially adopted as the basis for the countercyclical capital buffer in the Basel III framework (Borio & Drehmann, 2009, BIS Working Papers).
The relationship between these existing systems and this book's diagnostic engine is not replacement but complement. The IMF VE and the BIS credit gap have the strength of evaluating crisis vulnerability from a medium-to-long-term perspective, have high cross-country comparability, and have an institutional foundation that policymakers have long utilized. This book's engine has the strength of identifying the timing of a crisis at the short time resolution of a monthly unit, decomposing the type of blockade into the signal dimension (CAM) and the channel dimension (DLT), and structurally distinguishing an exogenous shock from an endogenous accumulation. The two approaches are tools that answer different questions, and when used together they can complement each other's limitations.
It is useful to make one concrete comparison. At the time of the 2008 global financial crisis, the IMF VE pointed out in advance the vulnerability of the subprime mortgage market, but it did not predict the exact point at which this vulnerability would escalate into a system-level crisis. This is because the original purpose of the vulnerability evaluation system is not timing prediction but structural risk diagnosis. This book's engine judged the precursor of dual blockade in the United States in March 2008 during the same period, and warned of the full-scale collapse of September six months ahead. These two results do not conflict. If the IMF VE said "this system is structurally precarious," this book's engine said "this system will collapse within the next six months." When the two statements are together, both the breathing room and the precision of the policy response are secured.
The Physical Basis of the Six-Month Lead
Why six months of all things? This question naturally follows after the measured result of this chapter emerges, and the Gradient Precedence Principle this book presents answers this question. The core of the principle is simple. The moment the system's state value Φ(t) crosses the threshold comes later than the moment its rate of change S(t) = dΦ/dt bends significantly in the negative direction. This principle, that the collapse of the gradient precedes the collapse of the value, is the physical foundation of all leading warnings.
The concrete lead time of several months is determined by the system's characteristic time constant. In an economic system, the time in which a one-step change of gradient propagates to the entire system is defined by several factors. The debt maturity structure of households and businesses, the capital adjustment cycle of financial institutions, the delay of the central bank's policy response, and the demand-and-supply adjustment delay of the real economy are those factors. In the case of Korea and the United States, the response time of the entire system that these factors synthesize and produce is observed to converge to roughly around six months. This six months is not a mathematical necessity but an empirical constant, and a different value can emerge in a different system.
This physical interpretation also connects with adjacent research in economics. A paper published by Marten Scheffer of Wageningen University in the Netherlands in Nature in 2009 formalized the "critical slowing down" phenomenon that complex systems show just before a critical transition, and this phenomenon is commonly observed in ecosystems, climate systems, and financial markets (Scheffer et al., 2009, Nature). Critical slowing down is the phenomenon in which the speed at which a system recovers from a small perturbation slows, and expressed in the language of gradient, it is a state in which S(t) approaches 0 and recovery resilience weakens. This book's Gradient Precedence Principle provides a physical cause for Scheffer's observation. The closer a system comes to the threshold, the earlier the negative-direction acceleration of the gradient occurs, and because that acceleration is observed ahead of the change in the surface state value, a leading warning holds.
Micro vs Macro: The Meaning of the Two-Way Classification of Crisis Types
The most important typological distinction that runs through the four crises this chapter addressed is the two-way classification of the endogenous Micro Flow Crisis and the exogenous Macro Flow Crisis. The previous three crises (1997 Asian and 1998 LTCM, 2003 credit unwind, 2008 global financial crisis) all belong to the first type. The feature of this type is that the system collapses nonlinearly after a particular trigger on top of the long-term accumulation of gradient, and in this type the engine can provide a leading warning of around six months. The 2020 COVID shock belongs to the second type, and in this type, because an external shock paralyzes the system immediately without the accumulation of gradient, a leading warning is impossible in principle.
This two-way classification structurally makes the response method for a crisis different. For the response to the Micro type, preemptive intervention that unwinds the accumulated deposition in the stretch where the gradient has not yet fully collapsed is most effective. Concretely, the gradual reduction of excessive debt, the capital strengthening of financial institutions, and the early calming of an asset market where a bubble has formed correspond to this stretch. By contrast, for the response to the Macro type, the ex-post intervention that unwinds the paralysis of the system with immediate liquidity supply and fiscal transfers at the moment the shock occurs is key. That the Fed introduced zero interest rates and unlimited quantitative easing within weeks in March 2020, and that the Bank of Korea urgently cut the base rate by 50bp, are both cases of the immediate response suited to the Macro type.
This distinction provides practical guidance to policymakers. When a crisis occurs, one must first judge which type it belongs to, and according to that judgment select the time structure and means of the response. Applying a Macro response to a Micro is an overreaction, and applying a Micro response to a Macro is an underreaction. This book's engine has, within the structure of the binary verdict, a time profile that can distinguish the two types built in, so the verdict result itself suggests the direction of the response. If there is a leading warning, it is the Micro type, and if there is not, it is the Macro type. This simple distinction clearly divides the direction of the response.
What the Economy Learns from the Human Body: The Possibility of Preventive Economics
The most important implication the measured result of this chapter points to is the possibility that the concept of prevention can be introduced into the management of economic crises. Just as modern medicine has moved since the mid-20th century from a focus on acute treatment to a focus on the early intervention of chronic disease, economic policy also has room to move from a focus on recovery after a crisis to a focus on gradient management before a crisis. The theoretical basis of this move is the grammar of gradient of this book, and one example of a practical tool is the diagnostic engine this chapter has shown.
The establishment of preventive medicine required three conditions. First, the scientific establishment of the fact that the progression of disease is staged. Second, the development of objective indicators that can measure each stage. Third, the accumulation of clinical evidence that intervention at each stage makes a meaningful difference in the long-term prognosis. For preventive economics to be established, the same three conditions are needed. That the progression of an economic crisis is staged is shown by the five-stage reconstruction of this chapter, the measurement indicators of each stage are provided by the 59 gauges, and that intervention at each stage makes a difference is preliminarily shown by the case in which the Fed's six-month leading response in 2008 played a role in limiting the scale of the crisis.
This possibility is work that requires long-term accumulation, and this book does not claim the completion of that work. What this chapter does goes only as far as showing that preventive economics is theoretically possible, and that the grammar of gradient that forms the basis of that possibility is supported by measurement. On this foundation, the research that will accumulate going forward will concretize this possibility into an actual policy tool. This book provides the first coordinate of that.
The Progression of an Economic Crisis Seen Through the Five Stages
Organizing the discussion of this entire chapter according to this book's formula (accumulation of determinants, trigger, dual blockade and the collapse of flow, manifestation, collapse) yields the following.
Stage 1, the accumulation of determinants. Structural conditions over several years create the long-term worsening of gradient within the system. The gradual rise of household and corporate debt relative to GDP, the persistent divergence between asset prices and the income growth rate, the expansion of leverage in the financial sector, and the increased dependence on short-term foreign debt in the external sector are the typical factors of this stage. At this stage no indicator crosses the threshold, and the composite indices generally remain in the normal range.
Stage 2, the trigger. The worsening of a gauge begins on a particular axis, and the engine judges the blockade of a single axis for the first time. At this stage the other axes are still normal, and the system is absorbing the worsening of that axis with the room of the other axes. On the surface it is still not a crisis, but at the level of the gradient it is the point at which movement in one direction has already begun. Intervention at this stage is most effective and, at the same time, most difficult. Because the surface evidence is weak, the basis for policy resolve appears lacking.
Stage 3, dual blockade and the collapse of flow. Blockade forms simultaneously on two or more axes, and the engine hands down a system-level dual-blockade verdict. This point is the reference point that this chapter's engine identifies, and from here the system switches from recovery mode to survival mode. In the case of an endogenous crisis, this point precedes the actual crisis manifestation by an average of six months.
Stage 4, manifestation. Dual blockade surfaces and is recognized as an official crisis. The bankruptcy of major financial institutions, the plunge of stock prices, the sudden change of exchange rates, the surge of unemployment, and the emergency intervention of the central bank are the clinical signs of this stage. The response at this stage aims at damage limitation, and the best is to prevent additional chain reactions in a system in which collapse is already progressing.
Stage 5, collapse. As dual blockade spreads to multiple sectors, it becomes fixed as a system-level long-term recession. The resolution of bad debt, restructuring, the prolongation of unemployment, and the readjustment of debt are the content of this stage, and the time to reach recovery varies from several years to several decades. At this stage the natural recovery of the gradient is very slow, and without active policy intervention it is generally the case that the new normal settles at a level lower than the previous normal.
These five stages repeat in the same form in the progression of the body's chronic disease and the progression of an economic crisis. The concrete names of the factors differ, but the logical structure of the stages is the same. This is the place where the universality of the grammar this book has argued for is revealed most compactly.
The Last: The Same Grammar, Different Media
In closing this chapter, let me confirm once more the place this chapter occupies within the structure of this book as a whole. Chapters 9 and 10 read the grammar of gradient in the human body. That reading overlapped in many parts with the perspective of existing medicine, and the overlap was a basis for persuasiveness while at the same time an element that weakened the basis for novelty. This chapter was the testing ground that asked whether that reading holds outside medicine as well. The fact that the same grammar was confirmed with measured data in the economy, the complex system farthest from the human body, supports that the reading of the previous chapters is not confined to medicine but is the medical expression of a domain-independent physical law.
The media will differ. In the body it was calcium, and in the economy it is cash. In another system there will be yet another medium. But beneath the difference in media there is the same gradient, the same dual blockade, and the same nonlinear collapse. This sameness is the main body of the landscape this book has bound together under the name of gradient. That this main body exists was confirmed by measurement in four historical crises, and this is the sole achievement of this chapter, and on top of this achievement the remaining chapters of this book unfold.
The next chapter extends this confirmation to other academic domains. Until now we have seen the same grammar operating in the two domains of medicine and the economy. If the hypothesis of this book is correct, the number of domains should not remain at two. The extension to physics and chemistry, engineering, information systems, ecosystems, and social systems awaits, and the next chapter is the place that draws out the map of this extension. It will be the chapter in which the ambition of this book is revealed most greatly, and at the same time the chapter that honestly shows how much verification this book requires going forward.
참고문헌
- Schularick, M., & Taylor, A. M. (2012). Credit booms gone bust: Monetary policy, leverage cycles, and financial crises, 1870-2008. American Economic Review, 102(2), 1029-1061.
- Borio, C., & Drehmann, M. (2009). Assessing the risk of banking crises: revisited. BIS Quarterly Review, March 2009, 29-46.
- Kang, T. S., & Ma, G. (2009). Credit card lending distress in Korea in 2003. BIS Papers No. 46, 95-106.
- Financial Crisis Inquiry Commission. (2011). The Financial Crisis Inquiry Report: Final Report of the National Commission on the Causes of the Financial and Economic Crisis in the United States. Washington: U.S. Government Printing Office.
- Gorton, G. B., & Metrick, A. (2012). Getting up to speed on the financial crisis: A one-weekend-reader's guide. Journal of Economic Literature, 50(1), 128-150.
- Bernanke, B. S. (2015). The Courage to Act: A Memoir of a Crisis and Its Aftermath. W. W. Norton & Company.
- Edwards, F. R. (1999). Hedge funds and the collapse of Long-Term Capital Management. Journal of Economic Perspectives, 13(2), 189-210.
- Scheffer, M., Bascompte, J., Brock, W. A., Brovkin, V., Carpenter, S. R., Dakos, V., Held, H., van Nes, E. H., Rietkerk, M., & Sugihara, G. (2009). Early-warning signals for critical transitions. Nature, 461(7260), 53-59.
- Reinhart, C. M., & Rogoff, K. S. (2009). This Time Is Different: Eight Centuries of Financial Folly. Princeton University Press.
- Kim, K. (2006). The 1997-1998 Korean financial crisis: Causes, policy response, and lessons. IMF High-Level Seminar on Crisis Prevention in Emerging Markets, Singapore.
- Jordà, Ò., Schularick, M., & Taylor, A. M. (2013). When credit bites back. Journal of Money, Credit and Banking, 45(s2), 3-28.
- Minsky, H. P. (1986). Stabilizing an Unstable Economy. Yale University Press.
This article is the final installment (3/3) of the three-part series covering Chapter 11 of The Universal Law: Gradient. The body text follows the original manuscript and is provided for informational purposes.