تجزیه و تحلیل دینامیکی اثرات سیاست های وام مسکن در بازار املاک و مستغلات

The Korean government announced new policies to regulate mortgage lending that aim to decrease both the loan-to-value ratio and the debt-to-income ratio, in 2008. These policies were implemented on the expectation that they will control housing demand and stabilize house prices, focusing only on the current market status. However, it is difficult to analyze the effectiveness of these kinds of policies using an empirical approach. Consequently, a comprehensive and dynamic method is necessary for analyzing the effects of policies. This paper, therefore, develops an integrated and dynamic model for analyzing policy impacts. Using this model, the validity of mortgage-lending policies is assessed, and the interplay between various factors (including mortgage loans, housing prices, and demand) is examined. The model is also used to analyze unnoticed side effects in the real estate and financial markets. The dynamic analysis in this research can be applied not only to policy implication, but also to other dynamic fields such as project management, financial planning and demand analysis.

In the second half of 2008, the Korean government adopted policies aimed at invigorating mortgage lending and housing demand. The policies involved an increase in the maximum allowable loan-to-value ratio (LTV: amount borrowed as a percentage of the total appraised value of the property) and the maximum allowable debt-to-income ratio (DTI: the percentage of borrower’s monthly gross income used to repay debt). The result of these policies is a steady increase in the number of mortgage loans, as well as increased demand for housing and upward pressure on housing prices. In response to increases in housing prices, the government announced a comprehensive real estate program in 2008 to regulate mortgage lending by commercial banks. One policy included in the plan was to decrease the maximum allowable LTV and DTI in order to control housing demand and stabilize house prices with a short-term perspective, focusing only on immediate market impacts. However, housing market forecasting and its link to mortgage lending are difficult to assess empirically. Intuitive and empirical approaches can overlook the side effects of mortgage-lending policies on housing and real estate financial markets. Specifically, it is likely that house prices and demand are impacted by the policy to a lesser degree than expected. Mortgage loans from secondary lending agencies (e.g., nonmonetary institutions such as mutual savings banks and credit unions) become attractive because they are not restricted by the policies, and potential mortgage borrowers move to secondary lending agencies, reducing the impact of the policy. Furthermore, mortgage loans are an essential element of the real estate financial market, and for this reason, any policy that alters the mortgage market would induce a behavioral response from financial institutions. This change in behavior and its impact on the secondary financial market should also be considered. Many studies have addressed housing price market models and real estate policy analysis with a focus on several factors in the real estate market. Such studies have analyzed mortgage lender behaviors [1], bank lending restrictions [2], and existing futures contracts [3]. However, this research is limited by its empirical approach and in that it encompasses only a small number of potential factors. There is therefore a need for research that considers various perspectives as regards the real-world markets and systems at hand, and their interdependencies. To this end, many academics have attempted to develop integrated econometric models such as a dividend–price ratio model using time-series analysis [4], a rent–supply–demand model [5], impulse response functions for housing stocks derived from the VAR model [6], and a real house price growth model [7]. Although these econometric models are useful for point estimation, there are some limitations of such models in terms of separating the impact of correlated dependent variables, as well as difficulties in analyzing dynamic cause-and-effect relationships between model parameters. Therefore, as [8] demonstrated, in order to gain a comprehensive understanding of the real estate financial market and related policies, a comprehensive and dynamic approach is required. In this way, it is possible to forecast the sensitivity of the market to various policies, as well as the direction of the policy’s impact on various outcome parameters. In an effort to address this issue, this study outlines the development of system dynamics models replicating the Korean real estate and mortgage markets. These models are based on the fundamental principles and causal loops found in housing markets, and are set in motion by the economic activities of both consumers and financial agencies. Dynamics models represent numerous simultaneous interactions through a comprehensive and systematic approach elucidated by diagrams representing feedback loops. In addition, this research conducts a sensitivity analysis to validate, and estimate the impact of, the current and proposed policies of the Korean government as regards LTV and DTI ratios. The policy models are based on the price expectations of consumers as well as the law of supply and demand. Simulation models in this research are focused on the housing market and mortgage-lending agencies and their profit-seeking behavior. The secondary financial market for mortgage securitizations is not well established in Korea, and therefore it is difficult to gather data regarding its influence factors. For this reason the secondary market is not considered in this study.

Utilizing system dynamics modeling, this research attempted to analyze the Korean government’s mortgage-lending policies targeting housing demand. System dynamics analysis explains several positive and negative effects that could result from these policies. In these policy models, house price and demand are placed in balancing loops, such as demand and price-controlling loops. Further, a reinforcing loop that incorporates the expectation of future marginal profit is shown to have a great effect on house prices by increasing price volatility and growth tendencies. The regulation of excessive demand through mortgage-lending policies can assist in controlling this growth of house prices. Through sensitivity analysis, this research confirmed that LTV and DTI regulation has strong leverage in the housing market. However, the same regulations may cause potential mortgage borrowers to shift to the secondary market and thus offset the intended effect of regulations. Consequently, secondary lending agencies need to be regulated equally with primary agencies to maximize the effectiveness of government policies. Depending on the market structure and its responses, these policies could potentially have damaging effects rather than yielding the desired positive effect. That is, if the Korean government adopts policies that regulate mortgage lending excessively, the Korean secondary financial market may become depressed, such that capital supplied by mortgage investors is not sufficient for agencies to lend, contrary to the policy’s intention to increase lending. On the other hand, if the government decides to activate the secondary mortgage market, the generation of risky derivatives would need to be prevented through respective policies. Through quantitative analysis using system dynamics simulation, this research found influencing factors not established by empirical analysis of policy, and analyzed the dynamic impact of these factors. In this way, it was possible to expose the hidden side effects of mortgage-lending policies, and to propose comprehensive solutions and systematic approaches based on dynamic systems. Furthermore, various perspectives in the system dynamics analysis can provide decision-makers with evidence to support both policy regulation and deregulation. The results of the system dynamics simulation and sensitivity analysis tests the causal loops, validates the effects of various policies, and also estimates the extent of policy consequences. Dynamic analysis presented in this research can be also applied to project management such as financial planning and demand analysis in rapidly changing circumstances. Future research is required to expand simulation models such that it includes the behavior of secondary mortgage agencies and their investors. Also, the financial behavior of construction companies in relation to housing supply should be examined; doing so will more effectively highlight the impact of mortgage-lending policies on the market.