應用動態財務分析於地震保險基金

Abstract

隨著國內住宅地震險的投保率提升，台灣地震保險基金之規模也逐漸增加。但現行的地震基金預測模型仍著重在負債面的風險管理，資產部分則維持靜態的投資策略。本研究以兼顧資產與負債面的動態財務分析為工具，模擬不同投資情境下基金未來淨值的變化。本文中債券報酬與利率走勢模擬主要採用CIR模型，股價模型假設為符合幾何布朗寧運動；負債部分則以古騰堡-里克特法則（Gutenberg-Richter Law）與波松分配為模擬依據。根據模擬結果，建議地震基金減少現金部位並增加債券持有比例，以使其在一定的破產機率下將資金做更有效的運用。

The accumulated asset of Taiwan Residence Earthquake Insurance Fund (TREIF) has largely increased in recent years due to the growth of take-up rate in Taiwan. The forecast model used by TREIF, however, focuses only on the management of liability side and simulates the fund asset by static investment strategy. In this paper, we apply dynamic financial analysis (DFA) technique to simulate the accumulated fund value under various customized scenarios. We use CIR model to simulate bond prices and interest rates, and assume that the stochastic process can be applied to our model where we use Geometric Brownian Motion to execute stock simulations. As for the liability model, it is based on Gutenberg-Richter Law and Poisson distribution. According to the simulation results, we suggest TREIF to reduce its cash position but increase the weight on bonds investment, so that the asset can be better allocated.