過去十年來，我國產險業的主權益急劇的上升，其速度遠超過同時期市場規模增加的速度。而保險公司業主權益大幅上升，很可能是因爲市場對保險公司償付能力的要求，但也有可能單純是保險公司在生產資源配置上的無效率，若增資是基於第一個原因，則隨資本的上升，保險公司的收益將隨之增加，反之，若只是因為生產資源配置的無效率，則增資將造成業主權益報酬率下降。 為探討這個問題，本文以我國十四家在民國81年前成立之產險公司為樣本，利用DEA無母數邊界分析法，求出最適資本投入量後，再依公司實際資本投入對目標資本投入之比率，用以衡量公司的資本投入是否呈現過多或不足的情形。實證結果發現公司投入過多的資本並未增加公司的收入效率，反面導致業主權益報酬率下降，換言之，資本之大幅增加並非完全是因為市場對保險公司清償能力的要求，而主要是因為生產要素配置不當所致。

The capital in the property-liability insurance industry in Taiwan has increased dramatically in the last decade. However, the size of the insurance market in terms of premium volume has not increased commensurately and the return on equity has been decreasing in the same period. Literatures suggest that committing more capital could be beneficial since lower insolvency risk due to more capital could improve contractual terms with policyholders. However, it is also possible that higher capital is simply the result of a/locative inefficiency of input resources. Therefore, the paper investigates the use of capital by insurers to provide evidence on whether the capital increase is the response to the demand for insurers' solvency or the allocative inefficiency of input resources. If the increased capital is to enhance insurer's financial strength, then more capital should lead to higher revenue efficiency and thus higher return on equity. Conversely, if the increased capital is simply the result of resources misallocation, then higher capital would have negative impact on revenue efficiency as well as return on equity. The data for the empirical test is the property-liability insurance companies in Taiwan from 1992 to 2001 when the capital of the industry increased dramatically. We use data envelopment analysis to derive the optimal capital an insurer should have, and use the optimal capital as a benchmark to measure the extent that the actual capital deviates from the optimal. The regressions are then conducted to investigate the relationship between the revenue efficiency and the extent of the deviation of the actual capital from the optimal capital, and the relationship between return on equity and the deviation of the actual capital from the optimal capital. The empirical evidences indicate that most insurers over-utilize capital, and the excessive capital has no significant impact on revenue efficiency. Thus, it suggests the overcapitalization represents allocative inefficiency in the industry. In addition, the over-utilization on capital a/so leads to lower return on equity, suggesting the allocative inefficiency penalizes the industry for holding too much capital. Overall, the paper provides empirical evidences showing that the higher capital in the last decade is not the industry's response on the demand for insolvency The result is consistent with the fact that policyholders do not seem to demand the solvency of property liability insurers given the existence of guaranty fund reduces the incentive of searching financially sound insurers, and further, no property-liability insurer has ever bankrupted so far. Instead, the empirical evidences suggest that higher capital is simply the result of allocative inefficiency of input resources. This leads to a series of issues surrounding the capital of the industry. Since the industry is generally overcapitalized, the cost of capital for the industry becomes an interesting topic. Also, how the industry utilizes the excessive capital is worth of further investigation since an efficient capital market would not tolerate the industry to keep excessive capital for too long while the return on capital deteriorates. Finally, the paper generally confirms that the industry is overcapitalized. Therefore, the problem facing the industry is no longer the deficiency of capital as indicated by Shiu and Wang (2002). Instead, the industry can now be characterized by the overcapitalization accompanied with sluggish growth of premium and deteriorating return on equity.