【正文】 elative to potential) and the writeoff ratio – although not vice versa. Shocks to output are found to have a significant impact on the writeoff ratio up to six quarters ahead, with the maximum impact occurring after one year. The bank writeoff ratio is also found to increase following shocks to nominal variables such as annual rate of retail price inflation and nominal interest rates although here the impact is only significant at a longer time horizon (after foursix quarters).The results suggest that following a shock to the output gap of the same magnitude experienced in the early 1990s, the aggregate writeoff ratio of UK banks would increase by around percentage points. This, of course, is nontrivial but would still only equate to one third of the major UK banks’ average pretax average annual profits over the past three years. Therefore, according to these results, the UK banking system as a whole would appear to be robust to large adverse macroeconomic the mon shorter (post1993) sample period, the corporate sector writeoff ratio is found to be twice as sensitive to output shocks as the aggregate writeoff ratio. Household writeoffs, on the other hand, seem to be more sensitive to changes in ine gearing than changes in economic activity per se, even though the economic impact is quite small. This result is confirmed when separate VARs are estimated using arrears on credit cards and on mortgages. While the impact of ine gearing on mortgage arrears is statistically and economically significant, credit card arrears do not respond to ine gearing shocks. This result may be attributable to a more than offsetting structural change in the household unsecured borrowing market over the more recent is also noticeable that the sensitivity of aggregate writeoffs to adverse output shocks is twice as large when the model is estimated in the post1993 period than in the full sample from 19882021. This could imply that it is misleading to use estimates from the more recent sample period to assess the impact on banks’ balance sheets of a repeat of the early 1990s’ recession. The result is interesting though since one might have expected, a priori, that the impact of a givensized output shock on writeoffs to be larger during recessions if there is a nonlinear impact during times of issue of differences between stress and tranquil periods is worth investigating in future work as well as incorporating, if possible, structural changes in the banking industry (such as improvements in risk management). Noheless, we believe, that this VAR approach is a useful addition to the suite of models used to assess the fragility of the banking system to adverse macroeconomic shocks. 以 VAR 方法對(duì)英國(guó)銀行的壓力測(cè)試 目 錄 摘要 總結(jié) 1 引入 2 文獻(xiàn)綜述 3 宏觀經(jīng)濟(jì)變量的選擇與估算 4 數(shù)據(jù) 5 總體結(jié)果與分類(lèi)結(jié)果 6 堅(jiān)挺性檢驗(yàn) 7 變量解構(gòu) 8 結(jié)論 附錄 參考文獻(xiàn) 摘要 該文采用一種新的方式對(duì)英國(guó)銀行系統(tǒng)進(jìn)行壓力測(cè)試。 Stress tests of UK banks using a VAR approach Contents Abstract Summary 1 Introduction 2 Literature review 3 Choice of macroeconomic variables and estimation 4 Data issues 5 Aggregate and sectoral results 6 Robustness checks 7 Variable deposition 8 Conclusions Appendix References Abstract: This paper adopts a new approach to stress testing the UK banking system. We attempt to account for the dynamics between banks’ writeoffs and key macroeconomic variables, through conditioning our stress test on the historical correlation between the variables and allowing for feedback effects from credit risk to the macroeconomy. In contrast to most existing empirical stress testing work, this paper uses a direct measure of banks’ fragility – the writeoff to loan ratio. We find that both UK banks’ total and corporate writeoffs are significantly related to deviations of output from an adver