Knowledge that Transforms

The accuracy of dynamic stress-test capital models remains undocumented. Three methodologies: a CLASS-style approach, Bayesian model averaging, and a Lasso specification are used to forecast the performance of 14 large US banks during the financial crisis. Individual bank models are calibrated using bank historical data while regulatory models are calibrated using representative bank data. Representative bank model forecasts differ dramatically from the forecasts from bank-specific models and from actual outcomes. The Lasso methodology is most accurate, but its superiority may be sample-specific and is only apparent ex post. The results highlight the policy uncertainty inherent in regulatory stress tests.

Consistent with flight-to-quality, the coskewness between developed market sovereign bonds and global equity markets can help explain bond returns. A coskewness factor, defined as the return difference between the most negative coskewness bond portfolio and the most positive coskewness bond portfolio, carries a statistically significant unit price of risk of 43.8 basis points per month. Decreases in coskewness are also significantly associated with declines in bond yields. Coskewness declines during recessions when interest rates become low. Moreover, countries with a monetary policy regime which explicitly targets monetary aggregates have lower coskewness.

We explore the effects of bank bailouts on competition in the underwriting business. We exploit a sample of underwriters active in the European corporate bond markets from 2006 to 2013 and find that reputable underwriters suffer market share losses (of 12.43 %) after being bailed out. However, the market share of non-reputable underwriters is found to increase after a bail out. An exploration of the firm–bank underwriting matching reveals that the probability of being chosen as underwriter in a given deal decreases for reputable bailed-out banks, while it increases for non-reputable bailed-out banks. These results provide evidence of the effects of bailouts on underwriting competition. The economic impact depends on the ex-ante reputational capital of the bailed-out bank.

In this study, we examine whether political connections affect market reactions to rating changes. Using a new and comprehensive dataset on ten advanced emerging markets, we find that political connections attenuate negative reactions linked to rating downgrades. This effect is shaped by political ties involving influential politicians and concerns all sample firms and a subsample of nonfinancial companies. Moreover, we establish that state-owned companies or companies with significant state ownership do not benefit from the mitigation effect of political connections.

This paper examines the impact of geographic technological diversification on firm innovativeness. Our empirical study conducted on a panel of U.S. manufacturing companies shows that firms with geographic technological diversification are more innovative (as measured by both patents and citations) than firms without. Furthermore, we find that the positive relation between geographic technological diversification and firm innovation is driven by domestic technological diversification, while international technological diversification is negatively related to firm innovation. Our valuation tests further confirm the detrimental effect of international technological diversification on shareholder wealth.

The Financial Crisis began and accelerated in short-term money markets. One such market is the multi-trillion dollar sale-and-repurchase (“repo”) market, where prices show strong reactions during the crisis. The academic literature and policy community remain unsettled about the role of repo runs, because detailed data on repo quantities is not available. We provide quantity evidence of the run on repo through an examination of the collateral brought to emergency liquidity facilities of the Federal Reserve. We show that the magnitude of repo discounts (“haircuts”) on specific collateral is related to the likelihood of that collateral being brought to Fed facilities.

We consider consumer responsiveness to changes in the macroeconomic environment (i.e. product demand sensitivity) to be a systematic industry characteristic useful to studying how industrial diversification adds value. We argue that diversified firms, with the advantages of their internal capital market and the imperfectly correlated cash flows of their segments, will perform better than focused firms when the demand sensitivity of a product increases. Our empirical evidence reveals a significant diversification premium associated with an increase in sensitivity. Moreover, such premium predominantly exists during recessionary periods, disappearing in the presence of low coinsurance and inefficient use of the internal capital market. Our results are robust to alternative measures of sensitivity and performance metrics, different empirical model specifications, and to the concern of possibly biased estimations associated with the sample of diversified and focused firms.

Abundant references to threats to financial stability likely posed by systemic risk-taking in the euro area investment fund industry in an era of persistent low interest rates have not been accompanied by robust supportive empirical evidence. This is the first study that assesses the effects of euro area conventional and unconventional monetary policy shocks on coherent systemic risk measures applied to the investment fund industry. This research finds evidence of systemic risk-taking notably in the forms of contagion and increased vulnerability. It seems more material following conventional than unconventional monetary policy shocks. There is heterogeneity in the results, as the investment focus is important for assessing investment funds’ contribution to systemic risk. Fund types most affected by significant systemic risk-taking are bond funds, mixed funds and real estate funds. Some evidence of heightened vulnerability in equity funds is also present. Increase in leverage is part of the risk-taking mechanism. A key policy implication is that persistently accommodative monetary policy geared toward preserving price stability may face an intertemporal trade-off with financial stability, making it necessary to coordinate monetary and macroprudential policies.

We consider predicting systemic financial crises one to five years ahead using recurrent neural networks. We evaluate the prediction performance with the Jórda-Schularick-Taylor dataset, which includes the crisis dates and annual macroeconomic series of 17 countries over the period 1870−2016. Previous literature has found that simple neural net architectures are useful and outperform the traditional logistic regression model in predicting systemic financial crises. We show that such predictions can be significantly improved by making use of the Long-Short Term Memory (RNN-LSTM) and the Gated Recurrent Unit (RNN-GRU) neural nets. Behind the success is the recurrent networks’ ability to make more robust predictions from the time series data. The results remain robust after extensive sensitivity analysis.

European banks are exposed to a substantial amount of risky sovereign debt. “Missing capital” in the banking system resulting from the zero-risk weight exemption for European sovereign debt amplifies the co-movement between sovereign CDS spreads and facilitates cross-border crisis spillovers. Risks spill over from risky peripheral sovereigns to safer core countries, but not in the opposite direction nor for exposures to countries not exempted from risk-weighting. Unfunded non-domestic sovereign bond exposures primarily affect CDS spreads of non-GIIPS banks, while domestic sovereign-to-bank linkages are particularly important for GIIPS banks. Spillovers are attenuated when banks fund their sovereign bond exposures with capital.