Azamat Abdymomunov

Time variation of CAPM betas across market volatility regimes

We investigate time variation in Captial Asset Pricing Model (CAPM) betas for Book-to-Market (B/M) and momentum portfolios across stock market volatility regimes. For our analysis, we jointly model market and portfolio returns using a two-state Markov-switching process, with beta and the market risk premium allowed to vary between ‘low’ and ‘high’ volatility regimes. Our empirical findings suggest strong evidence of time variation in betas across volatility regimes in almost all the cases for which the unconditional CAPM can be rejected. Although the regime-switching conditional CAPM can still be rejected in many cases, the time-varying betas help explain portfolio returns much better than the unconditional CAPM, especially when market volatility is high.

Predicting output using the entire yield curve

Many papers find that the term spread of the term structure of government bond yields can predict future output growth. This paper extends that literature by exploiting information in the entire term structure of interest rates. I apply a dynamic version of the Nelson–Siegel yield curve model to jointly model real GDP growth and yield factors. I find that the dynamic yield curve model produces better out-of-sample forecasts of real GDP than those generated by the traditional term spread model. The main source of this improvement is in the dynamic approach to constructing forecasts versus the direct forecasting approach used in the term spread model. While I confirm the importance of the term spread as a predictor of future output, there is also a gain from using information in the curvature factor.

Regime-switching measure of systemic financial stress

In this paper, I propose an approach to measuring systemic financial stress. In particular, abrupt and large changes in the volatility of financial variables that represent the dynamics of the US financial sector are modeled with a joint regime-switching process, distinguishing “low” and “high” volatility regimes. I find that the joint “high” volatility regime for the TED spread, return on the NYSE index, and capital-weighted CDS spread for large banks is closely related to periods of financial stress. This result suggests that the probability of the joint high volatility regime of these financial variables can be considered as a measure of systemic financial stress. Copyright Springer-Verlag 2013

The Effects of Monetary Policy Regime Shifts on the Term Structure of Interest Rates

We investigate how the entire term structure of interest rates is influenced by changes in monetary policy regimes. To do so, we develop and estimate an arbitrage-free dynamic term-structure model which accounts for regime shifts in monetary policy and price of risk. Our results for US data from 1985 through 2008 indicate that (i) the Federal Reserve’s reaction to inflation has changed over time, switching between “active” and “passive” monetary policy regimes; (ii) on average, the term spread in the “active” regime was wider than in the “passive” regime; and (iii) the yields in the “active” regime were considerably more volatile than in the “passive” regime. The wider term spread in the “active” regime reflects higher term premia associated with a more sensitive response of the short-term interest rate to inflation. Additionally, our analysis suggests that the model fit improves substantially when we account for regime switching in monetary policy and price of risk.

Banking Sector Operational Losses and Macroeconomic Environment

We investigate the relationship between operational losses at large U.S. banking organizations and the macroeconomic environment. We find evidence of a negative relationship between macroeconomic growth and operational losses in two Basel II loss event type categories: Clients, Products, and Business Practices and Execution, Delivery, and Product Management. Losses in these two categories comprise about 90 percent of the total industry losses in our sample. Our analysis suggests that the negative correlation of losses in these two categories with macroeconomic growth is concentrated in the tails of loss distributions.

Forecasting the Term Structure of Government Bond Yields Using Credit Spreads and Structural Breaks

In this paper, we investigate whether credit spread curve information helps forecast the government bond yield curve and whether the joint dynamics of the government bond yields and credit spreads have structural changes. For this purpose, we use a joint dynamic Nelson-Siegel (DNS) model of the term structures of U.S. Treasury interest rates and credit spreads. We find that this joint model produces substantially more accurate out-of-sample Treasury yields forecasts compared with a standard DNS yield curve only model. We also find that the predictive gain from incorporating the credit spread curve information substantially increases if the joint model accounts for structural changes in the dynamics of yield and credit spread curves. In addition, our model incorporates a zero lower bound restriction ensuring that our predictions are economically plausible.

Approaches to Calculate Tail Quantiles of Compound Distributions

The literature proposes several alternatives for estimating compound distributions, which are widely used for risk quantification in the banking and insurance industries. In this paper, we evaluate the accuracy and time-efficiency of different approaches for estimating quantiles of compound distributions. We focus on three approaches: 1) Single Loss Approximation (SLA), 2) Perturbative Expansion Correction (PEC), and 3) Fast Fourier Transform (FFT). We find that the PEC approach is an accurate and time-efficient methodology for sub-exponential distributions, but only for quantiles greater than the 95th. The SLA performs similarly, but only for tail quantiles greater than the 99th. Neither the SLA nor PEC approaches are accurate for non-sub-exponential distributions. The FFT approach consistently gives the most accurate estimates for every distribution, however it is substantially less time-efficient than the PEC or SLA approaches. We recommend applying the FFT approach for estimating lower quantiles and non-sub-exponential distributions. We contribute to the literature by providing comprehensive guidance for selecting an appropriate approach for various parametric distributions used in the banking and insurance industries.

Operational Risk and Risk Management Quality: Evidence from U.S. Bank Holding Companies

This study documents the association between the quality of risk management practices and operational losses at large U.S. financial institutions. Using detailed supervisory data, we find that companies with weak risk management practices experience higher and more volatile operational losses. We also present evidence that the strength of risk management practices prior to the 2008–2009 Financial Crisis has explanatory power over losses during the crisis period. Our analysis provides new evidence of the importance of risk management practices for curtailing risk at financial institutions.

Tail Dependence and Systemic Risk in Operational Losses of the US Banking Industry

Using supervisory operational loss data of the US banking industry, we analyze dependence among operational losses within banks and across banks. We find evidence of relatively strong dependence among tail losses of different operational loss types within banks. Applying a copula framework, we estimate that the median correlation parameter for the key operational loss types is around 30% and exceeds 50% for some banks in our sample. Our results contrast with the previous literature that documents that correlation parameter estimates are in the range of 5–10% and typically do not exceed 20%. Further, we demonstrate significant model risk from not accounting for dependence among tail losses, resulting in material underestimation of operational risk. In addition, we investigate dependence of operational losses across banks. Using a copula framework, we estimate correlation parameters between losses of large banks in our sample to be 42% on average. This result suggests the presence of systemic risk from the simultaneous occurrence of operational tail losses in different large banks.