A logistic regression point of view toward loss given default distribution estimation

We propose a new procedure to estimate the loss given default (LGD) distribution. Owing to the complicated shape of the LGD distribution, using a smooth density function as a driver to estimate it may result in a decline in model fit. To overcome this problem, we first apply the logistic regression to estimate the LGD cumulative distribution function. Then, we convert the result into the LGD distribution estimate. To implement the newly proposed estimation procedure, we collect a sample of 5269 defaulted debts from Moody’s Default and Recovery Database. A performance study is performed using 2000 pairs of in-sample and out-of-sample data-sets with different sizes that are randomly selected from the entire sample. Our results show that the newly proposed procedure has better and more robust performance than its alternatives, in the sense of yielding more accurate in-sample and out-of-sample LGD distribution estimates. Thus, it is useful for studying the LGD distribution.     

In-sample vs. out-of-sample tests of stock return predictability in the context of data mining

We undertake an extensive analysis of in-sample and out-of-sample tests of stock return predictability in an effort to better understand the nature of the empirical evidence on return predictability. We find that a number of financial variables appearing in the literature display both in-sample and out-of-sample predictive ability with respect to stock returns in annual data covering most of the twentieth century. In contrast to the extant literature, we demonstrate that there is little discrepancy between in-sample and out-of-sample test results once we employ out-of-sample tests with good power. While conventional wisdom holds that out-of-sample tests help guard against data mining, Inoue and Kilian [Inoue, A., Kilian, L., 2004. In-sample or out-of-sample tests of predictability: which one should we use? Econometric Reviews 23, 371–402.] recently argue that in-sample and out-of-sample tests are equally susceptible to data mining biases. Using a bootstrap procedure that explicitly accounts for data mining, we still find that certain financial variables display significant in-sample and out-of-sample predictive ability with respect to stock returns.     

The predictability of aggregate returns on commodity futures

This paper provides evidence that aggregate returns on commodity futures (without the returns on collateral) are predictable, both in-sample and out-of-sample, by various lagged variables from the stock market, bond market, macroeconomics, and the commodity market. Out of the 32 candidate predictors we consider, we find that investor sentiment is the best in-sample predictor of short-horizon returns, whereas the level and slope of the yield curve have much in-sample predictive power for long-horizon returns. We find that it is possible to forecast aggregate returns on commodity futures out-of-sample through several combination forecasts (the out-of-sample return forecasting R² is up to 1.65% at the monthly frequency).     

Out-of-sample equity premium predictability and sample split–invariant inference

For a comprehensive set of 21 equity premium predictors we find extreme variation in out-of-sample predictability results depending on the choice of the sample split date. To resolve this issue we propose reporting in graphical form the out-of-sample predictability criteria for every possible sample split, and two out-of-sample tests that are invariant to the sample split choice. We provide Monte Carlo evidence that our bootstrap-based inference is valid. The in-sample, and the sample split invariant out-of-sample mean and maximum tests that we propose, are in broad agreement. Finally we demonstrate how one can construct sample split invariant out-of-sample predictability tests that simultaneously control for data mining across many variables.     

The Explanatory Power and the Forecast Performance of Consumer Confidence Indices for Private Consumption Growth in Turkey

In this study, we assess empirically whether consumer confidence indices contain information about future private consumption growth in Turkey. To this end, we estimate models for quarterly total, durable, and nondurable consumption growth with and without sentiment indicators. We evaluate in-sample forecasts and one-step-ahead out-of-sample forecasts from recursive ordinary least squares (OLS) estimates. We also test permanent income and precautionary savings hypotheses with our results. We use overall indices of CNBC-e and Turkstat-CBRT Surveys, and Consumer Expectations Index (CEI) and Propensity to Consume Index (PCI) from the CNBC-e Survey as sentiment measures. We show that the lagged values of consumer sentiment have explanatory power on consumption growth. However, when used in conjunction with other economic variables such as real labor income, real stock price, real interest rate, and exchange rate, only CNBC-e for total consumption, and CBRT and PCI for nondurable consumption provide independent information about future consumption growth. Similarly, the gains in out-of-sample forecasts are observed under the absence of other variables and disappear in almost all cases following their inclusion to the estimations. Finally, we find no clear evidence for either precautionary savings motive or permanent income hypothesis on the link between consumer sentiment and future total consumption changes.     

恒生指数和沪深300股指期货套期保值效果对比研究

本文利用OLS、ECM、ECM-GARCH模型对沪深300股指期货和恒生指数期货的最优套期保值率进行了估算,并在风险最小化框架下对它们的套期保值效果进行了对比研究。结果发现:无论是哪种股指期货,不考虑期现货间存在的协整关系会使估算的最优套期保值率偏高,影响套期保值效果;其次是虽然在样本内外,沪深300股指期货的套期保值效果比恒生指数期货的好,但是沪深300股指期货套期保值效果的稳定性比恒生指数差。此时,ECM-GARCH和OLS模型分别为样本内外投资者利用沪深300指数期货进行套期保值时的最佳选择;对于恒生指数股指期货,最优模型是ECM。     

Limits to mutual funds' ability to rely on mean/variance optimization

Our evidence suggests that estimation error in the required statistics is an important factor inhibiting investors' ability to rely on mean/variance analysis. We compare the returns reported by mutual funds to the returns obtained from a mean/variance optimized portfolio of fund holdings. The results suggest that funds tend to outperform the optimized portfolio out-of-sample (when means/variances/covariances are unknown), but under-perform in-sample (when the required statistics in the optimization are known). Therefore, a popular assumption in asset pricing models that investors rely on a basic mean/variance analysis with known underlying statistics is likely to be grossly violated in the case of mutual funds.     

The Gilt-Equity Yield Ratio and the Predictability of UK and US Equity Returns

A number of financial variables have been shown to be effective in explaining the time-series of aggregate equity returns in both the UK and the US. These include, inter alia , the equity dividend yield, the spread between the yields on long and short government bonds, and the lagged equity return. Recently, however, the ratio between the long government bond yield and the equity dividend yield – the gilt-equity yield ratio – has emerged as a variable that has considerable explanatory power for UK equity returns. This paper compares the predictive ability of the gilt-equity yield ratio with these other variables for UK and US equity returns, providing evidence on both in-sample and out-of-sample performance. For UK monthly returns, it is shown that while the dividend yield has substantial in-sample explanatory power, this is not matched by out-of sample forecast accuracy. The gilt-equity yield ratio, in contrast, performs well both in-sample and out-of-sample. Although the predictability of US monthly equity returns is much lower than for the UK, a similar result emerges, with the gilt-equity yield ratio dominating the other variables in terms of both in-sample explanatory power and out-of-sample forecast performance. The gilt-equity yield ratio is also shown to have substantial predictive ability for long horizon returns.     

The Gilt-Equity Yield Ratio and the Predictability of UK and US Equity Returns

A number of financial variables have been shown to be effective in explaining the time-series of aggregate equity returns in both the UK and the US. These include, inter alia , the equity dividend yield, the spread between the yields on long and short government bonds, and the lagged equity return. Recently, however, the ratio between the long government bond yield and the equity dividend yield – the gilt-equity yield ratio – has emerged as a variable that has considerable explanatory power for UK equity returns. This paper compares the predictive ability of the gilt-equity yield ratio with these other variables for UK and US equity returns, providing evidence on both in-sample and out-of-sample performance. For UK monthly returns, it is shown that while the dividend yield has substantial in-sample explanatory power, this is not matched by out-of sample forecast accuracy. The gilt-equity yield ratio, in contrast, performs well both in-sample and out-of-sample. Although the predictability of US monthly equity returns is much lower than for the UK, a similar result emerges, with the gilt-equity yield ratio dominating the other variables in terms of both in-sample explanatory power and out-of-sample forecast performance. The gilt-equity yield ratio is also shown to have substantial predictive ability for long horizon returns.     

The output gap and stock returns: Do cyclical fluctuations predict portfolio returns?

This study examines whether the output gap leads portfolio stock returns. The paper conducts in-sample and out-of-sample forecasting of US stock portfolios formed on the basis of size and value. First, the paper finds cross-sectional portfolios are predictable in-sample by the output gap. Out-of-sample evidence is weaker but still generally supports the finding that the historical average benchmark can be beaten. Secondly and most importantly, we find mixed evidence that the Fama–French factor mimicking portfolios can be forecasted by the output gap. In particular, there is some out-of-sample predictability of the size effect (SMB) suggesting this lags the output gap. However, the output gap, a key business cycle indicator, cannot predict the value effect (HML) either in-sample or out-of-sample. Our results add to the prior literature which finds that the factor mimicking returns are related contemporaneously (Petkova and Zhang, 2005) or lead (Liew and Vassalou, 2000) economic indicators.     

Forecasting variance using stochastic volatility and GARCH

This paper estimates the conditional variance of daily Swedish OMX-index returns with stochastic volatility (SV) models and GARCH models and evaluates the in-sample performance as well as the out-of-sample forecasting ability of the models. Asymmetric as well as weekend/holiday effects are allowed for in the variance, and the assumption that errors are Gaussian is released. Evidence is found of a leverage effect and of higher variance during weekends. In both in-sample and out-of-sample comparisons SV models outperform GARCH models. However, while asymmetry, weekend/holiday effects and non-Gaussian errors are important for the in-sample fit, it is found that these factors do not contribute to enhancing the forecasting ability of the SV models.     

Portfolio performance of linear SDF models: an out-of-sample assessment

We evaluate linear stochastic discount factor models using an ex-post portfolio metric: the realized out-of-sample Sharpe ratio of mean–variance portfolios backed by alternative linear factor models. Using a sample of monthly US portfolio returns spanning the period 1968–2016, we find evidence that multifactor linear models have better empirical properties than the CAPM, not only when the cross-section of expected returns is evaluated in-sample, but also when they are used to inform one-month ahead portfolio selection. When we compare portfolios associated to multifactor models with mean–variance decisions implied by the single-factor CAPM, we document statistically significant differences in Sharpe ratios of up to 10 percent. Linear multifactor models that provide the best in-sample fit also yield the highest realized Sharpe ratios.     

Portfolio optimization under Expected Shortfall: contour maps of estimation error

The contour maps of the error of historical and parametric estimates of the global minimum risk for large random portfolios optimized under the Expected Shortfall (ES) risk measure are constructed. Similar maps for the VaR of the ES-optimized portfolio are also presented, along with results for the distribution of portfolio weights over the random samples and for the out-of-sample and in-sample estimates for ES. The contour maps allow one to quantitatively determine the sample size (the length of the time series) required by the optimization for a given number of different assets in the portfolio, at a given confidence level and a given level of relative estimation error. The necessary sample sizes invariably turn out to be unrealistically large for any reasonable choice of the number of assets and the confidence level. These results are obtained via analytical calculations based on methods borrowed from the statistical physics of random systems, supported by numerical simulations.     

Noise fit,estimation error and a Sharpe information criterion

When the in-sample Sharpe ratio is obtained by optimizing over a k-dimensional parameter space, it is a biased estimator for what can be expected on unseen data (out-of-sample). We derive (1) an unbiased estimator adjusting for both sources of bias: noise fit and estimation error. We then show (2) how to use the adjusted Sharpe ratio as model selection criterion analogously to the Akaike Information Criterion (AIC). Selecting a model with the highest adjusted Sharpe ratio selects the model with the highest estimated out-of-sample Sharpe ratio in the same way as selection by AIC does for the log-likelihood as a measure of fit.     

Are Smart Beta strategies suitable for hedge fund portfolios?

In the equity context different Smart Beta strategies (such as the equally weighted, global minimum variance, equal risk contribution and maximum diversified ratio) have been proposed as alternatives to the cap-weighted index. These new approaches have attracted the attention of equity managers as different empirical analyses demonstrate the superiority of these strategies with respect to cap-weighted and to strategies that consider only mean and variance. In this paper we focus our attention to hedge fund index portfolios and analyze if the results reported in the equity framework are still valid. We consider hedge fund index and equity portfolios, the approaches used for portfolio selection are the four ‘Smart Beta’ strategies, mean–variance and mean–variance–skewness. In the two latter approaches the Taylor approximation of a CARA expected utility function and the Polynomial Goal Programing (PGP) have been used. The obtained portfolios are analyzed in the in-sample as well as in the out-of-sample perspectives.     

Empirical Analysis of Business Cycle Synchronization and Shock Similarity Between Romania and the Eurozone

This study investigates benefits from a trading strategy based on the spillovers from international stock markets to the Polish emerging stock market. The analysis is conducted within the framework of factor and predictive generalized autoregressive conditional heteroskedasticity (GARCH) models of the Warsaw Stock Exchange main index, WIG. We apply an approach in which the mean equation of the GARCH model includes a deterministic part incorporating cross-markets linkages. Both in-sample and out-of-sample forecasts from the estimated models are calculated. The trading strategy is based on signals from the out-of-sample predictions. The models' performance and benefits from adopting such a strategy are evaluated using direction quality measures. Our results suggest that predictive models using cross-market linkages can produce superior out-of-sample forecasts compared to benchmarks.     

Should investors include commodities in their portfolios after all? New evidence

This paper investigates whether an investor is made better off by including commodities in a portfolio that consists of traditional asset classes. First, we revisit the posed question within an in-sample setting by employing mean-variance and non-mean-variance spanning tests. Then, we form optimal portfolios by taking into account the higher order moments of the portfolio returns distribution and evaluate their out-of-sample performance. Under the in-sample setting, we find that commodities are beneficial only to non-mean-variance investors. However, these benefits are not preserved out-of-sample. Our findings challenge the alleged diversification benefits of commodities and are robust across a number of performance evaluation measures, utility functions and datasets. The results hold even when transaction costs are considered and across various sub-periods. Not surprisingly, the only exception appears over the 2005-2008 unprecedented commodity boom period.     

Do commodities add value in multi-asset portfolios? An out-of-sample analysis for different investment strategies

An essential motive for investing in commodities is to enhance the performance of portfolios traditionally including only stocks and bonds. We analyze the in-sample and out-of-sample portfolio effects resulting from adding commodities to a stock-bond portfolio for commonly implemented asset allocation strategies such as equally- and strategically-weighted portfolios, risk-parity, minimum-variance as well as reward-to-risk timing, mean-variance and Black–Litterman. We analyze different commodity groups such as agricultural and livestock commodities that currently are critically discussed. The out-of-sample portfolio analysis indicates that the attainable benefits of commodities are much smaller than suggested by previous in-sample studies. Hence, in-sample analyses, such as spanning tests, might exaggerate the advantages of commodities. Moreover, the portfolio gains greatly vary between different types of commodities and sub-periods. While aggregate commodity indices, industrial and precious metals as well as energy improve the performance of a stock-bond portfolio for most asset allocation strategies, we hardly find positive portfolio effects for agriculture and livestock. Consequently, investments in food commodities are not essential for efficient asset allocation.     

Short-run forecasting of the euro-dollar exchange rate with economic fundamentals

We propose a fundamentals-based econometric model for the weekly changes in the euro-dollar rate with the distinctive feature of mixing economic variables quoted at different frequencies. The model obtains good in-sample fit and, more importantly, encouraging out-of-sample forecasting results at horizons ranging from one-week to one month. Specifically, we obtain statistically significant improvements upon the hard-to-beat random-walk model using traditional statistical measures of forecasting error at all horizons. Moreover, our model obtains a great improvement when we use the direction of change metric, which has more economic relevance than other loss measures. With this measure, our model performs much better at all forecasting horizons than a naive model that predicts the exchange rate as an equal chance to go up or down, with statistically significant improvements.     

Implementing statistical criteria to select return forecasting models: what do we learn?

Statisticl model selection criteria provide an informed choiceof the model with best external (i.e., out-of-sample) validity.Therefore they guard against overfitting ('data snooping').We implement several model selection criteria in order to verifyrecent evidence of predictability in excess stock returns andto determine which variables are valuable predictors. We confirmthe presence of in-sample predictability in an internationalstock market dataset, but discover that even the best predictionmodels have no out-of-sample forecasting power. The failureto detect out-of-sample predictability is not due to lack ofpower.