Hedge Portfolio Research Articles

Explainable machine learning for financial risk management: two practical use cases

We explore the potential of machine learning (ML) models applied in two financial risk management areas, i.e., credit risk management and financial risk hedging, through two practical use cases. This comparative study starts with the issue of explainability in complex ML models used in peer-to-peer lending for credit risk management. The first use case examines the limitations of using Kernel-SHAP with dependent features and evaluates different methods for estimating these dependencies using the Lending Club dataset. Our results suggest that accounting for feature dependence improves the understanding and robustness of prediction explanations. The second use case investigates a dynamic method for hedging foreign exchange risk in international equity portfolios, emphasizing the importance of accurate forecasts of currency returns. The analysis demonstrates that predictions yielded by ML models can significantly enhance the hedging of portfolios against currency risk. These findings highlight the transformative potential of advanced ML models in financial risk management, showcasing their capability to improve financial risk measurement and management. Further, our study outlines future research directions to advance this field.

On bank stock return spillovers in South Africa: Implications for portfolio hedging

The purpose of this study is to investigate the return spillovers among the five systemically important banks in South Africa. The study employs a time-varying parameter vector autoregression (TVP-VAR) framework utilising daily data between 2000 and 2024. A minimum return connectedness portfolio is constructed from the spillover indexes and compared with traditional portfolio optimisation methods in terms of risk reduction and reward-to-risk ratios. The research findings reveal that Capitec, the smallest bank in terms of assets, has weaker return connectedness with other banks while ABSA, FirstRand, Nedbank and Standard Bank are moderately connected. Moreover, the two largest banks in terms of assets, Standard and FirstRand, are more connected than any other pairwise connection in the network. During significant events such as national elections, the global financial crisis and the COVID-19 pandemic, system-wide connectedness increases.

Limiting out-of-sample performance of optimal unconstrained portfolios

This paper studies the out-of-sample Sharpe ratio of an unconstrained portfolio that combines the global minimum-variance with a hedge portfolio. Furthermore, we investigate how this ratio behaves as the number of risky assets and observations approaches infinity while maintaining a constant ratio. Under these conditions, it becomes possible to simultaneously account for estimation risk and achieve analytical tractability when optimizing the out-of-sample Sharpe ratio. This analysis also provides valuable insights to enhance out-of-sample performance in the finite case by introducing additional deterministic factors to the portfolio components.

Dynamic Returns Connectedness: Portfolio Hedging Implications During the COVID‐19 Pandemic and the Russia–Ukraine War

ABSTRACTWe apply a Time‐Varying Parameter Vector Auto Regressive (TVP‐VAR) connectedness approach on global assets to investigate time‐varying dynamic connectedness, portfolio performance, and hedge effectiveness during COVID‐19 and the Russia–Ukraine war. With increased connectedness and the changing role of energy and soft commodities during these two events, we find the minimum correlation (connectedness) portfolio performing better during COVID‐19 and the Russia–Ukraine war and that cumulative returns of portfolios are higher during COVID‐19. Additionally, we find varying (stable) hedge effectiveness of equity market indices and soft commodities (cryptocurrencies). This paper provides specific insights to investors about using optimal portfolios and hedging during pandemics and military conflicts.

Hedging global currency risk: A dynamic machine learning approach

We propose a dynamic method to hedge foreign exchange risk of international equity portfolios. The method is based on the currency return forecasts derived from a set of alternative machine learning models, built on the main factor components of the currency return time series. To illustrate our method, we take the perspective of a US portfolio manager investing in a global equity portfolio of developing and developed economies. The analysis of several model performance indicators, back-tested on data from 27 October 2008 to 30 December 2022, allows to conclude that accurate predictions of global currency factor returns, such as those obtained with non linear machine learning models, can improve currency risk hedging of global equity portfolios.

Deep learning for quadratic hedging in incomplete jump market

We propose a deep learning approach to study the minimal variance pricing and hedging problem in an incomplete jump diffusion market. It is based on a rigorous stochastic calculus derivation of the optimal hedging portfolio, optimal option price, and the corresponding equivalent martingale measure through the means of the Stackelberg game approach. A deep learning algorithm based on the combination of the feed-forward and LSTM neural networks is tested on three different market models, two of which are incomplete. In contrast, the complete market Black–Scholes model serves as a benchmark for the algorithm’s performance. The results that indicate the algorithm’s good performance are presented and discussed. In particular, we apply our results to the special incomplete market model studied by Merton and give a detailed comparison between our results based on the minimal variance principle and the results obtained by Merton based on a different pricing principle. Using deep learning, we find that the minimal variance principle leads to typically higher option prices than those deduced from the Merton principle. On the other hand, the minimal variance principle leads to lower losses than the Merton principle.

Clean energy market connectedness and investment strategies: New evidence from DCC-GARCH R[formula omitted] decomposed connectedness measures

In this study, we investigate the return propagation mechanism across four clean energy indices, namely, the NASDAQ OMX Green Economy Index, NASDAQ OMX Solar Energy Index, NASDAQ OMX Wind Energy Index, and NASDAQ OMX Geothermal Energy Index ranging from December 21st, 2010 until June 2nd, 2023 by using a novel DCC-GARCH-based R2 decomposed connectedness approach. This framework allows us to efficiently decompose dynamic conditional R2 goodness-of-fit measures into its decomposed components. Furthermore, we introduce the concept of minimum R2 decomposed connectedness portfolios and multivariate hedging portfolios. We find that the dynamic total connectedness is heterogeneous over time and economic-event dependent. In addition, the empirical results highlight that the NASDAQ OMX Green Economy Index is a net transmitter of shocks while all others are net receivers of shocks. Finally, we find that our proposed portfolio technique outperforms the NASDAQ OMX Green Economy Index as well as all alternative multivariate portfolio techniques regarding the hedging effectiveness score and the Sharpe ratio.

When you need them, they are not there: hedge capacities of cryptocurrencies disappear in downtrend markets

We provide empirical evidence supporting the economic reasoning behind the impossibility of diversification benefits and the hedge attributes of cryptocurrencies remaining in force during the downside trends observed in bearish financial markets. We employ a spillover connectedness model driven by time-varying parameter vector autoregressions on daily data covering January 2018 to November 2022 to analyze spillover transmissions between conventional and digital markets, focusing on the role of stablecoin issuances. We study the stock, bond, cryptocurrency, and stablecoin markets and find very high connectedness, which varies over time in response to up/down trends in financial markets. The results show that during financial turmoil, cryptocurrencies amplify downside risks rather than serve as diversifiers. In addition to risky assets from conventional financial markets, cryptocurrencies champion the transmission of spillovers to digital and conventional markets. In contrast, changes in stablecoin issuances produce few shocks because of their pegged prices, but they facilitate investors’ switch from volatile cryptos to more stable digital instruments; that is, we observe a phenomenon designated by us as the “flight-to-cryptosafety.” We draw insightful conclusions, provoking new thinking regarding portfolio hedge strategies that could potentially benefit investors when searching for less volatile investment performance.

Hedging an Efficient Portfolio against Expected Inflation Risk: An Applied Research in the Iraq Stock Exchange

This research aims to hedge the efficient portfolio of the investor against the expected inflation risk and to evaluate the extent of improvement in the quality of its performance. It has applied to an intentional sample of companies whose shares traded on the Iraq Stock Exchange, consisting of (37) companies, with (120) monthly observations for each company from 2012 to 2021. The simple ranking model of Elton et al. (1978) has been used to build the nominally efficient portfolio and the inflation-adjusted model of Chen and Moore (1985) to hedge a portfolio against expected inflation risk. The Sharpe, Treynor, Jensen, and M2 models have been used to evaluate the performance of portfolios. The research has reached several results. The most important of which is the presence of a big difference in the components of the efficient hedged portfolio compared to the nominal unhedged portfolio, in addition to the presence of a big difference in the amounts of investment weights between the two portfolios. The results of the analysis have also shown a significant improvement in the quality of the performance of the efficient portfolio that has hedged against the expected inflation risk compared to the unhedged nominal portfolio. The originality of the research and its scientific value lie in the fact that it is the first to adopt the inflation-adjusted model in hedging the efficient portfolios of investors against the inflation risk. In addition, it is the first knowledge contribution with empirical evidence about the efficient portfolio's hedging against that risk on the Iraq Stock Exchange. Paper type: Research Paper

Extreme downside risk connectedness and portfolio hedging among the G10 currencies

This study investigates the frequency connectedness among foreign exchange markets of G10 countries, focusing on tail risk and its implications for portfolio management. To do so, we use a novel framework that combines the Conditional Autoregressive Value-at-Risk (CAViaR) model with the novel time-varying frequency and quantile connectedness method developed by Chatziantoniou et al. (2022) based on Baruník and Křehlík (2018) and Ando et al. (2018) approach. A key value of this paper to the literature is the provision of fresh empirical evidence on the extreme downside linkages among the markets examined. From the average connectedness measures, the top shock transmitters within the network were EUR, NOK, AUD, SEK, and NZD, while the main shock receivers emerged to be JPY and CHF, followed by CAD and GBP. We note that events in the major funding markets (the Eurozone, Japan) have a higher impact on the participants in these same markets than in relatively small markets (New Zealand, Norway). From the dynamic connectedness results, the magnitude of connectedness for the entire sample period increased during the COVID-19 era, compared to the magnitude before the COVID-19 outbreak. The cumulative spillover also reveals that USDNOK is the vastest net transmitter of spillovers to other markets, including SEK, CHF, and AUD. However, the EUR is the largest net beneficiary followed by JPY and CAD. Findings from the time-varying extreme downside analysis suggest that throughout the period, SEK and NOK are the other currencies' strongest and most frequent net spillover shock emitters for the short-, medium-, and long-term dynamics. Currency portfolio implications are discussed.

The dynamic connectedness between collateralized loan obligations and major asset classes: a TVP-VAR approach and portfolio hedging strategies for investors

Motivated by the increasing demand for alternative assets that can contribute to reducing portfolio risk, this paper examines the volatility spillovers between collateralized loan obligations (CLOs) and various in-demand investment instruments, including equities, bonds, crude oil, commodities, gold, bitcoin, shipping and real estate. The applied methodology comprehends the time-varying parameter vector autoregressive (TVP-VAR) modification of the classical spillover approach, for the period from January 1, 2012, to August 31, 2023. The empirical findings show moderate levels of dynamic connectedness; albeit several external shocks strengthened the interconnection among the assets. Moreover, we compare the ability of CLOs for hedging, during the overall sample period and multiple subperiods, by estimating hedge ratios and optimal portfolio weights, in order to inform investors about feasible portfolio adjustments. Our results indicate that CLOs constitute an effective hedging tool, irrespective of the period covered, as the short position in their volatility provides high hedging effectiveness for investors holding long positions in the volatility of all the remaining assets.

Relative performance evaluation in spillover networks

In a multi-agent contracting problem, agents are linked in performance through two channels, effort spillover, governed by spillover network, and risk correlation, governed by risk structure. Assigning compensation weights on peers' performances can not only filter out common risks but also alter agent's incentives. We study how the network and risk structure jointly determine the optimal linear contract. First, the relative compensation sensitivity is determined by ratio of the dot product, between spillover vector and pure hedge portfolio, to unhedgeable risk. We then propose an index named informativeness along the spillover direction and argue that this index measures how precisely the principal can infer the agent's effort. By showing that both the implemented effort and induced welfare are increasing in the informativeness index, we argue that this index captures how central each agent is in this economy. Finally, results regarding relative sensitivities still hold under optimal contract with bounded compensation.

Firm-specific climate risk and market valuation

In alignment with the call in Engle et al. (2020) to improve the measurement of firm-level climate risk exposure, we explore usefulness of a firm-specific measure based on 10-K disclosures of climate risk. We find that our measure has incremental explanatory power for firm valuation over climate risk-related measures currently used in the literature. Further, our measure is broadly available, which extends the span of questions related to firm-specific climate risk that can be investigated. Corroborating its usefulness, we find that relative to hedge portfolios based upon the other measures, the return on a hedge portfolio that is long (short) on firms with high (low) values of our climate risk measure has the strongest negative association with a US-news based measure of climate change concerns. By exploiting the broad climate-related disclosures in 10-Ks, our measure provides a tool to better understand valuation implications of climate risk for firms.

Can Intraday Return Reversals Predict Future Stock Returns?

Based on the research of the US stock market, it is found that the fierce long-short battle between heterogeneous investors with different views is manifested as the high-frequency inversion between overnight returns and daytime returns over a period of time. This kind of long-short battle has an impact on asset prices over time, resulting in pricing anomaly on a cross-section. Based on the research of A-share market, this paper finds that after controlling the characteristics of scale, book-to-market ratio, reversal effect, profitability, turnover rate, etc., the abnormal frequency of opening high and closing low (AB_NR) is significantly positively correlated with the future return of stocks. The hedge portfolio constructed according to AB_NR has excess returns in time series (opening high and closing low anomaly) which cannot be explained by the existing asset pricing model. Further testing shows that the anomaly is significant in the period of high arbitrage cost and high investor sentiment. The empirical results show that the anomaly is due to mispricing rather than risk premium.

Hedging Effectiveness on the MISO Exchange

The purpose of this paper is to evaluate the effectiveness of four commonly used hedging techniques on wholesale electricity prices within the Midcontinent Independent System Operator (MISO) energy market. This analysis provides and in-depth assessment of hedging performance on one of the largest power markets in North America. In a break from extant literature, hourly time series are used to create hedge portfolios for each of the regional hubs within the MISO footprint. Each risk management strategy is employed on 24 hourly time series for each of the eight regional hubs within the MISO footprint. This results in a total of 768 hedged portfolios across 192 hourly time series. While the naïve approach tends to outperform the others in terms of variance reduction, none of the strategies examined meet the generally accepted standard for a highly effective hedge. In several instances, the hedged portfolio contains a higher amount of risk than the unhedged position. This may be explained by the unique characteristics of wholesale electricity and hedging pressure, particularly during peak demand hours.

Managing the shortfall risk of target date funds by overfunding

Abstract Is it possible to achieve almost riskless, nonfluctuating investment payoffs in the long run, at a fraction of the traditional funding requirement, using equity investments? What is their shortfall risk? These questions are motivated by the need to increase yields, while limiting the variability of investment results. We show how to use contingent claims, denominated in units of a stock index, to achieve an almost riskless investment outcome. To control the risk of the proposed hedge portfolios, we introduce an overfunded scheme and show its reliability using bootstrapping. Results show that a modest amount of overfunding is an effective risk-management approach that brings the probability of not achieving the target to less than 1 percent. Our results are based on the use of the minimal market model and a change of numeraire. Robustness tests support their validity under different market specifications.

Construction and Hedging of Equity Index Options Portfolios

Construction and Hedging of Equity Index Options Portfolios

Frequency Volatility Connectedness and Portfolio Hedging of U.S. Energy Commodities

Frequency Volatility Connectedness and Portfolio Hedging of U.S. Energy Commodities

Cobertura delta precisa de opciones europeas usando cálculo conformable

Objective: we aim to develop a method for delta hedging portfolios of European options based on the theory of conformable calculus which improves accuracy of risk management of listed options in a first-order approximation. Methodology: we allow the time derivative in the classic Black-Scholes-Merton model to have a fractional order0 ≤α≤ 1 and calculate the corresponding delta of a portfolio of listed options as a function of this conformableparameter. Results: applying this method to a portfolio consisting of eight European options on the SPX index, we find that conformable delta hedging offers more accurate average predictions than classical delta hedging. Limitations: this method is applicable for delta hedging in European options only. Originality: this is the first successful application of conformable calculus to delta hedging in European options. Conclusions: application of Conformable Calculus allows for a greater flexibility in the local approximation to price in delta-hedging European options and offers a new and more precise methodology to this objective.

인공신경망(ANN)을 이용한 옵션가격결정 및 딥헤징 (DH)에 대한 SNN 및 RNN 결과의 비교·분석

The purpose of this study is to present alternative methods to overcome market limitations such as the assumption of fixing the dynamics of underlying assets and market friction for the traditional Black-Scholes option pricing model. As for the research method of this paper, Adam is used as the gradient descent optimizer for the hedging model of the call option short portfolio using artificial neural network models SNN (simple neural network) and RNN (recurrent neural network) as analysis models, and deep neural network (deep neural network) is used as the hedging model. neural network) methodology was applied.
 The research results of this paper are as follows. Regarding the ATM call option price, the BS model showed 10.245, the risk neutral model showed 10.268, the SNN-DH model showed 11.834, and the RNN-DH model showed 11.882. Therefore, it appears that there is a slight difference in the call option price according to each analysis model. At this time, the DH analysis data sample is 100,000 (1×) training samples generated by Monte Carlo Simulation, and the number of testing samples is the training sample. 20% (20,000 pieces) was used, and the option payoff function is lambda(), which is -max(, 0). In addition, the option price and P&L (P&L) of SNN-DH and RNN-DH appear linearly on the basis, and the longer the remaining maturity, the closer the delta value of SNN-DH and BS to the distribution of the S-curve, and the closer the expiration date, the closer the two models are. Delta is densely distributed between 0 and 1. As for the total loss of delta hedging of the short call option position, SNN-DH was -0.0027 and RNN-DH was -0.0061, indicating that the total hedged profit and loss was close to 0.
 The implications of this study are to present DL techniques based on AI methods as an alternative way to overcome limitations such as fixing underlying asset dynamics and market friction of the traditional Black-Scholes model. In addition, it is an independent analysis model that considers valid robust tools by applying deep neural network algorithms for portfolio hedging problems. One limitation of the research is that it analyzed the model under the assumption of zero transaction costs, so future studies should consider this aspect.