Determinants of the Nordic hedge fund performance

Kolisovas, Danielius; Giriūnienė, Gintarė; Baležentis, Tomas; Štreimikienė, Dalia; Morkūnas, Mangirdas Article Determinants of the Nordic hedge fund performance Journal of Business Economics and Management (JBEM) Provided in Cooperation with: Vilnius Gediminas Technical University (VILNIUS TECH) Suggested Citation: Kolisovas, Danielius; Giriūnienė, Gintarė; Baležentis, Tomas; Štreimikienė, Dalia; Morkūnas, Mangirdas (2022) : Determinants of the Nordic hedge fund performance, Journal of Business Economics and Management (JBEM), ISSN 2029-4433, Vilnius Gediminas Technical University, Vilnius, Vol. 23, Iss. 2, pp. 426-450, https://doi.org/10.3846/jbem.2022.16170 This Version is available at: https://hdl.handle.net/10419/317564 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/ Copyright © 2022 The Author(s). Published by Vilnius Gediminas Technical University This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons. org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. *Corresponding author. E-mail: [email protected] Journal of Business Economics and Management ISSN 1611-1699 / eISSN 2029-4433 2022 Volume 23 Issue 2: 426–450 https://doi.org/10.3846/jbem.2022.16170 DETERMINANTS OF THE NORDIC HEDGE FUND PERFORMANCE Danielius KOLISOVAS 1*, Gintarė GIRIŪNIENĖ 2, Tomas BALEŽENTIS 3, Dalia ŠTREIMIKIENĖ 4, Mangirdas MORKŪNAS 3 1Faculty of Public Governance and Business, Mykolas Romeris University, Vilnius, Lithuania 2Lithuanian Military Academy, Vilnius, Lithuania 3Faculty of Economics and Business Administration, Vilnius University, Vilnius, Lithuania 4Kaunas Faculty, Vilnius University, Kaunas, Lithuania Received 14 April 2021; accepted 26 August 2021 Abstract. Hedge funds have become an important part of the financial sector. The development of the hedge funds in the Nordic countries has been rather robust. Therefore, it is important to identify the determinants of the hedge fund performance and isolate the managerial performance, i.e., the Jensen’s alpha. To this end, this paper construct cross sectional and panel model for the Nordic hedge funds over 2005–2018. The Fung-Hsieh 8-factor model and other models are developed to identify the determinants of the Nordic hedge fund performance. The effects of crises of different nature (local to global, hedge funds to banking sector) are also tested. The results indicate that Nordic hedge funds are capable to generate positive alpha during the crisis even exceeding the alpha of the economically stable time periods. Keywords: hedge funds, Nordic countries, asset pricing, panel models, crisis variable, risk factors. JEL Classification: G23, C23. Introduction Although hedge funds account for less than 5% of all Collective Investment Undertakings (CIU) industry reaching the Assets under Management (AUM) of 3.25 trillion in the end of 2018, hedge funds have received a great deal of coverage by the researchers due to their focus on the high return. The high returns of the hedge funds derive from dealing with higher risk assets often using leverage and derivatives and special management techniques involving the combination of long/short, market neutral, relative value arbitrage strategies. Hedge fund strategies are more contrarian and do not follow the market trend allowing to generate the excess return (alpha) of 2.4%, while mutual funds are more inclined to do the Journal of Business Economics and Management, 2022, 23(2): 426–450 427 opposite (Grinblatt etal., 2020). The analysis of the financial data allows one to identify the underlying trends in the performance of business entities (Zhao etal., 2020). This can be applied to the financial institutions and instruments as well. Dixon etal. (2012) categorised the higher risk of the hedge fund investment into the following channels: – Credit channel with the collapse of Long-Term Capital Management (LTCM) in 1998 occurred as a consequence of worldwide crises in Asia and Russia. – Capital market concentration with tight connection with major investment banks as Bear Stern and Lehman Brothers which collapsed at the peak of financial crisis of 2007–2008. – Liquidity risk which causes the sudden price fall during the sell-off the assets. The assessment of the hedge fund performance is determined by balancing the investment instrument-based risk profile and the unique dynamic strategy implemented by the hedge fund manager. CAPM, APT and other conventional methods face challenges when measuring the hedge fund performance (Fung & Hsieh, 2004). The common drawback in trying to apply those methods for hedge fund pricing model is that they both rely on linear risk factors, which hedge funds managers can easily eliminate in their strategies by applying derivatives or option-like strategies. Considering hedge funds contain financial instruments with linear and non-linear payoffs, they may employ hedging/derivative instruments and they may employ very dynamic trading, Fung and Hsieh (2004), provided a new view on the hedge fund pricing. They identified five major risk components out of the set of the most common ones in the hedge fund universe. These authors also created five drivers of return within an asset class in relationship to those five components. These drivers were attributed to categories of value, system/trend following, system/opportunity, distressed style factors and global/macro. These portfolios of look back straddles are considered as the trend following risk factors which resemble the returns of trend following hedge funds, providing a key link between hedge fund returns and market assets. Edelman etal. (2012) supplemented the model with emerging market index that is now called Fung and Hsieh 8-factor model initially designed to analyse the surviving hedge funds and their generated positive alpha. Dewaele etal. (2015) and other researchers contributed to hedge funds assets pricing by analysing and proposing other investment-based non-linear factors (or alternative risk premia) summarized by Robertson (2018). The largest share of the hedge funds is managed by the managers in the United States which were used to develop those hedge fund performance measurement tools. Meanwhile the research of the smaller regions (e.g., the Nordic hedge funds reported AUM of USD 31.52 billion and do not exceed 1% of Global hedge funds) are rather episodic. Do etal. (2005) analyse Australian hedge funds; Van Dyk etal. (2014) compare US, Europe and Asian hedge funds; Oueslati and Hammami (2018) built models for Saudi Arabian and Malaysian hedge funds; Kanuri (2020) analyses Japanese hedge funds performance. The other special focus of our research is striving to expose whether the performance measurement models are able to figure the main determinants of the performance during the different economic cycle conditions. Researchers globally usually seek the strategies and the actions (e.g., adjusting the strategy, reducing the leverage) leading to the positive results 428 D. Kolisovas et al. Determinants of the Nordic hedge fund performance during the crisis and are interested in identifying the hedge funds outperforming the other investment classes during the crisis. Although Nordic hedge funds indices outperform the global rivals (e.g., NHX Composite index surpasses HFR index in the crisis financial of 2007–2008 by 8%) this phenomenal performance is left overshadowed in the lights of “green economy” or Chinese investors’ impact onto the Nordic market. Thus, we set the objective of this paper to develop the pricing models that would reveal the factors determining the performance of the Nordic hedge funds. The special focus shall be accommodated to the selected Nordic hedge fund managers’ performance during the crisis time over the period of 2005–2018. We also intentionally did not include the latest Nordic hedge fund performance time series as the Nordic hedge fund market underwent significant transformation: there was a sharp decrease of Nordic hedge fund reported AuM in 2019 and Covid-19 crisis in 2020 with stunning positive results presented by Kolisovas (2021b). The paper is organized as follows: Section 1 analyses the performance of the NHX hedge funds and the composition of the Nordic hedge fund industry. Section 2 identifies Asset Pricing models and the variables, which would best explain the pricing of the Nordic countries hedge funds. Section 3 presents our empirical Nordic hedge fund pricing modelling results based on the theoretical assumptions. Section 4 examines the robustness of the models. Section 5 discusses the results on the global and wider context. The conclusions summarise the main Nordic hedge fund performance determinants providing the tools to the investors and recommendations for further in-depth research. 1. Nordic hedge funds The Nordic countries are extensively represented by Hedge Nordic database collected and published by Nordic Business Media Aktiebolag– a private limited company registered in Stockholm, Sweden. Hedge Nordic provides the main Nordic Hedge Index Composite– NHX Composite– an equally-weighted index tracking the performance of the universe of Nordic hedge fund managers on a monthly basis. To be admitted to the NHX index, the fund, the fund manager, or the investment advisor should be domiciled in one of the Nordic countries (i.e., Finland, Sweden, Norway, Denmark, or Iceland), or the investment theme of the fund should be clearly Nordic or exhibit strong and convincing ties to the Nordic region. Hedge Nordic breaks down the Nordic hedge fund universe into five categories of hedge funds, which are also reflected by NHX strategy equally weighted indices: NHX Equities, NHX Fixed income, NHX CTA, NHX Multi-strategy and NHX Fund of funds. Nordic equities and fixed income hedge funds contain Nordic region equity or debt securities and their derivatives, Nordic CTA may also use the commodity and currency instruments. These specific focus on the Nordic investment instruments impose the assumption Nordic hedge funds returns are affected by Nordic specific systemic risk measures. The distribution of Nordic hedge fund across countries and strategies is presented in Table1. Hedge funds with a track record of longer than 100 months from 2005 to March 2018 were analysed making in total 72 funds out of 219 reported in 2018. Estrada (2021) concludes the series of reports claiming, that most hedge funds fail: their average life span is about 5 Journal of Business Economics and Management, 2022, 23(2): 426–450 429 years. However, out of 72 analysed Nordic hedge funds, 57 survived for over 10 years making Nordic as a long-livers region (Kolisovas, 2021a). Table 1. Nordic hedge funds distribution Country Hedge fund strategy Sweden Denmark Finland Norway All selected funds Nordic Equities 19 1 – 6 26 Nordic Fixed income 2 7 – 1 10 Nordic Multi-strategy 5 5 1 – 11 Nordic CTA 8 1 3 – 12 Nordic Fund of funds 10 – 3 – 13 Total 44 14 7 7 72 Considering the Nordic region factors are significant and following Hespeler and Loiacono (2015) hedge funds were split into two main categories: the funds with returns significantly correlated with the industry / sector and the funds with neutral character. The values ≥ 0.3 and ≤ –0.3 are considered of high or moderate degree correlated and the values in the range between –0.3 and 0.3 are said to be a small correlation close to industry / strategy neutral. Such split into categories shall also reduce the possible effect of heteroscedasticity. Having considered this rule within the same NHX strategies indices the funds can be distributed in the following pools (Table 2). Table 2. Hedge funds by correlation with corresponding NHX index Hedge fund strategy Total number of selected funds Positive correlation Low correlation Negative correlation Nordic Equities 26 18 8 – Nordic Fixed income 10 5 5 – Nordic Multi–strategy 11 9 2 – Nordic CTA 12 8 4 – Nordic Fund of funds 13 10 3 – Total 72 50 22 – In contradiction to Hespeler and Loiacono (2015) Nordic hedge fund data did not reveal any significant negative correlation which could lead to negative coefficients. However, 22 hedge funds out of 72 are showing signs of neutrality and their strategies may be more contrarian. Ardia and Boudt (2018) and Canepa etal. (2020) also proposed analysing hedge funds data split into quantiles by the performance. However, the preliminary analysis revealed, that splitting the funds into larger and wider quantiles does not make any significant effect on the models. Splitting into four to five quantiles leave smaller groups indivisible. Kolisovas (2021b) analysed Nordic equity hedge funds only splitting the hedge funds into quintiles by the volatility allowing to track the performance of the Nordic equity hedge funds right before and during the pandemic of Covid-19. 430 D. Kolisovas et al. Determinants of the Nordic hedge fund performance Following the performance characterisation by correlation with the index, as well as following Teo (2009) and Kang etal. (2020) there were constructed equally weighted three different portfolios of hedge funds based on their correlation with the index returns. These portfolios are named “Total”, “Correlated” (positively) and “Neutral”. Table3 reports the summary statistics of these portfolios constructed for each hedge fund strategy. Table 3. Summary statistics of NHX hedge fund portfolios (monthly) Hedge fund strategy Mean Std. Dev. Sharpe Skew Kurtosis Equities hedge funds Total 0.51% 1.26% 40.48% –0.3 4.56 Correlated 0.55% 1.75% 31.59% –0.49 4.93 Neutral 0.42% 0.71% 58.69% 0.34 5.39 NHX Equities 0.48% 1.50% 32.23% –0.73 4.62 Fixed income hedge funds Total 0.61% 0.91% 66.56% –0.89 9.33 Correlated 0.77% 1.57% 48.77% –2.02 1.69 Neutral 0.44% 0.57% 77.52% 0.6 3.51 NHX Fixed income 0.46% 1.35% 34.15% –3.84 27.05 Multistrategy hedge funds Total 0.48% 1.92% 25.07% –1.64 10.39 Correlated 0.44% 2.25% 19.51% –1.79 10.49 Neutral 0.69% 1.89% 36.53% 0.34 3.31 NHX Multistrategy 0.41% 1.11% 37.28% –0.51 3.54 CTA hedge funds Total 0.36% 2.45% 14.81% 0.12 2.96 Correlated 0.29% 3.47% 8.26% –0.12 3.04 Neutral 0.51% 2.26% 22.77% 0.23 3.95 NHX CTA 0.50% 1.95% 25.38% 0.21 3.29 Fund of funds hedge funds Total 0.26% 0.69% 37.02% –0.45 2.96 Correlated 0.26% 0.88% 30.27% –0.42 2.94 Neutral 0.23% 0.41% 56.90% 0.47 4.36 NHX Fund of funds 0.18% 0.90% 19.76% –0.92 5.87 For comparison purposes the summary statistics of the corresponding hedge fund indices is also presented next to the hedge fund portfolios statistics. The mean returns of “Total” portfolios of Nordic Equities, Nordic Fixed income, Nordic Multi-strategy and Nordic Fund of funds exceed the NHX index mean return what corresponds to the survivorship bias characteristic to most of hedge fund databases. Fung and Hsieh (2004), Hespeler and Loiacono (2015), Joenväärä etal. (2019) and others who made their analysis on the global scale Journal of Business Economics and Management, 2022, 23(2): 426–450 431 choose several or even all five global databases combined data to reduce the bias. Dou etal. (2021) detected the material survivorship bias of small databases similar to Nordic hedge funds which cannot be reduced. Alliance Bernstein’s (2012) research noted that alive funds reported 2.5% (or 0.2% monthly) better annual returns comparing with portfolios including the funds discontinued reporting. An example of NHX CTA index, however, is opposite: the mean return is 0.5% monthly, whereas Nordic CTA “Total” portfolio only is 0.36%. Taking in consideration hedge fund return reporting biases, we admit the paper represents the pricing models of the long living funds, not the entire Nordic hedge fund universe. 2. Methods and data The development of the hedge fund asset pricing models began from relying on the traditional risks (such as traditional asset value or the size of the company issuing the capital market instrument). These traditional risks are gathered in Conventional CAPM, APT or Fama-French three-factor model. Agarwal etal. (2018) noted that investors also started to recognise the exotic risks (such as momentum and option-like investments) usually used by hedge fund managers. Besides Fung and Hsieh 7-factor model initiated in 1997 and developed through 2004, Carhart (1997), Agarwal and Naik (2004), Capocci etal. (2005) and other worked on developing exotic risk factors, which would allow calculating more accurate hedge fund alpha. Agarwal etal. (2018) as well Robertson (2018) linked the risk factors with their relative price, considering exotic factors are more difficult and expensive to achieve. The factors were grouped by the cost and the difficulty from asset-based beta and increasing to smart beta, alternative beta and alpha factor on the top. Such an approach distinguishes alpha as the most contributing to the top results of the hedge funds and therefore need to be adequately rewarded. Agarwal etal. (2018) also argue, that CAPM model explains alpha better than more sophisticated models. Duanmu etal. (2020) analysed the contribution of various factors and found both beta dominant and alpha dominant hedge funds. Fung and Hsieh 8-factor model enhanced by Edelman etal. (2012) was used as the basis in our research. Fung and Hsieh 8-factor model covers a broad range of the factors including: equity risks consisting of equity market, size spread and emerging market factors; bond risk that consists of bond market and credit spread factors, and trend following risk that consists of bond trend-following, currency trend-following, and commodity trend-following factors. Following the arguments of Agarwal etal. (2018) these factors were diluted with other exotic risk factors of Fama and French 3-factor model Dewaele etal. (2015). The special attention was paid to commodity factors, which are more common in the hedge funds during the difficult conditions (Stafylas etal., 2018). The above-mentioned researchers were analysing global hedge fund databases (Barclay Hedge; Eureka Hedge; HFR; Morningstar and TASS) which altogether gather over 25 thousand single funds on aggregate. As the vast majority of hedge funds are registered or related to USA, global risk factors were dominant in those researches. Our research focuses on Nordic hedge funds investing into Nordic assets, therefore stock index related factor SPRF, stock size related factor RLSP and 10 years yield related bond factor TYRF were replaced with the corresponding Nordic national factors (Table4). 432 D. Kolisovas et al. Determinants of the Nordic hedge fund performance Table 4. Substituted risk factors Risk factor Descriptions Substituted factor OMXSRF Monthly OMX[1] Stockholm 30 Index (SE0000337842) minus monthly[2] Sweden 3-Month Bond Yield SPRF OMXCRF Monthly OMX Copenhagen Ex OMXC20 (DK0060487064) minus monthly Denmark 3-Months Bond Yield OMXHRF Monthly OMX Helsinki 25 (FI0008900212) minus monthly Finland 2-Years[3] Bond Yield OSEBXRF Monthly Oslo Børs[4] Benchmark Index minus monthly Norway 3-Months Bond Yield OMXN40FR Monthly OMX Nordic 40 Index (SE0001809476) minus Risk-free rate SizeSprS Monthly OMX Stockholm Small Cap minus monthly OMX Stockholm 30 Index (SE0000337842) RLSP SizeSprC Monthly OMX Copenhagen Small Cap minus monthly OMX Helsinki 25 (FI0008900212) SizeSprH Monthly OMX Helsinki Small Cap minus OMX Helsinki 25 (FI0008900212) SizeSprO Monthly Oslo GICS Small Caps minus monthly Oslo Børs Benchmark Index 10YSwed Sweden 10Y Bond Yield[5] minus Sweden 3-Month Bond Yield TYRF 10YDen Denmark 10Y Bond Yield minus Denmark 3-Months Bond Yield 10YFin Finland 10Y Bond Yield minus Finland 2-Years Bond Yield 10YNor Norway 10Y Bond Yield minus Norway 3-Months Bond Yield Notes: [1] OMX indexes are uploaded from www.nasdaqomxnordic.com/indexes website. [2] Here and where other yield (or rate) is reported annual, monthly rate is computed by division by 12. [3] 3-Months Yield is not available in Finland, however 2-Year bonds are not much different in other Nordic countries, therefore 3-Month yield was not extrapolated. [4] Oslo Børs indexes were uploaded from https://www.oslobors.no/ob_eng/markedsaktivitet website. [5] All yield information was uploaded from https://www.investing.com/rates-bonds website. Most of the Nordic hedge funds and risk factors presented in Table4 are accounted in the local currency (i.e., Sweden– SEK; Finland– EUR; Denmark– DKK and Norway– NOK). To eliminate the possible distortion of the models deriving from the currency exchange, hedge fund returns, and national factors were recalculated (discounted) using the local currency and USD spot market exchange rate deviation. Almeida etal. (2020) benefited from using panel data models when splitting the hedge funds into narrower pools by performance or interaction with the benchmark what we are also aiming in this research. The ultimate panel data models based on Fung and Hsieh 8-fac- tor model with replaced local variables were also supplemented with commodity and other variables as suggested by Stafylas etal. (2018), Racicot and Theoret (2019), Bohl etal. (2020), Mensi etal. (2021). The list of risk factors is presented in Table5. The table also presents the summary statistics and the augmented Dickey-Fuller (ADF) test to assure the values in the regression models are stationary. If the variables in the regression model are not stationary, then it can be proved that the standard assumptions for asymptotic analysis will not be valid. Journal of Business Economics and Management, 2022, 23(2): 426–450 433 Table 5. Final model risk factors with corresponding summary statistics Risk factor Mean Std. Dev. Shapre Skew Kurtosis ADF-p OMXSRF 0.37% 4.63% 7.90% –0.89 5.89 0.0000 OMXCRF 0.66% 5.25% 12.63% –1.29 5.91 0.0000 OMXHRF 0.40% 5.35% 7.39% –0.31 5.36 0.0000 OSEBXRF 0.64% 5.87% 10.81% –1.74 9.61 0.0000 SizeSprS 0.53% 3.71% 14.26% 0.56 5.80 0.0003 SizeSprC –0.54% 2.57% –21.12% 0.30 4.24 0.0000 SizeSprH 0.10% 3.63% 2.77% 0.64 5.30 0.0000 SizeSprO –0.14% 3.36% –4.25% 0.33 3.69 0.0000 10YSwed 0.11% 0.07% 159.73% 0.53 3.45 0.1758* 10YDen 0.08% 0.07% 120.03% –1.00 4.02 0.0834* 10YFin 0.10% 0.06% 176.45% 0.16 2.19 0.4776* 10YNor 0.07% 0.07% 101.47% –0.83 4.60 0.0216 BAATY 2.66% 0.85% 311.15% 1.59 6.89 0.0526* MSEMKFRF 0.59% 6.33% 9.34% –0.55 5.11 0.0000 PTFSBDRF –3.45% 14.48% –23.85% 1.30 4.86 0.0000 PTFSFXRF –1.46% 19.59% –7.46% 1.38 5.06 0.0000 PTFSCOMRF –0.68% 15.08% –4.48% 0.88 3.34 0.0000 PTFSIRRF –3.08% 29.54% –10.43% 4.34 29.02 0.0000 PTFSSTKRF –4.62% 14.41% –32.04% 1.72 8.66 0.0000 SMB[1] 0.08% 2.27% 3.43% 0.24 2.68 0.0000 HML –0.03% 2.55% –1.02% 0.00 5.66 0.0000 LIQ[2] 0.12% 3.55% 3.46% –0.39 4.52 0.0000 OCMDRWT[3] 0.33% 3.93% 8.41% –0.86 5.90 0.0000 GOLD[4] 0.70% 4.07% 17.27% 0.06 3.37 0.0000 COPPER 0.68% 8.03% 8.44% –0.16 6.91 0.0000 SILVER 0.90% 9.49% 9.46% 0.09 3.27 0.0000 BROIL 0.71% 9.64% 7.40% –0.16 4.49 0.0000 NGAS 0.26% 13.38% 1.91% 0.76 5.98 0.0000 COCOA 0.64% 7.48% 8.58% 0.03 3.14 0.0000 VIX[5] 0.33% 3.93% 8.41% –0.86 5.90 0.0000 Notes: [1] SMB and HML– Fama and French (1993) factors also used by Dewaele etal. (2015). [2] Liquidity factor of Pástor and Stambaugh is presented in https://faculty.chicagobooth.edu/-/media/ faculty/lubos-pastor/data/liq_data_1962_2019.txt. [3] Risk Weighted Enhanced Commodity Ex grains Index tracked by Ossiam ETF, includes 20 out of 24 components from the S&P GSCI TR. This strategy aims to offer volatility reduction and a better participation from all commodity sectors, especially by avoiding the concentration in the energy markets (weighting approximatively 70% of the S&P GSCI allocation). Source https://www.next-finance.net/ Ossiam-ETF-on-the-Risk-Weighted. [4] Commodity (Gold, Copper, Silver, Brent oil– BROIL, Natural gas– NGAS and Cocoa are represented by corresponding monthly spot price change less risk-free rate of return). 440 D. Kolisovas et al. Determinants of the Nordic hedge fund performance fixed income hedge funds are less focusing on commodities, however, the increase of adjusted R2 originates from the other specific factors except commodities spot prices. Adjusted R2 of Stafylas etal. (2018) models range from 0.76 to 0.89. The models of Swartz and Emami- Langroodi (2018) and Racicot and Theoret (2019), which both sought connection with VIX volatility index and commodities spot prices adjusted R2 vary from 0.2 to 0.7. They proved commodity related factors usually have positive expression, however, did not analyse CTA strategy hedge funds. Based on the regularity of the other specific factors between three panel data models (i.e., Total, Correlated and Neutral), the following regular significant factors shall be used in addition to Fung and Hsieh 8-factor model factors with national stock and bond factors, when pricing the Nordic hedge funds’ performance. Nordic equity hedge funds performance is additionally determined by: 1. HML is present in all three equity hedge fund models ranging from 0.0521 to 0.0692. HML factor was also significant in Stafylas etal. (2018) with values ranging between 0.069 and 0.10 during the growth or bull market times, while negative in the recession and bear market (ranging between –0.131 and –0.253). Swartz and Emami- Langroodi (2018) did not find significant connection between equity hedge funds and HML. 2. LIQ is present in “Total” and “Correlated” models. The factor is negative and range from –0.0412 to –0.0604. Liquidity factor impact on the equity portfolios was widely analysed by Pástor and Stambaugh (2003) who called it liquidity beta and in most of the analysed portfolios it was also negative with similar proportion. Although VIX volatility index is present in “Total” and “Correlated” models and the coefficient value range from –1.3060 to –1.5888, they are significantly different from the results of other studies. VIX index coefficients range from –0.021 to –0.034 in research of Stafylas etal. (2018) and range from –0.0149 to –0.0179 in the research of Swartz and Emami-Langroodi (2018) who only found significant connection of VIX volatility index with related value Volatility funds. These VIX coefficients of Stafylas etal. (2018) and Swartz and Emami-Langroodi Table 11. Nordic hedge fund panel data global and national models’ comparison Nordic hedge fund strategy Equity Fixed income CTA Multistrategy Fund of funds Adj. R2 Global factors (Total) 0.4353 0.4570 0.1370 0.5088 0.4782 Adj. R2 Other factors (Total) 0.4584 0.4938 0.1768 0.5214 0.5269 Adj. R2 Global factors (Correlated) 0.4940 0.5053 0.2062 0.5750 0.5000 Adj. R2 Other factors (Correlated) 0.5126 0.5295 0.2829 0.5812 0.5505 Adj. R2 Global factors (Neutral) 0.3015 0.4163 0.0519 0.2903 0.3131 Adj. R2 Other factors (Neutral) 0.3497 0.4696 0.0811 0.3652 0.3602 Journal of Business Economics and Management, 2022, 23(2): 426–450 441 (2018) are of the same range when comparing with VIX coefficient of –0.0195 presented in the multiple linear regression models compiled for NHX equity index (Table10). Other factors selected by the NHX equity hedge funds’ models are random in nature, therefore will not be included in further analysis of crisis impact on the pricing models either. Nordic fixed income hedge funds performance is additionally determined by LIQ factor, which is present in all three hedge fund models. The factor is negative and range from –0.0896 to –0.1260, which is two times higher than is Nordic hedge fund models providing Nordic fixed income hedge funds are more dependent on the changes in liquidity premiums. Other factors selected by the NHX fixed income hedge funds’ models are random in nature, therefore will not be included in further analysis of crisis impact on the pricing models. Nordic CTA hedge funds performance is additionally determined by: 1. SMB is present in “Total” and “Correlated” models ranging from –0.2407 to –0.3878. SMB factor was also significant equities in hedge funds models of Stafylas etal. (2018) ranging from 0.127 for small funds to 0.173 for funds with lockups. Swartz and Emami-Langroodi (2018) analysed various hedge funds strategies and the closest to CTA would be Energy Infrastructure funds (both are related with energy commodities) which present SMB coefficient of –0.0209, whereas other strategies present SMB coefficients in the positive ranges similar to Stafylas etal. (2018). SMB factor of –0.1448 present in the multiple linear regression models compiled for NHX CTA index (Table10). 2. GOLD is present in “Total” and “Correlated” models. GOLD ranging from 0.0091 to 0.0900 comparing with models of Billio etal. (2012), where GOLD coefficient takes a negative value of –0.09 for Short bias strategy and ranges from 0.03 in equity neutral to 0.16 in Emerging strategy hedge funds. Swartz and Emami-Langroodi (2018) only found significant connection of GOLD with Volatility portfolio with negative value of –0.06. Although none of aforementioned researchers analysed CTA strategy in particular, based on APT theory of Ross (1976) the presence of GOLD in CTA hedge funds pricing model raises the presumption there is actual investment into GOLD or XAU2 index in these funds. 3. SILVER is present in “Total” and “Correlated” models. SILVER is ranging from 0.0551 to 0.0975. Using of SILVER in hedge fund pricing models is rather uncommon by other researchers, however it shall not be discarded from the model, as in some interim calculations SILVER even prevailed the GOLD factor. SILVER factor also presented in the multiple linear regression models compiled for NHX CTA index (Table10) and amount 0.0745, which is close to the values of the panel data models. As with equities strategy, VIX volatility index is present in all models and the coefficient equal –2.8468, –6.0041 and 2.5290. Such values are significantly different from the results of other studies mentioned above and therefore shall not be included in the models. Other factors selected by the NHX CTA hedge funds’ models are random in nature, therefore will not be included in further analysis of crisis impact on the pricing models. All Nordic hedge fund pricing models with National factors and with aforementioned other specific factors were tested for cross-section Fixed effect presence. Breusch-Pagan LM 2 XAU index– Gold Bullion index monthly rate of return. 442 D. Kolisovas et al. Determinants of the Nordic hedge fund performance random effect, Hausman fixed effect and Breusch-Pagan LM cross-section dependence tests were applied for all models. Equities and fixed income all panel data models (i.e., “Total”, “Correlated” and “Neutral”) proved to be able to apply cross-sectional Fixed Effect after successful Cross Section Dependence diagnostic test of Breusch-Pagan (1980) Lagrange Multiplier (LM). This led to further increase for the adjusted R2 as presented in Table12. Table 12. Nordic panel data models with applied cross-sectional Fixed Effect Total Correlated Neutral Adj. R2 Equity Panel Least Squares 0.5373 0.5846 0.4892 Adj. R2 Equity Panel EGLS (Cross-section weights) 0.6887 0.7353 0.5387 Adj. R2 Fixed income Panel Least Squares 0.6603 0.7053 0.6702 Adj. R2 Fixed income Panel EGLS (Cross-section weights) 0.6622 0.7162 0.6884 The high level of adjusted R2 of the models with applied cross-sectional Fixed Effect provides the opportunity to analyse alpha of every single hedge fund within the pool. Adding the crisis dummy variables in all models (i.e., “Total”, “Correlated” and “Neutral”) for all Nordic hedge fund strategies yielded the following results: 1. Banking Crisis variable was significant in most of the models of fixed income, CTA, multi-strategy and fund of funds models, and was rejected by all equity models. 2. Global Drawdown variable was significant only for all CTA models and rejected in all other hedge fund strategies. 3. Global Crisis including Brexit period as a crisis was significant in all fixed income, most multi-strategy and most fund of funds models, and was rejected by equity and CTA models. 4. Global Crisis which does not include Brexit as a crisis period was also significant in all equity hedge fund models. All crisis related factor coefficients and their standard errors are presented in Table13. Table 13. Nordic hedge fund panel data crisis alpha Crisis Strategy / Models Equity Fixed income CTA Multistrategy Fund of funds Banking Crisis “Total” –0.0133 (0.0017) 0.0127 (0.0048) 0.0106 (0.0027) 0.0149 (0.0029) “Correlated” –0.0104 (0.0021) –0.0076 (0.0028) 0.0160 (0.0029) “Neutral” – 0.0113 (0.0031) 0.0209 (0.0057) 0.0198 (0.0070) – Global Drawdown “Total” – – 0.0167 (0.0037) – – “Correlated” – – 0.0112 (0.0046) – – “Neutral” – – 0.0180 (0.0047) 0.0119 (0.0058) – Journal of Business Economics and Management, 2022, 23(2): 426–450 443 Crisis Strategy / Models Equity Fixed income CTA Multistrategy Fund of funds Global Crisis including Brexit “Total” –0.0077 (0.0012) –0.0048 (0.0017) 0.0026 (0.0011) “Correlated” –0.0077 (0.0017) – – 0.0034 (0.0012) “Neutral” 0.0044 (0.0017) 0.0064 (0.0017) –0.0068 (0.0034) – Global Crisis excluding Brexit “Total” 0.0051 (0.0011) 0.0109 (0.0013) –0.0078 (0.0018) 0.0043 (0.0012) “Correlated” 0.0045 (0.0013) 0.0126 (0.0017) –0.0062 (0.0020) 0.0053 (0.0013) “Neutral” 0.0051 (0.0017) 0.0076 (0.0018) –0.0110 (0.0035) – Note: Standard errors in parentheses. Coefficients in boldface indicate statistical significance at the 99% one-tailed level (otherwise– 95%). All crisis alpha coefficients presented in Table13 are positive meaning the crisis alpha is greater than alfa in the quiet times. This can be explained by the hedge fund managers’ experience to prevent the value of the hedge fund from dropping to the level of market declines. Carhart (1997) found the connection between the good returns of the current years and the negative returns in the past. Berglund etal. (2018) relate the increasing in the events when the crisis erupts with monetary policy actions in times of crisis, and then possibly lose alpha as unconventional policies restrain volatility. Cao etal. (2018) analysed the alpha during the liquidity crisis including bankruptcy of Lehman Brothers on September 15, 2008 and discovered, hedge funds what used the leverage provided by Lehman Brothers were reduced opportunities to borrow. However other (non-Lehman Brothers related funds) outperformed the other financial institutions. Liang and Qiu (2019) made an in-depth analysis of leverage prior, during and after the crisis concluding, that among the other, the stronger fund governance is associated with higher hedge fund leverage. The long living Nordic hedge funds analysed in this paper are rather experienced and generate robust returns. Nordic region had not faced a major banking crisis in the research horizon, the borrowing possibility had not been extinct. This also imposes the explanation why all models generated positive crisis alpha. Increased alpha during the crisis times can also be explained by hedge fund managers’ skills in finding the opportunities and employing short strategies, which may not be the case with Bull markets widely discussed by Siegel (2005), who concluded that this additional alpha could be considered as additional liquidity premia or opacity of other risk factors. The positive crisis alpha, however, contradicts the alpha yielded by the models with split time series into crisis and non-crisis period. Research of Metzger and Shenai (2019) compiled separate models using financial crisis of 06/2007– 03/2009 and non-crisis period after the crisis until 01/2017. While alpha of the 9 500 hedge funds collected in Credit Suisse’s Hedge Index database calculated using Carhart’s Four factor model (Carhart, 1997) is dominantly negative during the crisis, it remains negative in some strategies even after the crisis. Al- End of Table 13 444 D. Kolisovas et al. Determinants of the Nordic hedge fund performance though such approach of splitting the time-series into crisis and non-crisis periods enables adjusting the hedge fund long / short strategy, the Long / Short strategy of models of Metzger and Shenai (2019) generate alpha of –0.0004 during the crisis and –0.0008 after the crisis. However, regardless some controversy between the models’ estimated alpha, there is a consensus between researchers (e.g., Sung etal., 2020; and Denk etal., 2020 among the latest) who agree the hedge funds have better results than other types of investment during the crisis period. This exceptional performance during the crisis suggests the skills of the hedge fund managers are well executed and are fairly reflected on the crisis alpha factors. 4. Robustness analysis The robustness of the main results is examined in this section. The models and selected factors, determining the Nordic hedge fund performance are selected using the stepwise regression technique and were mostly consistent through the different hedge fund pools within the same strategy (i.e., “Total”, “Correlated” and “Neutral”). Panel data models are usually facing heteroscedasticity and endogeneity problems. Racicot (2015) developed and widely used in panel data models instrumental variables (IV) estimation in the context of the generalized method of moments (GMM) introduced by Hansen (1982). Racicot etal. (2018) applied GMM method when testing liquidity factor of Pástor and Stambaugh (2003) adding to Fama and French 4-factors model which works when number of cross sections greater than time-series period. Due to low number of Nordic hedge funds we added the lagged dependent variable as the control variable for residual autocorrelation used by Racicot and Théoret (2016), Ardia and Boudt (2018). The main finding of the robustness analysis is that including the extra control variable does not remove the statistical significance of the specific factors we suggested in this paper. In Nordic equity, fixed income and CTA models we also avoided the endogeneity of the liquidity ratio, which Adrian etal. (2017) and Racicot etal. (2018) stressed in their research. The robustness of the models is also satisfied by seeking for the superior adjusted R2 significance factors, which were consistent through adding National factors and adding the other specific factors into the models. CTA models are least consistent; however, this is a common issue of Commodity and financial derivative related hedge funds observed by Stafylas etal. (2018) and others. The detailed robustness analysis results are available from the corresponding author upon request. The result of this research is also influenced by survivorship bias; therefore, the conclusions shall only be applicable to the selected funds. In order to assess how the results could change shall all funds including “dead” funds returns were included in the models, crisis impact factor was included into NHX strategy indices multiple linear regression models. The crisis alpha factors of multiple linear regression models are presented in Table14. In most of the cases the crisis factors were insignificant to the models, however, there are four models with negative crisis alpha what contradicts the crisis alpha obtained in the Nordic hedge fund panel data models (Table13). Such contradiction in crisis alphas may either be caused by: – Survivorship bias as “dead” hedge funds are included into the NHX strategy index at the time of their reporting, however not included into the pool of long living hedge funds, or Journal of Business Economics and Management, 2022, 23(2): 426–450 445 – The panel data models in this paper are based on National stock and bond factors, which may have different reaction to the global crisis factors, therefore may show the opposite results. Lastly, the models use long term time-series for compiling the factors. This diminishes the significance of some exotic risk factors characteristic to hedge funds as presented by Agarwal etal. (2018). Momentum of Fama and French (1993) and trend following factors of Fung and Hsieh (2001) may have different compensation during the crisis and quiet times, however models in this research only have generalised factor coefficients. 5. Discussion This research analyses and focuses more on hedge fund asset pricing models (e.g., Fung and Hsieh) using a combination of traditional and exotic risk factors as outlined by Agarwal etal. (2018). All those risk factors represent the systemic risk factors and are not connected with any individual fund manager’s investment decision-based attributes. Among those could be a degree of leverage, short positions, frequent trading, fee structure, etc. These attributes and risk factors could result in more fund specific model, however due to small number of Nordic hedge funds and limited information on the investment strategy elements does not allow analysing hedge funds on the micro level. The quantitative Nordic hedge fund return figures themselves present superior Nordic hedge fund performance comparing with the global indices, which raise further questions on what peculiarities of the Nordic hedge fund market or managers make this happen. The stability of the Nordic economies and a high focus on the regulation? Or the peculiarity of the Nordic temperament and attitude of the fund managers? Or ranking Nordic countries as top happiest countries in the world for over three years now, even regardless of the impact of the pandemic of Covid-19 (Helliwell etal., 2021). Extending the research horizon including the pandemic and other events shall divulge more consistent risk factors determining the performance of the hedge funds. There are separate studies the authors are undertaking to answer the specific questions on that specific period. There is, however, a significant return abnormality observed as of April 2021, which will raise even more questions about the performance measurement of the Nordic hedge funds and all over the Global hedge fund universe. Table 14. NHX strategy index crisis alpha Crisis NHX equity NHX fixed income NHX CTA NHX multistrategy NHX fund of funds Banking Crisis ***** Global Drawdown ––––– Global Crisis including Brexit –0.0063 (0.0014) – – –0.0029 (0.0012) –0.0022 (0.0010) Global Crisis excluding Brexit –0.0042 (0.0015) 0.0041 (0.0017) ––– Note: * Banking crisis was not included in the model, since Banking crisis factor is individual to each Nordic country, whereas NHX strategy indexes are general and do not assess the influence of each country individually. 446 D. Kolisovas et al. Determinants of the Nordic hedge fund performance As regards the hedge fund return data used, the analysis could be extended by taking in to account the other funds operating in different regions. However, due to the differences in consolidating the hedge funds into pools by strategies and corresponding NHX strategy indices, more alignment by strategy and sub-strategy is needed. More detailed results could also be obtained by applying methods used by other researches: Vector Autoregression, Generalized Autoregressive Conditional Heteroskedasticity, applying Granger causality test to transfer significant lagged variables into Homogeneous Panel data. Conclusions We analysed the performance of the Nordic hedge funds. The funds embarking on different strategies were covered in the analysis. The capital asset pricing models were established to identify the effects of the environment on the fund returns. Multiple econometric models were specified based on different theoretical premises. The analysis of performance of hedge fund pools selected for the research showed superior pooled hedge funds returns by 0.03–0.15% comparing with corresponding NXH strategy index. The selected hedge funds represent the long living hedge funds, whereas the indices returns are calculated based on all alive hedge funds during the reporting period. Due to its uniqueness as a strategy, CTA funds showed the opposite relationship between returns, where the highest returns are achieved by the hedge funds with lowest correlation with the NHX CTA index. The most significant factors of the return of the Nordic hedge funds are national stock and bond risk factors (whereas less importance is attached to global stock and bond factors). The effect of national risk factors on the pricing of Nordic CTA funds was negligible leading to a conclusion CTA funds returns are least determined by the local or national stock and bond risk factors. The other specific factors were considered in light of the connections and recommendations made by other researchers. Liquidity factor was consistent through equity and fixed income hedge funds with HML factor– in equity hedge fund pricing models. However, CTA hedge fund models yielded significant relationship on SMB, Gold and Silver prices deviations. VIX volatility index is present also in most of the hedge fund strategies, however the coefficient values were significantly different from other studies and therefore need further adjustments and interpretation. The utilisation of dummy factors to indicate various crisis periods allowed assessing the effect of the crisis on the alpha during the crisis and the quiet times. Equity hedge funds models significantly impacted by the Global crisis factor representing the financial crisis of 2007–2008, European dept crisis of 2009–2011 and continuation in 2012–2013, but not including the Brexit-related crisis. On the contrary, CTA funds models were significantly impacted by the Banking crisis and Global hedge fund drawdown periods. Fixed income models are somewhere in between: impacted both by Banking crisis and Global crisis periods, but not the Global hedge fund drawdown. The impact of all above mentioned crisis factors (also called a crisis alpha) persisted through all models showing the positive effect. It is important to note, those models are based on those Nordic hedge funds, which withstood more than one crisis and therefore already made Nordic region famous for long living hedge funds. Journal of Business Economics and Management, 2022, 23(2): 426–450 447 However, when taking the entire NHX strategy indices returns into the consideration and assessing the crisis impact, the controversy derives from “dead” hedge funds unaccounted in the pooled hedge funds models compiled in this paper. Analysis of the time-series covering such long analysis horizon and covering both quiet and crisis periods presupposes the main limitation of the research. The models used the diminished effects of the exotic risk factors that may vary in the quiet times and crisis period. The future research shall explore the possibility to include in the analysis “dead” Nordic hedge funds and dividing the time horizon into crisis and quiet time. References Adrian, T., Fleming, M., Shachar, O., & Vogt, E. (2017). Market liquidity after the financial crisis. Annual Review of Financial Economics, 9, 43–83. https://doi.org/10.1146/annurev-financial-110716-032325 Agarwal, V., & Naik, N. Y. (2004). Risks and portfolio decisions involving hedge funds. Review of Financial Studies, 17(1), 63–98. https://doi.org/10.1093/rfs/hhg044 Agarwal, V., Green, T. C., & Rena, H. (2018). Alpha or beta in the eye of the beholder: What drives hedge fund flows? Journal of Financial Economics, 127(3), 417–434. https://doi.org/10.1016/j.jfineco.2018.01.006 Alliance Bernstein. (2012). Rooting Out Biases in Hedge-Fund Data. https://blog.alliancebernstein.com/ post/en/2012/12/rooting-out-biases-in-hedge-fund-data Almeida, C., Ardison, K., & Garcia, R. (2020). Nonparametric assessment of hedge fund performance. Journal of Econometrics, 214(2), 349–378. https://doi.org/10.1016/j.jeconom.2019.08.002 Ardia, D., & Boudt, K. (2018). The peer performance ratios of hedge funds. Journal of Banking and Finance, 87(C), 351–368. https://doi.org/10.1016/j.jbankfin.2017.10.014 Babecký, J., Havráneka, T., Matˇej˚u, J., Rusnáka, M., Smídková, K., & Vašíček, B. (2014). Banking, debt, and currency crises in developed countries: Stylizedfacts and early warning indicators. Journal of Financial Stability, 15, 1–17. https://doi.org/10.1016/j.jfs.2014.07.001 Berglund, A., Guidolin, M. & Pedio, M. (2018). Monetary policy after the crisis: Threat or opportunity to hedge funds’ alphas? (BAFFI CAREFIN Centre Research Paper No.2018-84). https://doi.org/10.2139/ssrn.3225600 Bernard, C., Vanduffel, S., & Ye, J. (2019). A new efficiency test for ranking investments: Application to hedge fund performance. Economics Letters, 181, 203–207. https://doi.org/10.1016/j.econlet.2019.05.023 Bohl, M. T., Siklos, P. L., Stefan, M., & Wellenreuther, C. (2020). Price discovery in agricultural commodity markets: Do speculators contribute? Journal of Commodity Markets, 18, 100092. https://doi.org/10.1016/j.jcomm.2019.05.001 Canepa, A., González, M., & Skinner, F. S. (2020). Hedge fund strategies: A non-parametric analysis. International Review of Financial Analysis, 67, 101436. https://doi.org/10.1016/j.irfa.2019.101436 Cao, Ch., Liang, B., Lo, A. W., & Petrasek, L. (2018). Hedge fund holdings and stock market efficiency. The Review of Asset Pricing Studies, 8(1), 77–116. https://doi.org/10.1093/rapstu/rax015 Capocci, D., Corhay, A., & Hubner, G. (2005). Hedge fund performance and persistence in bull and bear markets. The European Journal of Finance, 11(5), 361–392. https://doi.org/10.1080/1351847042000286676 Carhart, M. M. (1997). On persistence in mutual fund performance. Journal of Finance, 52(1), 57–82. https://doi.org/10.1111/j.1540-6261.1997.tb03808.x 448 D. Kolisovas et al. Determinants of the Nordic hedge fund performance Chen, Y., Sun, J., Xu, W., & Jin, H. (2019). Empirical test of CAPM in Shanghai securities market. Finance, 9(1), 28–33. https://doi.org/10.12677/FIN.2019.91004 Denk, K., Djerroud, B., Seco, L., Shakourifar, M., & Zagso, R. (2020). Option-like properties in the distribution of hedge fund returns. Frontiers of Engineering Management, 7(2), 275–286. https://doi.org/10.1007/s42524-020-0095-3 Dewaele B., Pirotte, H., Tuchschmid, N., & Wallerstein, E. (2015). Assessing the performance of Funds of Hedge Funds (Working Papers CEB 11-041). Universite Libre de Bruxelles. Dixon, L., Clancy, N., & Kumar, K. B. (2012). Hedge funds and systemic risk. RAND Corporation, Santa Monica, CA. https://www.rand.org/pubs/monographs/MG1236.html Do, V., Faff, R., & Wickramanayake, J. (2005). An empirical analysis of hedge fund performance: The case of Australian hedge funds industry. Journal of Multinational Financial Management, 15(4–5), 377–393. https://doi.org/10.1016/j.mulfin.2005.04.006 Dou, W., Kogan, L., & Wu, W. (2021). Common fund flows: Flow hedging and factor pricing (Jacobs Levy Equity Management Center for Quantitative Financial Research Paper). https://doi.org/10.2139/ssrn.3543675 Duanmu, J., Li, Y., & Malakhov, A. (2020). Capturing hedge fund risk factor exposures: Hedge fund return replication with ETFs. The Financial Review, 55(3), 405–431. https://doi.org/10.1111/fire.12221 Edelman, D., Fung, W., Hsieh, D. A., & Naik, N. Y. (2012). Funds of hedge funds: Performance, risk and capital formation 2005 to 2010. Financial Markets and Portfolio Management, 26(1), 87–108. https://doi.org/10.1007/s11408-011-0180-z Estrada, J. (2021). No hedge funds, no cry. https://doi.org/10.2139/ssrn.3807815 Evestment. (2018) Hedge fund exposure and tail risk industry report. https://www.evestment.com/resources/research-reports Fama, E. F., & French, K. R. (1993). Common risk factors in the returns on stocks and bonds. Journal of Financial Economics, 33(1), 3–56. https://doi.org/10.1016/0304-405X(93)90023-5 Fung, W., & Hsieh, D. A. (1997). Empirical characteristics of dynamic trading strategies: The case of hedge funds. Review of Financial Studies, 10(2), 275–302. https://doi.org/10.1093/rfs/10.2.275 Fung, W., & Hsieh, D. A. (2001). The Risk in Hedge Fund Strategies: Theory and Evidence from Trend Followers. Review of Financial Studies, 14(2), 313–341. https://doi.org/10.1093/rfs/14.2.313 Fung, W., & Hsieh, D. A. (2004). Hedge fund benchmarks: A risk-based approach. Financial Analysts Journal, 60(5), 65–80. https://doi.org/10.2469/faj.v60.n5.2657 Gregoriou, G. N., Racicot, F. É., & Théoret, R. (2021). The response of hedge fund tail risk to macroeconomic shocks: A nonlinear VAR approach. Economic Modelling, 94, 843–872. https://doi.org/10.1016/j.econmod.2020.02.025 Grinblatt, M., Jostova, G., Petrasek, L., & Philipov, A. (2020). Style and skill: Hedge funds, mutual funds, and momentum. Management Science, 66(12), 5505–5531. https://doi.org/10.1287/mnsc.2019.3433 Hansen, L. (1982). Large sample properties of generalized method of moments estimators. Econometrica, 50(3), 1029–1054. https://doi.org/10.2307/1912775 Helliwell, J. F., Layard, R., Sachs, J. D., De Neve, J. E., Aknin, L. B., & Wang, S. (2021). World Happiness Report. United Nations Program. https://happiness-report.s3.amazonaws.com/2021/WHR+21.pdf Hespeler, F., & Loiacono, G. (2015). Monitoring systemic risk in the hedge fund sector (ESMA Working Paper, 2). https://www.esma.europa.eu/sites/default/files/library/esma_wp-2015-2_working_pa- per_no.2_2015_monitoring_systemic_risk_in_the_hedge_fund_sector.pdf Joenväärä, J., Kauppila, M., Kosowski, R., & Tolonen, P. (2019). Hedge fund performance: Are stylized facts sensitive to which database one uses? (Critical Finance Review). https://ssrn.com/abstract=3359325 Journal of Business Economics and Management, 2022, 23(2): 426–450 449 Kang, B. U., Kim, J. M., Palmon, O., & Zhong, Z. (2020). Are college education and job experience complements or substitutes? Evidence from hedge fund portfolio performance. Review of Quantitative Finance and Accounting, 54(4), 1247–1278. https://doi.org/10.1007/s11156-019-00824-5 Kanuri, S. (2020). Hedge fund performance in Japan. Review of Pacific Basin Financial Markets and Policies, 23(3), 2050023. https://doi.org/10.1142/S021909152050023X Kolisovas, D. (2021a). Secrets of long livers: Crisis alpha. In Nordic hedge fund industry report 2021 (pp.48–53). https://hedgenordic.com/wp-content/uploads/2021/03/HNIR2021.pdf Kolisovas, D. (2021b). In the Face of COVID-19: Unusual crisis performance, Finding Alpha in Equities, (pp. 38–43). https://hedgenordic.com/wp-content/uploads/2021/05/Equities_2021.pdf Liang, B., & Qiu, L. (2019). Hedge fund leverage: 2002–2017. European Financial Management, 25(4), 908–941. https://doi.org/10.1111/eufm.12202 Mensi, W., Shafiullah, M., Vo, X. V., & Kang, S. H. (2021). Volatility spillovers between strategic commodity futures and stock markets and portfolio implications: Evidence from developed and emerging economies. Resources Policy, 71, 102002. https://doi.org/10.1016/j.resourpol.2021.102002 Metzger, N., & Shenai, V. (2019). Hedge fund performance during and after the crisis: A comparative analysis of strategies 2007–2017. International Journal of Financial Studies, 7(1), 15. https://doi.org/10.3390/ijfs7010015 Oueslati, A., & Hammami, Y. (2018). Forecasting stock returns in Saudi Arabia and Malaysia. Review of Accounting and Finance, 17(2), 259–279. https://doi.org/10.1108/RAF-05-2017-0089 Pástor, L., & Stambaugh, R. F. (2003). Liquidity risk and expected stock returns. Journal of Political Economy, 111(3), 642–685. https://doi.org/10.1086/374184 Racicot, F. E. (2015). Engineering robust instruments for GMM estimation of panel data regression models with errors in variables: a note. Applied Economics, 47(10), 981–989. https://doi.org/10.1080/00036846.2014.985373 Racicot, F. E., & Théoret, R. (2016). Macroeconomic shocks, forward-looking dynamics, and the behavior of hedge funds. Journal of Banking & Finance, 62, 41–61. https://doi.org/10.1016/j.jbankfin.2015.10.004 Racicot, F. É., & Théoret, R. (2019). Hedge fund return higher moments over the business cycle. Economic Modelling, 78, 73–97. https://doi.org/10.1016/j.econmod.2018.08.016 Racicot, F. E., Rentz, W. F., & Théoret, R. (2018). Testing the new Fama and French factors with illiquidity: A panel data investigation. Finance, 39(3), 45–102. https://doi.org/10.3917/fina.393.0045 Racicot, F. E., Théoret, R., & Gregoriou, G. N. (2021). The response of hedge fund higher moments risk to macroeconomic and illiquidity shocks. International Review of Economics and Finance, 73, 289–318. https://doi.org/10.1016/j.iref.2020.12.004 Robertson, J. (2018). Replica localization in East Asia: The case of the Asian hedge fund industry. Globalizations, 15(3), 407–421. https://doi.org/10.1080/14747731.2018.1424286 Ross, S. A. (1976). The arbitrage theory of capital asset pricing. Journal of Economic Theory, 13(3), 341–360. https://doi.org/10.1016/0022-0531(76)90046-6 Siegel, J. J. (2005). The Future for Investors: Why the tried and the true triumph over the bold and the new. Crown Publishing. Stafylas, D., Anderson, K., & Uddin, M. (2018). Hedge fund performance attribution under various market conditions. International Review of Financial Analysis, 56, 221–237. https://doi.org/10.1016/j.irfa.2018.01.006 Sung, S., Chun, D., Cho, H., & Ryu, D. (2020). Hedge fund market runs during financial crises. Economic Research-Ekonomska Istraživanja, 34(1), 266–291. https://doi.org/10.1080/1331677X.2020.1782245