Bài viết Khung cơ sở quản trị rủi ro trong hoạt động của các quỹ phòng hộ (Hedge fund) phân tích hai cách tiếp cận trong việc ghi nhận lợi nhuận phi tuyến tính của các quỹ phòng hộ. Phương pháp đầu tiên cung cấp một khuôn khổ phù hợp sử dụng cho việc phân tích các phong cách đầu tư năng động của các quỹ phòng hộ.
QUẢN TRỊ NGÂN HÀNG & DOANH NGHIỆP Khung sở quản trị rủi ro hoạt động quỹ phòng hộ (Hedge fund) Nguyễn Diệu Hương Ngày nhận: 13/04/2017 Ngày nhận sửa: 17/04/2017 Ngày duyệt đăng: 17/04/2017 Bài viết phân tích hai cách tiếp cận việc ghi nhận lợi nhuận phi tuyến tính quỹ phịng hộ Phương pháp cung cấp khn khổ phù hợp sử dụng cho việc phân tích phong cách đầu tư động quỹ phòng hộ Phương pháp cịn lại giải thích tính phi tuyến tính lợi nhuận quỹ phịng hộ mối quan hệ theo hợp đồng quỹ phòng hộ với nhà đầu tư nhà môi giới Ảnh hưởng mối quan hệ thể thông qua giá xuống (short position) quỹ phòng hộ hai hợp đồng quyền chọn: quyền chọn tài trợ quyền chọn mua lại Bài viết đề xuất hướng dẫn quy trình quản lý rủi ro cho nhà quản lý quỹ phịng hộ Họ đạt mức địn bẩy tối ưu nhằm thu lợi nhuận tối đa mà không làm cho mức tiền hạ xuống mức thấp Tác động phân tích nhà đầu tư nhà hoạch định sách trình bày Từ khóa: Quỹ phịng hộ, quản trị rủi ro, quyền chọn tài trợ, quyền chọn mua lại, lợi nhuận phi tuyến tính Introduction Hedge funds have a lot of freedom of choosing asset classes and strategies to generate returns including derivative, short sales and leverage (David Stowell, 2010) Those advantages of hedge funds make their risk management process become much more complex and different from other portfolios While Sharpe’s regression (1992) is a useful method to estimate risk exposure of many portfolios, the R2 of this regression over hedge fund returns is very low (Fung and Hsieh, 1997) This result indicated that there are nonlinearities in hedge fund returns ince the first hedge fund in 1949, there have been a dramatically development in hedge funds industry, especially in the past two decades (Hull, 2012) While in good states, hedge funds perform very well and gross a lot of money, the market position for hedge funds industry fails significantly in financial distress Some researchers agree that hedge funds’ performance is not persistent over time (Koh et la., 2003) © Học viện Ngân hàng ISSN 1859 - 011X 54 Tạp chí Khoa học & Đào tạo Ngân hàng Số 179- Tháng 2017 This paper attempts to answer the question what is reason for the nonlinearities in hedge funds returns and how to measure their risk exposure The answer for this question is multifaceted Firstly, the flexibility of hedge funds on choosing different asset classes and dynamic strategies makes their returns correlated with some “style factors” In another word, hedge fund returns depend on not only the assets selection but also the strategies used to generate returns from those assets Secondly, because of using dynamic strategies, hedge fund returns are also correlated with some “abnormal return factors” such as Fama and French’s (1993) value factor and size factor or Carhart’s (1997) momentum factor Thirdly, it is the fact that in bad states of the world, hedge fund investors might withdraw their money or prime brokers might want to raise the requirement margin or get their credit back Those actions will reduce the liquidity of hedge funds and make them have to deleverage to meet those requirements Two risk factors above have the same effects to hedge funds that short two valuable options: funding option and redemption option The reason for nonlinearities in hedge fund returns is blamed for the presence of those two options In this paper, we will introduce two risk management perspectives of hedge funds Both perspectives concentrate on nonlinearities of hedge fund returns While one perspective focuses on the nonlinearities arising from dynamic portfolio strategies followed by hedge funds, the other perspective introduces significant nonlinearities that are captured by two options: funding option and redemption option Section starts by explaining the idea that deploying dynamic strategies is the source of nonlinearities in hedge fund returns Then two models that are based on simple return model are introduced A different approach to explain nonlinearities in hedge fund returns will be presented in section The reason for the nonlinearities is from shorting two valuable options of hedge funds This section also discusses a measure for optimal level of leverage and unencumbered cash level Section uses frameworks developed in section and section to suggest a guideline for risk management process of hedge funds Some implications for hedge fund investors and policy makers will also be reviewed Tạp chí Khoa học & Đào tạo Ngân hàng QUẢN TRỊ NGÂN HÀNG & DOANH NGHIỆP Finally section concludes Nonlinearities in hedge fund returns arising from hedge funds’ portfolio strategies It is apparent that the financial security world is a multi-factor one (Agarwal and Naik, 2004) In this world there are many different types of securities, each is comprised different risk factors involved with their own premium returns Therefore, which strategy will be chosen depends on which risk premium associated with its risk factor investors want to earn Because hedge fund can flexibly choose among many asset classes and are freedom from Investment Company Act (1940) controls on leverage, short-selling, cross-holding and derivative position, they are bound to deploy more dynamic investment strategies to earn risk premium than mutual fund strategies There are many linear factor models used to capture risk exposures of portfolios, for example: the arbitrage pricing theory (APT), the capital asset pricing model (CAPM) or Sharpe’s (1992) “style regression” Those models can attribute portfolios’ returns to risk factors only if the returns are linear functions of risk factors Fund and Hsieh (1997) conduct this Sharpe’s (1992) style regression on hedge fund returns The result is that: a considerably fraction of chosen hedge funds in this regression has significant small R2 25% of those funds’ returns are even negative correlated with standard asset classes returns (Fung and Hsieh, 1997) Therefore, those models for linear relationship between returns and risk factors cannot be used for hedge fund returns Numbers of journals have priced securities whose returns are nonlinear with the risk factors using nonlinear asset pricing frameworks, for example: Rubinstein (1973), Kraus and Litzenberger (1976), Dybvig and Ingersoll (1982), Bansal and Viswanathan (1993), Bansal, Hsieh and Viswanathan (1993) and more recently Harvey and Siddique (2000a, 2000b) Now in this paper we will estimate the following regression: Rit = αit + Ʃ bikFkt + uit (1) This model looks like the Sharpe’s “style model” or other models for linear relationship where Rt is excess returns on hedge fund index i, αt is the Số 179- Tháng 2017 55 QUẢN TRỊ NGÂN HÀNG & DOANH NGHIỆP intercept for hedge fund i, bik is avareage loading kth factor for hedge fund i over regression period, Fkt is excess return of kth factor over period t, ui is error term of hedge fund index i However some risk factors are added to this model to capture the risk exposure of hedge fund returns Next we will deeply look into two ways to choose the risk factors to analyze the nonlinearities in hedge fund returns below 2.1 An extension of Sharpe’s style regression The first approach we will discuss is an extension of Sharpe’s (1992) “style regression” According to Fung and Hsieh (1997), there are two aspects that contribute to the concept of “style”: location choice and trading strategies While location choice is asset selection decision made by hedge fund managers to produce return, trading strategy term refers to how to manage those assets in funds’ portfolios: direction (long or short) and quantity (leverage) (Fung and Hsieh, 1997) In the regression equation (1), location choice is Fkt and trading strategies is the bt Hence returns are the product of trading strategies and location choice However hedge fund returns are not that simple because of not only various asset classes it can choose (F is large) but huge range of b: b can be from negative infinity to positive infinity Moreover, b can even change rapidly when hedge funds deploy dynamic trading strategies In Fung and Hsieh (1997) article, there are two steps to find down a model to capture the nonlinearities of hedge fund returns For the first step, Fung and Hsieh choose eight asset classes and high yield corporate bonds for the buy-and-hold strategy They are two bond classes (JP Morgan U.S government bonds and JP Morgan non-U.S government bonds), three equity classes (IFC emerging market equities, MSCI U.S equities and MSCI non-U.S equities), one commodity (gold), 1-month euro deposit for cash and the Federal Reserve’s Trade Weighted Dollar Index for currency The result of this regression on hedge fund returns show that the hedge fund returns are nonlinear with those asset classes returns The reason is that this model is for buy-andhold strategy; however hedge fund managers use dynamic strategies That’s the incentive for Fung 56 Số 179- Tháng 2017 and Hsieh to take place the second step In the second step, Fung and Hsieh factor analyze nearly 500 hedge funds to point out five qualitative style factors that are used to delineate trading strategies by hedge funds industry They are: Systems/ Trend Follow, Systems/ Opportunistic, Value, Distressed and Global/ Macro The concern here is that whether those strategies are dynamic strategies that make the hedge fund returns nonlinear or they are just location choice The regression based on equation (1) will be run over five “style” hedge funds on eight standard asset classes and high yield bonds found in the first step It will give us the information about whether this style use dynamic strategy or not by answer the question how the return of each style factor change in different states of the world The result is that: three factors: System/ Trend Follow, Systems/ Opportunistic, Global/ Macro are dynamic strategies because their returns move extremely in utmost cases The returns of hedge fund portfolios followed those strategies are nonlinear with the returns of standard asset classes Fung and Hsieh (1997) decide to include those three style factors into original model to build a more complete model that can capture the nonlinearities of hedge fund returns 2.2 A model based on option-like feature In Agarwal and Naik study, they also run regression on equation (1) like Fung and Hsieh (1997) Although both studies agree that the strategies a hedge fund follows determine whether its return is linear or not However, there are some differences between two studies We will thoroughly present how the differences in Argarwal and Naik’s study make their model more flexible to capture the nonlinearities in returns of different strategies of hedge funds as follow Agarwal and Naik also follow the regression: (1) Rti = αi + Ʃ bki Fkt + uti All of these terms in the above function are calculated over regression period Fkt here is risk factors which can be either buy-and-hold or option-based risk factors Agarwal and Naik choose three bonds such as Lehman high yield index, Salomon Brothers world government bond index and Salomon Brothers government and Tạp chí Khoa học & Đào tạo Ngân hàng corporate markets index; four equities including Morgan Stanley Capital International (MSCI) emerging markets index, MSCI world except USA index, Russell 3000 index and Russell 3000 lagged index; one cash index and one commodities index The improvement of this study is that Agarwal and Naik include strategies that can earn abnormal returns in this regression They are momentum factor (winners minus losers) (Carhart, 1997), size factor (small minus big) and value factor (high minus low) (Fama and French, 1993) They also use an additional factor to describe the credit risk exposure It is the change in the default spread Beside risk factors for buy-and-hold strategies, risk factors that capture option-based risk are used including liquid out-of-the-money (OTM) and atthe-money (ATM) European options on S&P 500 index The second difference between two studies is the hedge funds’ strategies that researchers choose to critically analyze Agarwal and Naik focus on two types of strategies: the one whose return arises from mispricing securities and the one whose return arises from taking directional bets In the first type, fund managers gain premium returns from mispricing securities rather than economic movements of markets They analyze six strategies in this category: equity hedge, restructuring, event driven, event arbitrage, convertible arbitrage and relative value arbitrage The other type including short-selling and equity non-hedge are strategies whose returns earned from taking directional bets After choosing the risk factors and categorizing the hedge funds strategies, the regression will be run to determine which strategies cause the nonlinearities in funds’ payoffs In general, the result shows that the relationship between risks and returns are nonlinear in most hedge fund strategies because the beta of each regression is significant Especially, the returns of hedge funds in the first category follow restructuring, event driven, relative value arbitrage, event arbitrage and convertible arbitrage are same as payoffs that arise from buying a put option on market indexes There are two main reasons: firstly, the movements of these strategies returns primarily correlate to movement of market indexes They go down when market goes down and vice versa The second reason might be that hedge fund managers Tạp chí Khoa học & Đào tạo Ngân hàng QUẢN TRỊ NGÂN HÀNG & DOANH NGHIỆP make better the Sharpe ratio or they reply to their incentive contract by designing hedge fund returns similar to payoff of writing a put option (Lo, 2001; Siegmann and Lucas, 2002) It is well noted that the risk exposure of event arbitrage strategy found by this model is consistent with Mitchell and Pulvino’s (2001) result In short, in this section, we deeply present the nonlinearities in hedge fund returns arising from their investment strategies The models that work for mutual fund returns cannot capture the risk exposure of hedge funds’ payoff because of the flexibility in dynamic strategies hedge funds can deploy Moreover, because of this flexibility, it is nearly impossible to separate fund managers’ skills into two groups: market timing and security selection We also clearly discuss two different methods used to capture the risk exposure of hedge funds Although the second method is more flexible and can capture more strategies of hedge funds, both of them agree on the nonlinearities of fund returns However, no method can capture all strategies followed by hedge funds because of the lack of information and the huge number of combination strategies of hedge funds The next section will thoroughly look into another method which can capture the nonlinearities in hedge fund returns arising from “funding option” and “redemption option” regarding the strategies those funds follow Nonlinearities in hedge fund returns arising from “funding option” and “redemption option” There have been many studies that try to capture the risk exposure of hedge funds Two of them have been critical analyzed in the above section However the credit crisis in mid 2007 with the collapses of many hedge funds indicated that there were something missed from our hedge fund risk analysis Dai and Sundaresan (2009) point out some lessons learnt from the crisis about a different way to look into the risk of hedge funds They show that the contractual relationships between hedge funds and their investors, their prime brokers are sources of risk In periods of crisis, hedge funds’ prime brokers and other counterparties tend to either raise the Số 179- Tháng 2017 57 QUẢN TRỊ NGÂN HÀNG & DOANH NGHIỆP margin requirement or/and take out their credit lines Their moves can make not only the hedge funds’ funding costs escalate but their profitable risk positions rickety Hedge funds are certain to involuntarily deleverage in bad states of the world, as a result incur losses, and even stand a chance to collapse Moreover, investors in those periods can withdraw their money too Those risks are faced by hedge funds that have even been performing effectively according to other performance standards before or even during the crisis Because of the existence of those two risks, there are nonlinearities in hedge fund returns “Funding option” and “redemption option” are introduced to capture those two risks, where “funding option” is the short option position with hedge fund prime brokers and “redemption option” is the one with their investors Then the effects of “funding option” and “redemption option” to hedge funds’ returns are discussed too We also introduce the concept “unencumbered cash” and its role in managing hedge funds’ risks Two options as well as the level of “unencumbered cash” will together play important parts in hedge fund return model 3.1 “Funding option” and “redemption option” In the first part of this section, the definition and effects of “funding option” and “redemption option” to hedge fund returns will be thoroughly discussed Shorting in those two options are thought to be risk factors that hedge fund faces The Figure below represents the relationships that hedge funds have with their prime brokers and their investors In financial distress period, prime brokers are bound to increase the requirement margin or withdraw their credit lines If these actions happen, hedge funds have to reduce their leverage level The effects of prime brokers’ actions to hedge fund are similar to the effects of shorting an option to deleverage in bad states of the world Shorting this option means that hedge funds have a commitment to decrease their leverage level when the option is 58 Số 179- Tháng 2017 Figure Contractual Relationships Source: Dai and Sundaresan (2009) exercised This option is called “funding option” held by prime brokers When hedge funds need money most to protect its survival is also when investors are likely to withdraw their money from the funds If these actions of investors occur, hedge funds have to be ready to pay back to investors The effects of these actions are similar to the effects of shorting an option by hedge funds This option is called “redemption option” held by investors If the funding options held by prime brokers and funding counterparties are exercised in financial distress, hedge funds have to deleverage If hedge funds not prepare well for this risk, they may have to involuntary decrease their leverage level that may cost a lot If hedge funds involuntarily deleverage, the cost of this may be very high In those circumstances, hedge funds may find difficult to keep their survivals Prime brokers often set a limit to NAV (net asset value) If asset value falls below this limit, the option will be exercised In the case of redemption options, AUM (asset under management) is the basis for investors to decide when they will exercise their options If too much money is withdrawn from the funds, NAV of funds may declines below prime brokers’ limit It will cause the funding option to exercise too and hedge funds face the risk of involuntarily deleverage According to the redemption requests of investors, hedge funds industry paid almost $400 billion of capital in 2008 It is clear that both options can be exercised in bad Tạp chí Khoa học & Đào tạo Ngân hàng states of the world The hedge fund returns or even their survivals will be influenced dramatically if those two options are exercised Therefore the fact is that the process of capture hedge fund returns as well as their risk is not only looking at the strategies they deploy but the hedge funds’ managements of two options In the next section the concept of “unencumbered cash” together with two options are the foundation to build a model that can capture the return of hedge funds given the existence of funding option and redemption option With this model, it is apparent that the relationship between hedge fund returns and their risk factors is nonlinear 3.2 Model of Hedge Fund Return If leverage level of a hedge fund is zero or this hedge fund is totally “self-financing”, its return can be written as follow: (2) Rt = αt-1 + σt-1 εt Where Rt is the return of hedge fund, αt is excess return without any leverage and σt is the volatility of hedge fund’s assets In fact, in order to earn premium return, hedge funds need to increase their leverage level However, the level of leverage is finite because an initial margin has to be paid to prime brokers or other counterparties to alleviate the risk those parties bear in case that hedge funds cannot pay the debt Given the existence of funding option and redemption option, if the leverage level of hedge funds (denote L) is low, the relationships between their expected returns and L may be linear When L increases, the volatility of hedge funds’ expected returns will be bigger Then the probabilities that two options are exercised increase Because the two options are shorted by hedge funds, the case when they are in the money will have negative effects on hedge fund returns If L is big enough, the negative effects will dominate and make the expected returns not linear with L It is clear that given strategies that hedge funds choose to deploy, the level of leverage they take will affect their expected returns We will present how two options and “unencumbered cash” affect expected returns, then show the model of expected returns given the presence of two options The effects of redemption option are captured Tạp chí Khoa học & Đào tạo Ngân hàng QUẢN TRỊ NGÂN HÀNG & DOANH NGHIỆP by the change of number of shares over time When the annual returns fall below the limit that investors set for hedge funds, investors will withdraw a fraction of AUM from the funds Such an action will make the number of shares decrease At the same time, hedge funds also have contracts with their prime brokers and other counterparties If annual returns are lower than the other limits that prime brokers set for hedge funds, prime brokers can increase the required margin or withdraw their credit lines from the funds The limits for prime brokers are often lower than the ones for investors It is apparent that if the hedge funds can prove that they are safe credit, both investors and prime brokers can lower their limits for annual returns One signal that investors and prime brokers can use to indicate the safety of hedge funds is their unencumbered cash level Unencumbered cash is the fraction of AUM that is not posted to primer brokers as margin The economical definition of unencumbered cash is the amount investors can take if all margins posted are lost or all counterparties fail to repay the margins Unencumbered cash is an important tool to manage the risk that hedge funds bear It might be the most important one for some reasons (Dai & Sundaresan, 2009) First of all, it is obvious that the higher the level of unencumbered cash is, the more confident investors feel about the funds In this case, unencumbered cash is a form of insurance for investors Secondly, unencumbered cash level is a signal for not only investors but also prime brokers Unencumbered cash level is correlated with the leverage of hedge funds So it helps prime brokers measure the level of risk they take to set an appropriate limit for hedge funds Last but not least, in a financial world with a lot of inefficiencies, a risk management tool that focuses on unencumbered cash level is certainly useful for different parts in managing risk of hedge funds Dai and Sundaresan (2009) also point out that the unencumbered cash level of a hedge fund is negatively correlated with its level of leverage, volatility of its assets and change in the limit prime brokers set to hedge fund but positively correlated with the change in AUM between two periods That means the unencumbered cash level will decrease if leverage level, volatility of hedge fund’s assets and limit posted by prime broker Số 179- Tháng 2017 59 QUẢN TRỊ NGÂN HÀNG & DOANH NGHIỆP increases and AUM decreases over time It is well noted that one practical risk management standard is keeping the level of unencumbered cash constant over time Because this level can change immediately over time if the level of AUM, volatility or limit set by prime brokers change, in order to keep unencumbered cash level constant, hedge fund has to voluntarily change the level of leverage It is important to note that this change in leverage level here is voluntary so the cost of this action may be insignificant and the benefit of this action is reduce the probability of exercising the two options If one or both options are exercised in bad states of the world, the cost to deleverage involuntary will be considerable or even can be the death of hedge fund With the critical analysis of two options and unencumbered cash level, Dai and Sundaresan (2009) come up with a model for return of hedge fund: ht+1 ≡ ln(NAVt+1(NAVt)-1) = Ltσt[Rt-1σt-1σ − a(1 − Lt+1Lt-1 1_{Rt-1σt-1 < R(Ltσt)-1} − b(1 − Lt+1Lt-1 (3) 1_{Rt-1σt-1 < R(Ltσt)-1}] Obviously, the returns of hedge funds are nonlinear with the level of leverage given the existence of funding and redemption options The part in brackets shows the effect of two options to the hedge fund returns The higher the leverage is, the higher value of the part outside the bracket is At the same time, the higher the level of leverage is, the higher the probability of two options being exercise, that makes value of the part in the bracket decrease Therefore, leverage level has a dynamic relationship with the return of hedge funds The second conclusion is that not only the level of leverage but the volatility of hedge fund’s assets together affect the expected returns In order to gain the maximum returns, funds should adjust both leverage and volatility levels More interestingly, if two funds agree on deploying the same strategies but have two different volatilities in asset, they should choose two different leverage levels as well to earn their optimal returns Lastly, Dai and Sundaresan also suggest that when choosing optimal levels for leverage, hedge funds must be careful with the level of unencumbered cash It is noted that the optimal leverage level can be associated with a too low unencumbered cash level Therefore hedge funds often set a limit level for unencumbered cash and then choose the optimal level of leverage that doesn’t make the unencumbered cash level fall below this limit With this function above, this approach not only shows the nonlinearities in hedge fund returns but also has advantages compared to other methods It provide a tool to calculate the optimal level of leverage and unencumbered cash that help hedge funds gain the maximum return and safe from financial distress 3.3 Advantages of this approach The first highlight of the approach analyzed above is the way it solves the risk budgeting problem Other methods often try to find a suitable VaR in order to earn a target expected return However, this approach makes an effort to get the biggest return keeping in mind that the two options can be exercised if the leverage level is too high In bad states of the world, if one or both options are exercised, each risk taking unit will cost hedge funds different value to involuntarily deleverage This approach also points out an important conclusion: with the presence of two options, the optimal leverage level is lower More significantly the higher the deleverage cost is, the lower the optimal level of leverage is This approach provides a new way to look into risk aversion and distance to default It is said that risk aversion is endogenous and determined by the deleverage cost if two options are exercised unexpectedly Whereas distance to default is endogenous too, hedge funds can choose the optimal levels of distance to default to keep the funds safe from unexpected increase in volatility or unexpected changes in limits set by investors and prime brokers Secondly, this approach also discusses the relative risk capital among risk taking units and the role (3) 60 Số 179- Tháng 2017 Tạp chí Khoa học & Đào tạo Ngân hàng of unencumbered cash level It is notably that the relative risk capital depends not only on Sharpe ratio but the correlation coefficient among risk taking units This term will be determined solely by Sharpe ratio if the correlation coefficient between two risk taking units is zero Dai and Sundarean (2009) also said that when choosing the optimal leverage level of hedge funds, managers should consider the level of unencumbered cash Because if one or both options are exercised and the unencumbered cash level is too low, hedge funds may be found in difficult to meet those new requirements and have to involuntarily deleverage or even fail Therefore, hedge funds should set a limit for unencumbered cash level and choose the optimal leverage levels to gain the maximum returns without make unencumbered cash level fall below the limit In brief, with presence of two options: funding option and redemption option, a new and useful method to capture hedge funds’ risk is introduced This method together with other traditional ones may help hedge fund managers have a better sight about their risks We will try to make a proper guideline on risk management for hedge fund managers, investors and policy makers in the next section Guideline for risk management process of hedge funds Three important lessons for risk managers of hedge funds are the nonlinearities in hedge fund returns, importance of good relationships with prime brokers and investors; and the role of unencumbered cash level in risk management process Moreover, this section also provides advices for investors in making better investment decisions on hedge funds and for policy makers in order to construct a more stable financial market given the risk exposure of hedge funds First and foremost, it is apparent that even each researcher uses different approaches, all of them agree on the nonlinearities of hedge fund returns This understanding will help hedge fund managers make better decisions on choosing dynamic strategies and measure expected returns and risk more accurately Hedge fund managers should be extra careful to mitigate the possibility of diversification Tạp chí Khoa học & Đào tạo Ngân hàng QUẢN TRỊ NGÂN HÀNG & DOANH NGHIỆP implosion and vulnerability of extreme events In the second place, building good relationships with prime brokers and investors holds no less important role Hedge funds with good risk management strategies will try to have long-term and strong relationships with multiple prime brokers It is unlikely that all prime brokers want to exercise funding options at the same time Therefore, having multiple prime brokers also keeps hedge funds safer from exercised options and gives other counterparties more confidence about survival of hedge fund even in financial crisis Moreover, hedge funds should pay attention in building relationships with investors too In redemption contract, there is often penalty for early exercising redemption option Since all these conditions are proposed by investors and agreed by both sides, all investors understand that exercising redemption options in bad states of the world helps them take back their money at a price Understanding well each type of investors and designing suitable contracts with them is an important mission for hedge fund managers In the third place, unencumbered cash level needs carefully considering When investors want to redeem their money or prime brokers want to increase the initial required margin or volatility of assets changes unexpectedly, hedge funds may find itself in difficult to provide those liquidities due to the shortage of unencumbered cash In extreme cases, hedge funds may have no other choice than fire sale In order to keep itself safe, hedge fund managers should set a threshold for unencumbered cash level, and choose an optimal level of leverage that always keep unencumbered cash level above this limit Last but not least, benefits of efficient risk management tools for investors and prime brokers are noteworthy The better understanding about risk exposure of hedge funds may help investors make accurate investment decisions about managers’ compensation This knowledge also provides policy makers a better perspective drawing of risk exposure hedge funds can bring to the whole financial market Moreover, benefits from good contractual relationships between hedge funds and prime brokers, investors have attracted much interest from three parties Besides, unencumbered cash level plays an important Số 179- Tháng 2017 61 QUẢN TRỊ NGÂN HÀNG & DOANH NGHIỆP role in helping policy makers make decisions too Dai and Sundaresan (2009) said that using unencumbered cash level to measure fund’s risk is easier and more transparent than using the traditional tool: VaR VaR is efficient but demands more assumptions and often depends on the tool prime brokers use to measure hedge funds’ risk More notably given many advantages this approach gives to investors, policy makers should try their best to make sure all investors have the same benefits Understanding the risk exposure of hedge funds is an important area of research We need a better understanding of this issue while making investment management decisions involving hedge funds Unfortunately this is a tricky issue, since hedge funds provide limited disclosure Moreover, hedge funds are not popular in Vietnam So the information of hedge funds is limited to be tested There are several papers analyzing risk management of hedge funds from investors’ perspective However, our approach in this paper focus more about risk management from hedge funs’ managers’ perspective In this context, our approach provides useful information to managers dealing with risk management related issues Estimation of hedge fund risks is also important, as a large number of hedge funds propose a risk free rate as a bench mark for claiming fees Benchmarking of hedge funds is an important area of research Investing in hedge fund involves significant costs for the investor and selecting the right manager is crucial in case of hedge funds Hence, a benchmark that accounts for the linear and non-linear risk exposures of hedge funds is necessary for their performance evaluation Our study contributes by providing a simple yet powerful approach to design a benchmark for hedge funds and to evaluate their performance It would be interesting to examine whether such a risk-adjusted performance is related to characteristic features of hedge funds such as size, lockup period, incentive fees, etc as well as to investigate the determinants of inter-temporal variation in the alphas of hedge funds These issues are being investigated as a part of ongoing research in the future Conclusion Hedge funds differ from others in many aspects Hedge funds have flexibility in choosing many asset classes and dynamics strategies Those advantages help them gain extra returns but also require a more complex risk management framework There are two main explanations for the nonlinearities of their returns: dynamics strategies and short positions of two options with investors and prime brokers Until now, hedge funds are still not popular in Vietnam One of the most important reasons for this is that hedge funds require a sophisticated and complex risk management There are also other problems in Vietnam that make hedge funds even more risky They are: the simplicity of financial market, the lack of financial instruments, the information asymmetric among investors and hedge funds’ managers and the lack of legal provisions If the two risk management approaches in this paper are considered to apply in Vietnam, there are many things need to be done However, all in all, for all the mentioned above, we strongly believe that hedge funds managers, investors and policy makers should take our analysis into attention to make effective decisions in the future to not only generate more profit but also keep a Tài liệu tham khảo Agarwal, V., & Naik, Y.N., 2004 Risks and portfolio decisions involving hedge funds The Review of Financial Studies, spring, 17(1), 63-98 Bansal, R., Hsieh, D., & Viswanathan, S., 1993 A New Approach to International Arbitrage Pricing Journal of Finance, 48, 1719-1747 Bansal, R., & Viswanathan, S., 1993 No Arbitrage and Arbitrage Pricing Journal of Finance, 48, 1231-1262 Carhart, M., 1997 On Persistence in Mutual Fund Performance Journal of Finance, 52, 57-82 Dai, Q., & Sundaresan, S.M., 2007 Risk Management Framework for Hedge Funds: Role of Funding and Redemption Options on Leverage Available at SSRN: http://ssrn.com/abstract=1439706 Dybvig, P H., & Ingersoll, J.E., 1982 Mean Varian Theory in Complete Markets Journal of Business, 55, 233-251 Fama, E F., & French, K.R., 1993 Common Risk Factors in the Returns on Stocks and Bonds Journal of Financial 62 Số 179- Tháng 2017 Tạp chí Khoa học & Đào tạo Ngân hàng QUẢN TRỊ NGÂN HÀNG & DOANH NGHIỆP Economics, 33,stable 3-56 financial markets ■ healthy and Fung, W., & Hsieh, D., 1997a Empirical characteristics of dynamic trading strategies: The case of hedge funds The Review of Financial Studies, 10, 275-302 Gatev, E G., Goetzmann, W.N., & Rouwenhorst, K.G., 1999 Pairs Trading: Performance of a Relative Value Arbitrage Rule Working paper, Yale University Glosten, L.R., & Jagannathan, R., 1994 A Contingent Claim Approach to Performance Evaluation Journal of Empirical Finance, 1, 133-160 10 Gregoriou, G N., 2002 Hedge Fund Survival Lifetimes Journal of Asset Management, 3(3), 237-252 Harvey, C., & Siddique, A., 2000a Conditional Skewness in Asset Pricing Tests Journal of Finance, 55, 1263-129 11 Harvey, C., & Siddique, A., 2000b, Time-Varying Conditional Skewness and the Market Risk Premium Research in Banking and Finance, 1, 27-60 12 Hull, C.J., 2012 Risk Management and Financial Institutions, Third Edition, John Wiley & Son, Inc INVESTMENT COMPANY ACT 1940 13 Jagannathan, R., & Korajczyk, R., 1986 Assessing the Market Timing Performance of Managed Portfolios Journal of Business, 59, 217-235 14 Koh, F., Lee, K.C., & Fai, K., 2004 Investing in Hedge Funds: Risks, Returns and Performance Measurement Research Collection Lee Kong Chian School of Business Paper 2341 15 Kraus, A., & Litzenberger, R., 1976 Skewness Preference and the Valuation of Risk Assets Journal of Finance, 31, 10851100 16 Lo, A., 2008 Hedge Funds: An Analytic Perspective Princeton University Press 17 Merton, R.C., & Henriksson, R.D., 1981 On Market Timing and Investment Performance II: Statistical Procedures for Evaluating Forecasting Skills Journal of Business, 41, 867-887 18 Mitchell, M., & Pulvino, T., 2001 Characteristics of Risk in Risk Arbitrage Journal of Finance, 56, 2135-2175 19 Sharpe, W.F., 1992 Asset Allocation: Management Style and Performance Measurement Journal of Portfolio Management, 18, 7-19 20 Siegmann, A., & Lucas, A., 2002 Explaining Hedge Fund Investment Styles by Loss Aversion: A Rational Alternative Working paper, Vrije Univesiteit, Amsterdam 21 Stowell, D., 2010 An introduction to Investment Banks, Hedge Funds, and Private Equity,The New Paradigm, Elsevier, US 22 Treynor, J., & Mazuy, K., 1966 Can Mutual Funds Outguess the Market? Harvard Business Review, 44, 131-136 Thông tin tác giả Nguyễn Diệu Hương, Thạc sỹ Khoa Ngân hàng, Học viện Ngân hàng Email: huongnd@hvnh.edu.vn Summary Framework for risk management of Hedge fund This paper analyzes two different approaches to capture the nonlinearities of hedge fund returns One approach provides an adequate framework used for style analysis of hedge funds’ dynamic strategies The other one uses the contractual relationships hedge funds have with their investors and prime brokers to explain the nonlinearities in their returns The effects of those relationships are captured by hedge funds’ short positions of two options: funding option and redemption option This paper also suggests a guideline in risk management process for hedge funds’ managers They can obtain the optimal leverage level to gain maximum returns without making unencumbered cash level too low Implication of this analysis for investors and policy makers are also presented Key- words: hedge fund, risk management, funding option, redemption option, nonlinearities returns Huong Dieu Nguyen, M.Ec Banking Faculty, Banking Academy Tạp chí Khoa học & Đào tạo Ngân hàng Số 179- Tháng 2017 63 ... long-term and strong relationships with multiple prime brokers It is unlikely that all prime brokers want to exercise funding options at the same time Therefore, having multiple prime brokers also... hedge funds from investors’ perspective However, our approach in this paper focus more about risk management from hedge funs’ managers’ perspective In this context, our approach provides useful... maximum return and safe from financial distress 3.3 Advantages of this approach The first highlight of the approach analyzed above is the way it solves the risk budgeting problem Other methods often