ISAM. ISAM Whitepapers And Research Memorandums. A Selected List June International Standard Asset Management

Transcription

1 International Standard Asset Management ISAM Whitepapers And Research Memorandums A Selected List June 2015 Cayman Islands 802 West Bay Road P.O. Box 30599, KY Grand Cayman, Cayman Islands BRITISH WEST INDIES Tel: London 52 Queen Anne Street London, W1G 8HL UNITED KINGDOM Tel Fax New York 205 Lexington Ave New York, NY UNITED STATES Tel Fax

2 All Portfolios Need Trend Abstract: In this paper we expand the analysis of the portfolio benefits of trend following to an array of assets. The purpose of doing so is to see whether the portfolio benefits hold across more exotic portfolio alternatives than just equities and bonds. We consider over one million potential portfolio combinations and evaluate the benefits of a trend following allocation. The results unquestionably confirm the value of trend following within all portfolios. The Cost of Ignoring the Trend Following Asset Class Abstract: This paper investigates the quantitative impact of not allocation to trend following within a portfolio. We ask the question: "What is the cost of ignoring Trend Following?" and demonstrate that an investor would very rarely regret an allocation to a trend following strategy. Additionally we provide evidence that at the portfolio level an investor would need to harbour extreme views in order to justify avoiding the strategy entirely. S&P 500, Is the Trend Your Friend Abstract: In this paper we study the performance of different trend-following systems on the S&P500 with a particular focus on the speed of trading. We ask the question of what kinds of systems have performed well on the S&P500 over the past 5 years, and put that into historical context using over 80 years of S&P500 data. Using both the recent steady recovery in equity markets post 2009 and the entire history of the S&P500 we examine various tradeoffs and portfolio benefits of fast/medium/slow systems from both a short term and long term perspective. Simple distributional analysis demonstrates dispersion in performance across speeds of trading and across time. We use this fact to discuss the benefits of diversification across trading speeds from a portfolio context. CTA Style Factor Partitions Abstract: CTA style factor analysis provides a new set of tools for helping investors understand performance. This paper is about taking a deeper look at CTA style factors focusing on construction, usage, and interpretations. Given the interest and ease of use of the CTA style factor framework, we clarify a few subtle points as well as pitfalls in the approach. In order to help explain factor construction at a more granular level, we discuss weighting schemes and factor partitioning for both the size and speed factors. We examine the two axes of size and speed, and also zoom in to view the results at intra-sector level. Empirical results demonstrate that despite different approaches to partitioning factors, CTA style factor analysis is a relatively robust tool for analysis. Partitioning factors and understanding factor construction leads to more sophisticated use of these tools. CONFIDENTIAL - DO NOT DISTRIBUTE 1

3 Style Factor Analysis of CTA Returns Abstract: The Fama-French Three Factor model has been widely utilized in the equity world, but similar analytic tools in the CTA world are lacking. The goal of this paper is to introduce tools capable of analyzing the detailed factors that drive much of CTA performance in a similar fashion to the Fama-French analysis, and thereby attribute performance to relevant style tilts. Utilizing trend-following as an example, we propose three factors (speed, equity, and size) to analyze a CTA s track record in the form of a return-based style analysis. Empirical data shows that the three factors in conjunction with the base system s returns can largely explain the return differential within the space. This forensic analysis of a CTA s return can reveal the portfolio styles, and at the same time address the frequent question about a CTA s relative performance against its peers and the corresponding benchmark. The Impact of Mark-to-Market on Return Correlations Abstract: There exists a general belief that the funds in the trend-following space are highly correlated with each other, while the hedge funds in other categories, e.g., credit arbitrage, are less correlated with each other. In this note, we demonstrate the impact of mark-tomarket on the correlations between hedge fund returns within the same category with less liquid markets, and provide a potentially powerful counter-argument to the myth of higher correlation amongst CTA funds as compared to the perceived lower levels of correlation noted in strategies where mark-to-market is needed. The Benefits of Pure Trend-Following: The Case against Diversifying the Diversifier Abstract: Trend-following provides substantial diversification benefits in a traditional portfolio. Recently however, trend-following as an asset class has lagged in standalone performance. It is thus tempting for some managers to add non-trend-following strategies to their portfolios in an effort to improve performance. In this paper we present the case against diversifying the diversifier, both by illustrating several hidden risks of a multi-strategy CTA and then by demonstrating the superior portfolio benefits of pure trend-following to an institutional investor. We show empirically that the desirable return characteristics of positive skewness and crisis alpha of trend-following are undermined by non-trend-following strategies. CONFIDENTIAL - DO NOT DISTRIBUTE 2

4 Hidden Risks of Leveraged Low Volatility Abstract: The volatility of some relative value strategies is inherently low. It is likely that a significant leverage is needed for these low volatility strategies in a portfolio consisting of strategies with heterogeneous risk profiles. For example, for risk-based allocation in a portfolio including both managed futures strategies and relative value strategies, the investor may need to leverage up the relative value strategies to achieve a desirable risk exposure. In this note, we first illustrate the need of leverage for relative value strategies. Then, we demonstrate the hidden risks of the leveraged volatility in extreme cases and in general when correlations change. Rising Interest Rates and Roll Yield Abstract: In this note we provide empirical data about the relationship between the two components of the futures price change: the spot price change and roll yield. We demonstrate that the spot price change and roll yield are fairly uncorrelated, which implies that, comparing with the falling interest rate regime in the past two decades, in a rising interest rate environment the roll yield is not more likely to work against any trend due to spot price decrease. The empirical evidences in this work do not support the pessimistic view of a rising interest rate environment s negative impact on trend-following profitability, in particular, in the bond sector. Diversification: Size Matters Abstract: In this paper we discuss the diversification benefit inside a trend-following portfolio as a function of the AUM size. There exists quite explicit association between AUM size and the diversification benefits along both axes of markets and speeds: when AUM increases substantially, allocations given to relatively less liquid markets must decrease, and faster systems have to cede ground to slower systems. Therefore, we can observe monotonically decreasing diversification when AUM grows. Additionally, we examine the impact of the AUM size on the diversification benefit of a trend-following program in a traditional portfolio. We demonstrate that the correlation is increased and the crisis alpha is reduced when the AUM size is substantially increased. CONFIDENTIAL - DO NOT DISTRIBUTE 3

5 Difference Between Two Correlated Return Streams Abstract: In this note we examine what kind of performance differences we can observe simply as a function of two managers volatility and correlation expectations. We derive a theoretical formula for this and then compare to actual observed performance differences within a peer group. The analysis shows that even at fairly high levels of expected correlations, significant performance differences are quite likely. Are All CTA s Created Equal? Abstract: This paper proposes a methodology with which to analyze the various drivers of return dispersion within the trend-following space. Delineating specific parameter choices and crafting subspaces along various dimensions allows for a greater degree of clarity into the return drivers of this differentiation. Additionally, we demonstrate that simple statistics can be used to classify a given track record to a specific subspace. Therefore, a more appropriate benchmark can be established for the track record, and potential style shifts can be exposed through time. The Multi-Centennial View of Trend-Following Abstract: In this paper we present a multi-centennial view (since the 1200 s) of the distinct return and risk characteristics of trend-following. The nearly 800 years of data that we have collected supports the view formed over the last few decades, that trend-following as an alternative asset class provides a desirable return profile with low correlation to traditional asset classes, positive skewness and meaningful capability to robustly deliver positive returns during crisis periods. The State of Trend-Following: 2013 Abstract: In this work we review the current environment from the aspects of industry crowdedness, low interest rate, government interventions, and market correlation. The empirical and historical data does not support an alarming scenario for trend-following in the current environment. Still, we find that correlations have increased recently and as a result the overall diversification has fallen. Even though there is little direct relationship between correlation and performance, it is true that increased correlation increases the tails at both ends of the performance spectrum. However, by adding less correlated new markets, especially at an AUM of a limited size, and reducing signal sensitivity we can tackle the potential challenges facing trend-following. We also establish specific benchmarks for a trendfollowing system as a function of various parameter choices, and examined the nature of what has led to such disperse outcomes and identified a sequence of parameter choices that polarized returns. CONFIDENTIAL - DO NOT DISTRIBUTE 4

6 Trend Following in Equity Markets: The Cost of Crisis Alpha Abstract: In this paper, we demonstrate that adding long bias to the equity sector is able to improve the Sharpe ratio of the equity sector and the whole trend following portfolio as well. It can also make a dramatic difference of the return of the equity sector since However, this improvement in terms of Sharpe ratio is at the cost of the crisis alpha. The standard trend following system, without the long bias in the equity sector, can deliver a higher crisis alpha, and thus more benefit to a traditional institutional portfolio. Chasing Performance vs Buying Dips: Investing in a Trend-Following Portfolio Abstract: In this paper, we discuss both theoretic and empirical results of chasing performance and buying dips for investing in a trend-following portfolio. Since the return series of a trend-following portfolio exhibits mean-reverting characteristic, the performancechasing investor is unlikely to achieve improved performance. Buying-dips can often improve the Sharpe ratio, but caveats of this type of dynamic leveraging scheme include a higher likelihood of catastrophic loss unless a leverage cap is employed. With the maximum leverage capped to limit the likelihood of catastrophe, a Buy-the-Dips investor might have to accept an improved risk-adjusted return (the Sharpe ratio) at the cost of a reduced total expected return. We then discuss four scenarios in which the corresponding investor profiles can increase the probability of achieving improved total expected returns. Moreover, in the framework of optimization under uncertainty, we investigate the optimum solution with regard to a given joint distribution of Sharpe ratio and serial correlation. The Fallacy of Trend-Following Trend-Following Abstract: Under certain assumptions and conditions, we theoretically prove that a generic trend-following portfolio exhibits mean-reverting characteristic. We utilize the equivalence between a trend-following portfolio and a lookback straddle option, of which the price can be expressed in a closed-form formula. By showing that the variance ratio statistic of the price changes of a lookback straddle is lower than 1, we conclude that the trend-following portfolio s return is mean reverting, which is further confirmed by empirical data. Therefore, trend-following a trend-following portfolio, i.e., trend-following squared, is not an optimal approach for one to invest in a trend following portfolio. CONFIDENTIAL - DO NOT DISTRIBUTE 5

7 Trend Following: The Impact of Interest Rates Abstract: The potential impact of interest rates on trend following performance originates from two sources: the additive interest income of the fully funded investments, and the futures prices inherent dependency on interest rates. In this paper, we demonstrate that the additive interest income from cash management can be quite significant in improving the performance, both Sharpe ratio and drawdown, even if the interest rate is only moderately increased. For the potential impact of interest rates on changes in futures price, we consider both the change in interest rates and the interest rate level, with the former focusing on the bonds and the latter focusing on the commodities. The overall empirical evidences appear to suggest that trend trading could benefit from an increase in the level of interest rates. Unravelling Hidden Risks: The Case of Trend Following Abstract: In this paper, based on a representative trend following system and alternative trend following systems with dynamic leveraging, we demonstrate that, while the typical riskadjusted returns on assets may appear better, a more complete profile of risks committed to generate such performance may actually indicate otherwise and reveal the hidden risks. We discuss the hidden risks from the aspect of inflated Sharpe ratios, margin to equity ratio, rolling volatility, and return distribution. This paper is a more comprehensive version of an earlier whitepaper on performance measurement by ISAM. The Characteristics of Drawdown Abstract: In this paper, we discuss the expected maximum drawdown and expected longest drawdown length based on theoretic analysis and Monte Carlo simulation. These expected values can provide a baseline for an investor to gauge an observed drawdown. Using trend following as an example, we examine the trade and signal statistics in the drawdown periods, and empirically confirm that a lower percent of winning markets, rather than the higher loss of losing markets, is often observed in larger drawdown periods. Finally we demonstrate that trend following typically possesses a relatively strong recovery ability. Due to the negative correlation between the drawdowns of trend following and the S&P 500 index, trend following provides significant diversification benefits to a traditional institutional portfolio. CONFIDENTIAL - DO NOT DISTRIBUTE 6

8 Dynamic Leveraging As a Factor of Performance Attribution Abstract: There exist extensive academic studies for hedge fund performance evaluation. Extending Sharpe (1992) s work on mutual funds performance attribution, hedge funds returns are generally characterized in the literature by returns of buy-and-hold strategies of assets in the managers portfolios and returns attributed to their dynamic trading strategies as well. Here we introduce a dynamic leveraging factor, and demonstrate its potential significant impact on performance. In addition, we propose a Fourier transform based methodology to detect dynamic leveraging and to decompose the return sources. We apply the analysis to several real track records. Finally, via Monte Carlo simulation, we try to quantify probabilistically the evaluation discount to a track record due to this factor. Trend Following: The Myth of Return Dispersion Abstract: This paper seeks to isolate the impact of a number of generic industry inputs on return dispersion among CTA's. In particular, we demonstrate that specific combinations of lookback and capital allocation among sectors may result in significant return dispersion in a short period among CTA s. However, the performance impact of these two factors appears to be random, and the return dispersion diminishes with increased investment horizon. Moreover, we empirically quantify the diversification needed to alleviate the return dispersion for a short investment horizon. Finally, we establish a return dispersion index which helps investors gauge the typical range of trend following performance. Trend Following: Empirical Evidence of the Stationarity of Trendiness Abstract: In this paper, we present some empirical evidence showing the stationarity of market trendiness, even after the CTA industry has witnessed the extraordinary increase of scale and some dramatic change of the economic environment. Instead of using performance statistics of a specific trend following program, here we introduce a strategy-independent market trendiness measure, the market divergence index (MDI), and demonstrate that the MDI is stationary along time via statistic tests and by Fourier transform analysis. The stationarity indicates that the market trendiness, even in the presence of the dramatic changes of the CTA industry and the economic environment, is still statistically consistent with what the markets exhibited in earlier years. In addition, we provide convincing empirical evidence that, even with the substantial increase of players and capital in the CTA industry, trend following is still able to capture the opportunities presented by market trends. CONFIDENTIAL - DO NOT DISTRIBUTE 7

9 Performance Measurement: Unravelling the Hidden Risks Abstract: Measuring performance based on returns on assets under management, e.g., by the Sharpe ratio or Omega ratio, is common and certainly provides a sensible way to judge managers. The Sharpe ratio is indeed based on risk-adjusted returns on assets, but this way of performance measurement hides a subtle but important factor, which is how much risk the manager takes to generate the returns. Note that the realized volatility of the returns in the Sharpe ratio is simply a single representation or outcome of the amount of risk a manager takes rather than the underlying risk itself. A few different approaches to taking risks into account in performance measurement are discussed in this paper. Using a representative trend following system and an alternative system with dynamic leveraging, we demonstrate that, while the risk-adjusted returns on assets may appear better, a more complete profile of risks committed to generate such performance may actually indicate otherwise. An Illustrative Comparison of Trend Following in Different Interest Rate Regimes Abstract: Since early 1980 s, the interest rates have been decreasing, and are currently at a record low. This study tries to provide several illustrative examples to alleviate investors concerns on the possible impact of rising interest rates on trend following. We look at two sectors: agriculture and fixed income. Several agriculture markets have futures data up to early 1960 s covering two decades of rising interest environment. While most of the fixed income markets do not have futures data until 1980 s, we derive the corresponding futures price from the yield data, which started as early as in 1960 s. The tests for both sectors demonstrate that trend following s performance is not degraded in a rising interest environment. The fact that the agriculture markets performed much better in the rising interest rate environment is supportive of including these less liquid markets in the portfolio at a forward-looking base, even though in the past decade the agriculture markets appeared to worsen the portfolio performance. Trend Following: Empirical Findings of Diversification by Less Liquid Markets Abstract: In this paper, we highlight a specific factor of capital allocation: inclusion or exclusion of less liquid markets. We will discuss the impact on returns, Sharpe ratio, drawdown, and crisis alpha by including an increasing number of less liquid markets in the portfolio. We have found empirically that the diversification impact provided by less liquid markets was generally negative on returns in the past 20 years, but the benefit of diversification by less liquid markets lies at an improved Sharpe ratio, a reduced maximum drawdown, and an increased crisis alpha as well. CONFIDENTIAL - DO NOT DISTRIBUTE 8

10 Crisis Alpha: ISAM Systematic Abstract: Crisis alpha measures a strategy s performance during equity market crisis. It represents the difference between the original Alternative Investment strategy and the strategy without crisis periods where the performance of the strategy is substituted with an investment in the short-term debt rate. In this study, we demonstrate that ISAM Systematic s crisis alpha is significantly positive, and is substantially higher comparing with the NewEdge index. Rather than subjectively specifying/picking the crisis periods, we use a VIX based criteria to define the crisis periods: any month with a VIX jump of above 20% over the value in the end of last month is identified as a crisis month. Is the 2/20 Worthwhile? Abstract: This study tries to provide one justification for the 2/20 fee schedule for a firm like ISAM versus the 1.x/0 fee schedule for a large mutual fund CTA. The mutual fund CTA s main focus would be to increase the asset base for the management fees. When they reach a large asset base, they are not be able to allocate sufficient capital to the commodity sector, which will lead to a degraded performance with a lower net Sharpe ratio and being less negatively correlated with the SPX down movement. Preliminary Results on Time-Reversed Prices Abstract: In this study we build a system based on time-reversed prices, and compare it with the system using the real prices. We have found that the time-reversed prices generate a much inferior system. However, when we conduct the same experiments over artificial price series of random walk, with or without shift, the model performances for both directions of the prices are very close. The asymmetry of model performance along forward and backward time directions for the real prices indicates that there exist some certain fundamental characteristics of a real price series, which are needed for some price-based models to perform. Trend Following: Myth of the Sharpe Ratio Abstract: Due to the specific payoff characteristic of option selling, there exists a significant probability for an option selling strategy to achieve seemingly superior performance in a short period such as two years. In this paper, we show that a trend following system with an option selling overlay is able to achieve a quite large Sharpe ratio in a short period at a much higher probability, even though the overall Sharpe ratio in an extended period is similar to what the trend following system achieves. We thus demonstrate that high Sharpe ratios over short periods of time (even apparently as long as several years) can be misleading, and may be hiding essential risks as well. CONFIDENTIAL - DO NOT DISTRIBUTE 9

11 I M P O R T A N T INF O R M A T I O N ISAM MAY USE IT S AFFILIATE, IS PRIME TO TRADE FX. IS PRIME DOES NOT EARN FEES FROM ANY ISAM PRODUCTS. This document has been issued and distributed by ISAM (USA) LLC when it pertains to U.S. investors, ISAM (EUROPE) LLP when it pertains to U.K. investors, ISAM Middle East Ltd when it pertains to investors in the Middle East and ISAM Asia Pacific Pty Ltd when it pertains to investors in Australia (each of ISAM (USA) LLC, ISAM (EUROPE) LLP, ISAM Middle East Ltd and ISAM Asia Pacific Pty Ltd is referred to herein as ISAM ). The information in this document does not constitute, or form part of, any offer to sell or issue, or any solicitation of an offer to purchase or subscribe for any fund or investment program described in this document; nor shall this document, or any part of it, or the fact of its distribution form the basis of, or be relied on, in connection with any contract. Some of the information contained in this document may provide estimated data only, is unaudited, is subject to change based on final data and should be relied upon as such. Some of this information may have been provided by third-party sources and, although believed to be reliable, it has not been independently verified and its accuracy or completeness cannot be guaranteed. This document is confidential and is intended solely for the information of the person to which it has been delivered by ISAM. It is not to be reproduced or transmitted, in whole or in part, to thirdparties without the prior consent of ISAM. Recipients of this document are reminded that any purchase or subscription to any investment program or vehicle may only be made solely on the basis of information contained in a formal offering document, describing, among other things, the risks associated with such investment program or vehicle, which may be different from the information contained in this document. No representation or warranty, express or implied, is made or given by or on behalf of ISAM or its directors or any other person as to the accuracy, completeness or fairness of the information or opinions contained in this document, and no responsibility or liability is accepted for any such information. The distribution of this document in certain jurisdictions may be restricted by law; therefore, people into whose possession this document comes should inform themselves about and observe any such restrictions. Any such distribution could result in a violation of the law of such jurisdictions. In the U.S., this financial promotion and the products and services it describes is directed only at recipients who are qualified eligible persons as defined under CFTC rule 4.7. In the United Kingdom, the Middle East and Australia, this financial promotion and the products and services it describes is directed only at recipients who are professional clients or eligible counterparties; retail customers may not rely on it. Past performance is not necessarily a guide to future performance. Investors may not get back the value of their original investment. ISAM (USA) LLC is registered and regulated by the commodities financial trading commission (CFTC). Pursuant to an exemption from the commodities futures trading commission in connection with accounts of qualified eligible persons, this brochure or account document is not required to be, and has not been, filed with the commission. The commodities futures trading commission does not pass upon the merits of participating in a trading program or upon the adequacy or accuracy of commodity trading advisor disclosure. Consequently, the commodity futures trading commission has not reviewed or approved this trading program or this brochure or account statement. ISAM (EUROPE) LLP is authorized and regulated by the financial conduct authority. ISAM Middle East Ltd is regulated by the Dubai Financial Services Authority. ISAM Asia pacific Pty Ltd is regulated by the Australian Securities & Investments Commission. Special considerations for Switzerland based investors: The Representative of the Fund in Switzerland is OpenFunds Investment Services AG, with its registered office at Seefeldstrasse 35, CH-8008 Zurich, Tel , Fax , The Paying Agent in Switzerland is is Società Bancaria Ticines SA, Piazza Collegiata 3, CH 6500 Bellinzona Tel , Fax The distribution of Shares in the Fund in Switzerland must be made exclusively to Qualified Investors. The place of performance and jurisdiction for the Shares in the Fund distributed in Switzerland is at the registered office of the Representative. Publications to Swiss investors in respect of the Fund are effected by the Representative. CONFIDENTIAL - DO NOT DISTRIBUTE 10