Long-only Strategy Research Articles

An analysis of conditional mean-variance portfolio performance using hierarchical clustering

This paper studies portfolio optimization through improvements of ex-ante conditional covariance estimates. We use the cross-section of stock returns over a 52-year sample to analyze trading performance by implementing the machine learning algorithm of hierarchical clustering. We find that higher out-of-sample risk-adjusted returns are achieved relative to the traditional Markowitz portfolio through hierarchical clustering using a 3-month buy-and-hold, long-only strategy. Additionally, the average change in portfolio weights at each rebalancing period is significantly lower for the portfolio formed using machine learning relative to Markowitz, decreasing investor trading costs. The results are robust to various settings and subsamples.

Comparison of Long-term Strategies Versus Long-short Combination Strategies in Different Industries

During the COVID-19 pandemic, the prospects for the development of the real economy are not very clear. Every investment decision an investor makes in this environment will bring different benefits to the investor. Comparing only long and long combinations allows investors to judge according to their own different situations what investment methods can make them more profitable or reduce the risk of investment. In Sina Finance, Investors obtained stock information on Li Ning, ANTA, Maotai, GUJING, Tencent and Power Assets Holdings Limited. Investors divided these stocks into three groups, and implemented a long-only strategy and a long-short combination strategy for them, and the final result made investors know that the long-short combination can alleviate some of the damage caused by only long, but at the same time, it may bring more pressure to investors. However, it does not specifically mean that a certain strategy can help investors get 100% returns, and investors need to choose the most suitable investment strategy according to their own situation and investment environment. That is, investors cannot blindly pursue interests and choose risks, nor can investors only pursue low risks and abandon interests, the two need to complement each other and make common progress, so as to bring investors the most suitable investment decisions.

Tax-Efficient Diversification of a Concentrated Portfolio through Margin and Shorting

We look at the rewards and risks of two tax-efficient strategies for diversifying a 10-stock basket of appreciated stocks in a separately managed account. Our long-only strategy requires an upfront partial liquidation whose after-tax proceeds are used to fund a diversified loss-harvesting portfolio. Losses are used to lower the weight of the remaining concentrated positions’ tax neutrally. Our margin-and-shorting strategy complements the 10-stock basket with a dollar-neutral loss-harvesting portfolio. An upfront liquidation is not required, although steep financing costs are associated with short positions in the loss-harvesting portfolio. Examining these strategies over many historical periods, we find that both diversification strategies typically generated portfolios with forecast tracking errors at roughly 1% on average at a 10-year horizon. The margin-and-shorting strategy delivered more tax alpha than the long-only strategy, however, for an investor with an initial investment of at least $10 million.

Real Estate ETNs in Strategic Asset Allocation

Previous research has shown that real estate serves as a diversifier in mixed-asset portfolios. However, this empirical finding is beset with some drawbacks associated with direct real estate investment. In order to overcome some of these drawbacks, we use theoretical real estate exchange traded notes (ETNs) in a mean-shortfall setting to optimize an international mixed-asset portfolio. In addition, the typical long-only strategy is abandoned in favor of a 130/30 long-short and an exchange rate hedge strategy. Not only in-sample but also out-of-sample portfolios yield significant diversification benefits by means of real estate ETNs in different portfolio strategies.

Stock market forecasting using a multi-task approach integrating long short-term memory and the random forest framework

Numerous studies have adopted deep learning (DL) in financial market forecasting models owing to its superior performance. The DL models require as many relevant input variables as possible to improve performance because they learn high-level features from these inputs. However, as the number of input variables increases, the number of parameters also increases, leaving the model prone to overfitting. We propose a novel stock-market prediction framework (LSTM–Forest) integrating long short-term memory and random forest (RF) to address this issue. We also develop a multi-task model that predicts stock market returns and classifies return directions to improve predictability and profitability. Our model is interpretable because it can identify key variables using the variable importance analysis from RF. We used three global stock indices and 43 financial technical indicators to verify the proposed methods experimentally. The root mean squared errors of LSTM–Forest with multi-task (LFM) in predicting returns from S&P500, SSE, and KOSPI200 were 25.53%, 22.75%, and 16.29% lower, respectively, than those of the baseline RF model. The model’s balanced accuracy in forecasting the daily return direction increased by 7.37, 1.68, and 3.79 points, respectively. Furthermore, our multi-task model outperforms our single-task model and previous DL approaches. In the trading test, LFM produced the highest profits—even compared with the long-only strategy when transaction costs were considered. The proposed framework is readily extensible to other tasks and fields that contend with high-dimensionality problems.

How Valuable Are Target Price Forecasts to Factor Investing?

The informativeness of financial analysts’ stock recommendations to investors has become the subject of much scrutiny and debate. This article investigates the investment value of target price (TP) forecasts under the specific angle of factor investing. It provides directions to quantitative portfolio managers for the integration of signals conveyed by those forecasts in systematic and rule-based investment decisions. Using TP data from 1999 to 2019 in Europe and North America, the author first documents a decline in analyst opinion on the performance of most equity factors and finds that their forecasts are style biased. He shows that a long-only strategy that picks best-ranked stocks according to consensus TP implied expected returns (CTPER) generates substantial alpha within large-cap, low-idiosyncratic-risk, and noncyclical stocks. Embedding the CTPER signal into the momentum factor helps materially improve its risk and return profile. Those findings support the idea of adequately using TP forecasts in the design of top-down multifactor portfolio construction frameworks.

Optimal Bitcoin trading with inverse futures

We consider an optimal trading problem for an investor who trades Bitcoin spot and Bitcoin inverse futures, plus a risk-free asset. The investor seeks an optimal strategy to maximize her expected utility of terminal wealth. We obtain explicit solutions to the investor’s optimal strategies under both exponential and power utility functions. Empirical studies confirm that optimal strategies perform well in terms of Sharpe ratio and Sortino ratio and beat the long-only strategy in Bitcoin spot.

Cryptocurrency Factor Portfolios: Performance, Decomposition and Pricing Models

The empirical distributions of cryptocurrency returns are highly non-normal, casting doubt on the performance metrics. So we apply almost stochastic dominance (ASD), which does not require any assumption about the return distribution, to examine cryptocurrency factor portfolios. Using portfolios based on factors that can be constructed from available market information, we find 13 factor portfolios that dominate our four benchmarks. The long-only strategy contributes more to this dominance than does the short-only strategy. We test whether returns on the 13 dominant factor portfolios can be explained by a coin market three-factor model. This model has limited success, and its performance is significantly improved by the inclusion of a mispricing factor.

Tax-Efficient Portfolio Transition: A Tax-Aware Relaxed-Constraint Approach to Switching Equity Managers

For a taxable investor with a highly appreciated equity portfolio, replacing the portfolio manager is likely to trigger substantial tax liabilities. We focus on transitioning an appreciated equity portfolio to an actively managed strategy. We compare transition from an appreciated portfolio to a traditional long-only tax-agnostic equity strategy with transition to equity strategies utilizing more-advanced portfolio management techniques such as tax-aware rebalancing and relaxed-constraint portfolio construction. We find that transition to a tax-aware relaxed-constraint strategy results in both high implementation efficiency and tax efficiency both during and after the transition. As a result, a tax-aware, relaxed-constraint, post-transition strategy significantly outperforms a traditional tax-agnostic, long-only strategy in its ability to preserve and grow the investor’s after-tax wealth over the long term. We also discuss risks and limitations of the tax-aware, relaxed-constraint approach. <b>TOPICS:</b>Wealth management, equity portfolio management <b>Key Findings</b> ▪ For an individual taxable investor with an appreciated equity portfolio, the tax costs of replacing the portfolio manager may be highly punitive, far outweighing the transaction costs of such replacement. ▪ Advanced portfolio management techniques can help alleviate the tax burden of portfolio transition resulting from manager replacement: We show that transition to a tax-aware relaxed-constraint strategy results in high implementation efficiency and tax efficiency both during and after the transition, leading to strong after-tax performance. ▪ We also show that tax-aware portfolio transition is a complex bespoke solution and stress the need for a careful evaluation of the client-specific situation by a client’s investment advisor and a prospective manager. Nonetheless, our results demonstrate that a tax-aware transition to a relaxed-constraint strategy has the potential for providing a substantial benefit to a taxable investor “locked-in” into a highly appreciated portfolio.

Implementing a Systematic Long-only Momentum Strategy: Evidence From India

We show that a monthly-rebalanced, long-only portfolio of top-decile stocks selected from the NIFTY100 using `off-the-shelf' momentum criteria significantly outperforms the NIFTY100 Index - both in terms of absolute returns (by 10.70% pa) and risk adjusted returns, with a mean turnover of 32.10% per month. We show that momentum persists in the near term but dissipates over time. We demonstrate that our long-only approach has a significant tilt to the momentum factor. We also show that time in the market rather than timing the market is important for momentum investing. The strategy has higher volatility and the occasional momentum crash. The strategy's out performance survives real-world implementation given the rise of discount brokers in India. In the absence of cheap ETFs to get exposure to momentum, the systematic long-only strategy from the most liquid part of the market using `off-the-shelf' criteria provides a practical, executable investment methodology that exposes an investor to momentum in the Indian market.

Tax-Efficient Portfolio Transition: &lt;i&gt;A Tax-Aware Relaxed-Constraint Approach to Switching Equity Managers&lt;/i&gt;

For a taxable investor with a highly appreciated equity portfolio, replacing the portfolio manager is likely to trigger substantial tax liabilities. We focus on transitioning an appreciated equity portfolio to an actively managed strategy. We compare transition from an appreciated portfolio to a traditional long-only tax-agnostic equity strategy with transition to equity strategies utilizing more advanced portfolio management techniques such as tax-aware rebalancing and relaxed-constraint portfolio construction. We find that transition to a tax-aware relaxed-constraint strategy results in both high implementation efficiency and tax efficiency both during and after the transition. As a result, a tax-aware relaxed-constraint post-transition strategy significantly outperforms a traditional tax-agnostic long-only strategy in its ability to preserve and grow the investors after-tax wealth over the long term. We also discuss risks and limitations of the tax-aware relaxed-constraint approach.

Implementing a Systematic Long-only Quality Strategy in the Indian Market

We believe investors should be willing to pay a higher price for higher quality companies. We build a composite quality score, including one which incorporates publicly available ESG scores, using 'off-the-shelf' criteria and publicly available financial data and show that a quarterly-rebalanced, long-only portfolio of top-decile stocks selected using our score in Hong Kong significantly outperforms the Hang Seng Composite LargeCap Index (HSLI) - both in terms of absolute returns (by 2.87% pa) and risk adjusted returns - while having an moderate annual turnover (a mean turnover of 42.15%). We show that our quality score predicts the persistence of quality for up to 3 years and there is a weak relationship between the price multiple and the quality score. Furthermore, we demonstrate that quality needs to be reviewed regularly - and the moderate turnover ensures this. In the absence of cheap ETFs to get systematic exposure to quality, the systematic long-only strategy using 'off-the-shelf' criteria provides a practical, executable systematic investment methodology that exposes an investor to quality in the Hong Kong market. We show that socially conscious investors can generate risk adjusted returns in excess of the HSLI in the Hong Kong market using a systematic strategy.

Implementing a Systematic Long-only Quality Strategy in the Hong Kong Market

We believe investors should be willing to pay a higher price for higher quality companies. We build a composite quality score, including one which incorporates publicly available ESG scores, using 'off-the-shelf' criteria and publicly available financial data and show that a quarterly-rebalanced, long-only portfolio of top-decile stocks selected using our score in Hong Kong significantly outperforms the Hang Seng Composite LargeCap Index (HSLI) - both in terms of absolute returns (by 2.87% pa) and risk adjusted returns - while having an moderate annual turnover (a mean turnover of 42.15%). We show that our quality score predicts the persistence of quality for up to 3 years and there is a weak relationship between the price multiple and the quality score. Furthermore, we demonstrate that quality needs to be reviewed regularly - and the moderate turnover ensures this. In the absence of cheap ETFs to get systematic exposure to quality, the systematic long-only strategy using 'off-the-shelf' criteria provides a practical, executable systematic investment methodology that exposes an investor to quality in the Hong Kong market. We show that socially conscious investors can generate risk adjusted returns in excess of the HSLI in the Hong Kong market using a systematic strategy.

The Specialness of Equally-Weighted Portfolios: The Rebalancing Fallacy

There is overwhelming evidence in equity markets that equally-weighted (𝐸𝑊) portfolios deliver higher average returns and higher Sharpe ratios than value-weighted (𝑉𝑊) portfolios. Rebalancing to fixed weights has typically been advanced as an argument to explain the overperformance of 𝐸𝑊 portfolios. The paper challenges the rebalancing to fixed-weights hypothesis and provides empirical evidence that the overperformance is entirely explained by portfolio volatility and that there is no additional contribution to returns from rebalancing. Conditionally, the long-short 𝐸𝑀𝑉 portfolios, (Equally Minus Value weighted), which are long 𝐸𝑊 portfolios and short 𝑉𝑊 portfolios are shown to overperform only to the extent that they are invested in high variance or in low market capitalization securities. Furthermore, when looking at their dynamics, the overperformance is observed only in a few states of the world. The paper identifies the relevant state variables and documents evidence of predictability using information contained in lagged returns and in lagged volatility. Both variables are found to predict the 𝐸𝑀𝑉 portfolio returns positively. The information is exploited to test simple dynamic strategies which allow the 𝐸𝑀𝑉 portfolio to be held long or short depending on the signals contained in the two conditioning variables. The Sharpe ratios of various long-short dynamic strategies significantly overperform the long-only strategy. When looking at risk-adjusted returns, factor models fail to explain the 𝐸𝑀𝑉 portfolio returns. However, 𝐸𝑀𝑉 portfolios display pricing performance by driving well known factors to statistical insignificance. More generally, the unconditional and the conditional results obtained for the 𝐸𝑀𝑉 portfolios are shown to depend on two cross-sectional correlations, the first between the securities’ active weights and their variances and the second between the securities’ active weights and their growth rates. The cross-sectional correlations are shown to be security and state dependent. The average return on the 𝐸𝑀𝑉portfolio can be positive or negative depending on which one of the two cross-sectional correlations dominates.

A Practical Machine Learning Approach for Dynamic Stock Recommendation

Stock recommendation is vital to investment companies and investors. However, no single stock selection strategy will always win while analysts may not have enough time to check all S&P 500 stocks (the Standard & Poor’s 500). In this paper, we propose a practical scheme that recommends stocks from S&P 500 using machine learning. Our basic idea is to buy and hold the top 20% stocks dynamically. First, we select representative stock indicators with good explanatory power. Secondly, we take five frequently used machine learning methods, including linear regression, ridge regression, step wise regression, random forest and generalized boosted regression, to model stock indicators and quarterly log-return in a rolling window. Thirdly, we choose the model with the lowest Mean Square Error in each period to rank stocks. Finally, we test the selected stocks by conducting portfolio allocation methods such as equally weighted, mean variance, and minimum-variance. Our empirical results show that the proposed scheme outperforms the long-only strategy on the S&P 500 index in terms of Sharpe ratio and cumulative returns.

Implementing a Systematic Long-only Quality Strategy in the Singapore Market

Implementing a Systematic Long-only Quality Strategy in the Singapore Market

Optimal Bitcoin Trading with Inverse Futures

We consider an optimal trading problem for an investor who trades Bitcoin spot and Bitcoin inverse futures, plus a risk-free asset. The investor seeks an optimal strategy to maximize her expected utility of terminal wealth. We obtain explicit solutions to the investor's optimal strategies under both exponential and power utility functions. Empirical studies con rm that optimal strategies perform well in terms of Sharpe ratio and Sortino ratio and beat the long-only strategy in Bitcoin spot.

Option Informed Stock Picking

A wide range of research has suggested that informed trading in options markets may effectively signal subsequent changes in equity prices. In this article, the authors analyze the performance of long/short strategies based on a number of signals from options markets. In addition, they create an easily implemented long-only strategy based on a subset of the signals (volatility risk premium, option/stock volume ratio, implied volatility skew, and realized volatility). In order to minimize transaction costs and liquidity issues, they restrict their analysis to S&P 500 constituents, rebalance the portfolio monthly, and limit the holdings to 50 individual stocks. The analysis of the period from 1996 through mid-2015 shows significant outperformance of a long-only, equal-weighted portfolio of 50 stocks (and found similar results when considering 10-stock portfolios), relative to the S&P 500 and the equal-weighted S&P 500. A return attribution analysis confirms that the outperformance is provided by individual stock selection rather than sector selection.

Predicting daily oil prices: Linear and non-linear models

In this paper, we assess the accuracy of linear and nonlinear models in predicting daily crude oil prices. Competing forecasts of crude oil prices are generated from parsimonious linear models which require no parameter estimation, as well as linear and nonlinear models. Two of the linear models that we employ exploit the informational content of oil demand and the increasing correlation between oil and equity prices and are novel to the literature. The nonlinear model that we consider is an artificial neural network. More specifically, we consider a bagged neural network, a neural network trained using the genetic algorithm as well as a neural network with fuzzy logic. We find that some of the linear models outperform the random walk in terms of out-of-sample statistical forecast accuracy. Our findings also suggest that while the buy-and-hold strategy dominates some of the models in terms of dollar payoffs and risk-adjusted returns under a long-only strategy, all the models that we consider generate higher dollar payoffs than the buy-and-hold strategy under the short-only strategy. An investor obtains the largest profits by trading based on the moving average convergence divergence which is a technical indicator.

Bitcoin, Portfolio Diversification and Chinese Financial Markets

This research explores the effects of adding bitcoin to an optimal portfolio (naive, long-only, unconstrained and semi-constrained) by relying on mean-CVaR in the Chinese market. Then backtesting to compare the performance of portfolios with and without bitcoin for each scenario is perfomed. Results show significant but weak correlations between various asset classes and bitcoin, implying a more mature financial profile of bitcoin in China compared to that in the west. Backtesting results show that the effect of adding bitcoin to optimal portfolios is not consistent over the entire out-of-sample period. The naive and the long-only strategy improved the risk-reward ratio up until the late 2013 price-crash with no significant advantages thereafter. Shorting strategies on the other hand, with or without leverage, fail to produce more efficient portfolios when bitcoin is added, and this is consistent over the entire out-of-sample period. The results also show that semi-annual rebalancing amplifies the advantages of adding bitcoin to most portfolios except for the semi-constrained portfolio, although the weights analysis show significant shifts in weights which might not represent a feasible strategy in realistic scenarios.