Applying Prospect Theory to Crypto Valuation and Portfolio Diversification

As cryptocurrencies become mainstream and gain acceptance, there is still no coherent investment framework for valuing them. Reference [1] explores the differences between equity and crypto investors and proposes an investment framework for cryptocurrencies based on the prospect theory.

The differences between equity and crypto investors are:

• Stock market investors typically rely on fundamental analysis, examining financial statements, market position, and industry trends to make informed decisions. In contrast, cryptocurrency investors often prioritize technological innovation and potential rapid appreciation, leading to greater volatility and asymmetry in returns compared to equities.
• Stock markets have a longer history and stricter regulations, resulting in more stable investor behavior. Meanwhile, the relatively new and less regulated cryptocurrency market experiences extreme volatility and speculative trading.
• Stock investors tend to have a long-term horizon, seeking steady returns and dividends, whereas cryptocurrency investors are often more focused on short-term gains, driven by high volatility. As a result, cryptocurrency markets exhibit more irrational investor behavior.

The paper then develops an investment framework built on a utility function, where crypto investors remain risk-averse when anticipating gains. However, investor risk attitudes shift during losses; they become risk-seeking in pursuit of recovery. The authors pointed out,

The results in Fig. 2 show the superior ability of our trading strategies to earn abnormal returns. From 2014, each $1 invested in the medium-PL, low-LV portfolio accelerates to $892 at the end of 2022, which is more than four times as in the Fama-French portfolio. While each $1 invested in the low-PL, high-LV portfolio accelerates to $789 at the end of 2022, which is more than three times as in the Fama-French portfolio. The comparison with the S&P 500 index generates similar results; the values of our PL and LV based strategies are much higher than that of the S&P 500.

Similarly, Table 8 shows the average returns, standard deviations, and Sharpe ratios of the portfolios. The medium-PL, low-portfolio and low-PL, high-LV portfolio generate the larger Sharpe ratios (0.411 and 0.396) than those of the equity portfolios, token portfolio, and market benchmarks. The results demonstrate that our trading strategy based on PL and LV with token can also earn superior risk-adjusted returns.

In short, constructing a portfolio that includes both equities and cryptocurrencies using the prospect theory framework results in superior risk-adjusted returns, demonstrating that cryptocurrencies add value to an equity portfolio.

Let us know what you think in the comments below or in the discussion forum.

References

[1] Zhan Wang, Xiang Gao, Jiahao Gu, Can cryptocurrencies improve portfolio diversification? Evidence from the prospect risk perspective, Research in International Business and Finance, Volume 76, April 2025, 102828

Post Source Here: Applying Prospect Theory to Crypto Valuation and Portfolio Diversification

source https://harbourfronts.com/applying-prospect-theory-crypto-valuation-portfolio-diversification/

Measuring Stock Market Dispersion: The Herd Behavior Index Approach

Dispersion in the stock market refers to the degree of variation in individual stock returns within an index or sector. High dispersion indicates significant differences in performance among stocks. Conversely, low dispersion suggests that stocks are moving more uniformly, often driven by broad market trends.

Usually, dispersion is measured by implied correlation. Reference [1] proposed a method to measure dispersion called the Herd Behavior Index (HIX). It is calculated as the ratio of the variance of a stock index to the variance of a hypothetical index that represents the extreme case of comonotonicity or perfect herd behavior. The variance is determined using model-free methods involving options, similar to the calculation of the VIX. The authors pointed out,

In this paper we made a modest contribution to this complicated matter by proposing a measure for the degree of co-movement or herd behavior present in equity markets. This measure compares the currently observed market situation with the comonotonic situation under which the whole system is driven by a single factor. More precisely, it compares an estimate of the variance of the market index with an estimate of the corresponding worst-case or comonotonic variance. In line with the VIX methodology, the estimate for the variance of the market index is based on the full spectrum of current option information on the index. Although the worst-case market situation is not observed, the comonotonic variance can easily be determined from the option prices on the constituents of the market index.

In short, the authors developed the Herd Behavior Index to measure stock market dispersion. They also explained how it differs from the implied correlation index,

Measuring the degree of co-movement with the HIX/CIX has several advantages compared to implied correlation. The HIX/CIX is able to capture all kinds of dependences between stock prices, whereas the implied correlation is a weighted average of pairwise correlations amongst the asset returns and hence, only focuses on linear dependences. Furthermore, making abstraction of the approximations involved in its calculation, the HIX reaches its maximal value of 1 if and only if the underlying random variables are comonotonic. On the other hand, there is no direct link between the degree of herd behavior and the value of the implied correlation.

This is an innovative proposal, but its practical application and effectiveness remain to be seen. One can apply it, for example, to option dispersion trading. A high HIX value suggests buying individual options and selling index options. The position can then be closed when the market stabilizes and the HIX decreases. Further research is needed to assess the profitability of this strategy and the effectiveness of the Herd Behavior Index in general.

Let us know what you think in the comments below or in the discussion forum.

References

[1] Jan Dhaene, Daniël Lindersy, Wim Schoutensz, David Vyncke, The Herd Behavior Index: A new measure for the implied degree of co-movement in stock markets, Insurance: Mathematics and Economics Volume 50, Issue 3, May 2012, Pages 357-370

Article Source Here: Measuring Stock Market Dispersion: The Herd Behavior Index Approach

source https://harbourfronts.com/measuring-stock-market-dispersion-herd-behavior-index-approach/

Technical Trading Strategies: Profitable or Just Curve-Fitting?

Technical analysis (TA) is a trading approach that evaluates past price movements and volume data to forecast future market trends. It relies on price patterns, indicators, and statistical measures rather than fundamental factors. Investors use tools like moving averages, RSI, and Bollinger Bands to identify trends, support and resistance levels, and potential entry or exit points.

The growing proliferation of academic articles on technical analysis suggests its acceptance and effectiveness. However, some researchers continue to question its validity. Reference [1] revisits the question of whether TA works, testing TA rules on three stocks—AAPL, MSFT, and NVDA—from January 2000 to December 2022. The study evaluates the profitability of various technical trading strategies both in-sample and out-of-sample using methods such as reality checks and stepwise tests.

The authors pointed out,

In this paper, we analyze a comprehensive dataset of AAPL, MSFT, and NVDA stocks from January 2000 to December 2022 to evaluate the profitability of various technical trading strategies both in-sample and out-of-sample, using February 2016 as the primary cutoff and May 2018 as an alternative. We construct strategies based on multiple indicators and timeframes, conducting thorough statistical analyses to ensure robustness against data-snooping bias. Our results consistently demonstrate that apparent profitability often stems from parameter selection rather than true market inefficiencies, supporting the efficient market hypothesis. This highlights the difficulty in predicting profitable strategies ahead of time, emphasizing the unpredictable nature of achieving sustained trading success.

In short, the article concludes that TA does not work, as it fails to identify any technical trading strategies that yield consistent profits across both periods. The results consistently show that apparent profitability often arises from parameter selection rather than genuine market inefficiencies, supporting the efficient market hypothesis.

We welcome this type of research that challenges prevailing beliefs. However, in our opinion:

• The sample size is small, covering only three stocks,
• While the examined period is long, market dynamics may have changed over time. Using an extended dataset is beneficial, but a trading system should account for shifts in market conditions,
• Only two types of tests were conducted; more comprehensive testing is needed.

That said, we look forward to seeing further research in this area. Let us know what you think in the comments below or in the discussion forum.

References

[1] Wang, Y., Chen, Y., Tian, H., & Wayne, Z. (2025). Evaluating Technical Trading Strategies in US Stocks: Insights From Data-Snooping Test. Journal of Accounting and Finance, 25(1).

Article Source Here: Technical Trading Strategies: Profitable or Just Curve-Fitting?

source https://harbourfronts.com/technical-trading-strategies-profitable-just-curve-fitting/

Breaking Down the Volatility Risk Premium: Overnight vs. Intraday Returns

The decomposition of the volatility risk premium (VRP) into overnight and intraday components is an active area of research. Most studies indicate that the VRP serves as compensation for investors bearing overnight risks.

Reference [1] continues this line of research, with its main contribution being the decomposition of the variance risk premium into overnight and intraday components using a variance swap approach. The study also tests the predictive ability of these components and examines the seasonality (day-of-week effects) of the VRP.

The authors pointed out,

This paper uses the P&L of a hypothetical variance swap position and breaks down the variance risk premium into its overnight and intraday components to provide some further insights on the nature of the variance risk premium. In the empirical analysis we use implied variance stock indices in US, Europe and Asia and find that the variance risk premium switches sign between overnight and intraday periods. During the overnight period the variance risk premium is negative and during the intraday trading period the variance risk premium becomes positive. Our findings suggest that the negative variance risk premium reported in numerous previous studies is primarily driven by the overnight period variance risk premium component. We also evaluate the ability of the intraday and overnight variance risk premium to predict future equity returns. We find that that the intraday component captures short-term risk and displays predictive ability at 1–3-month horizons, while the overnight component reflects longer term risk and displays predictive ability at 6-12-month horizons.

In summary, the study reaffirms that the variance risk premium is significantly negative during the non-trading overnight period, while it becomes positive and often insignificant during the intraday trading period.

An interesting finding is the day-of-week seasonality. For instance, going long volatility at the open and closing the position at the close tends to be profitable on most days, except Fridays.

Let us know what you think in the comments below or in the discussion forum.

References

[1] Papagelis, Lucas and Dotsis, George, The Variance Risk Premium Over Trading and Non-Trading Periods (2024). https://ift.tt/7zVDyfm

Originally Published Here: Breaking Down the Volatility Risk Premium: Overnight vs. Intraday Returns

source https://harbourfronts.com/breaking-volatility-risk-premium-overnight-vs-intraday-returns/

Using ChatGPT to Extract Market Sentiment for Commodity Trading

A Large Language Model (LLM) is an advanced AI system trained on vast amounts of text data to understand, generate, and analyze human language. In finance, LLMs are used for tasks like analyzing earnings reports, generating market sentiment analysis, automating financial research, and enhancing algorithmic trading strategies. They can process vast amounts of unstructured data, such as news articles, SEC filings, and analyst reports, to extract insights and identify patterns that may impact markets.

In this context, Reference [1] examines the effectiveness of ChatGPT in predicting commodity returns. Specifically, it extracts commodity news information and forecasts commodity futures returns. The study gathers over 2.5 million articles related to the commodity market from nine international newspapers across three countries, covering a diverse set of 18 commodities.

ChatGPT-3.5 is then used to assess the sentiment of these articles by analyzing their headlines, abstracts, or body content, classifying the news as either good or bad for the commodity market. It then constructs the Commodity News Ratio Index (CNRI) and conducts a comprehensive set of tests to evaluate its forecasting efficacy for commodity futures returns.

The authors pointed out,

This paper develops a text‐based CNRI using ChatGPT, derived from a comprehensive analysis of over 2.5 million articles sourced from nine newspapers. Our empirical results demonstrate that the ChatGPT‐based CNRI is significantly effective in forecasting 1‐ to 12‐month accumulated commodity futures index excess returns, as evidenced by both in‐sample and out‐of‐sample regression analyses. Moreover, we control for business variables and economic indicators to examine the significance of predictability.

Our analysis further reveals that the CNRI exhibits enhanced predictive power during periods of economic expansion, contango markets, and declining inflation. We also confirm that our constructed CNRI contains valuable predictive information regarding future macroeconomic performances. In summary, our study illustrates that ChatGPT can effectively predict trends in the commodity market, enhancing our understanding of the information processing capabilities of LLMs and their implications for investors in the financial market.

In short, ChatGPT proves useful in forecasting commodity market dynamics and provides valuable insights for investors and risk managers.

Let us know what you think in the comments below or in the discussion forum.

References

[1] Shen Gao, Shijie Wang, Yuanzhi Wang, Qunzi Zhang, ChatGPT and Commodity Return, Journal of Futures Markets, 2025; 1–15

Post Source Here: Using ChatGPT to Extract Market Sentiment for Commodity Trading

source https://harbourfronts.com/using-chatgpt-extract-market-sentiment-commodity-trading/

Market Reactions to Corporate Earnings Under Different Volatility Regimes

Earnings announcements impact a company's stock volatility. Reference [1] examined the earnings reports of Dow Jones Industrial Average components over 10 years to better understand the relationship between different volatility regimes and the time it takes for earnings information to be fully absorbed by the market.

The study focuses on the past ten years, a period that includes three distinct volatility regimes, multiple economic cycles, and shifts from low to rising and ultimately declining interest rates. These evolving market conditions provide a diverse backdrop for assessing how corporate earnings disclosures influence investor behavior and price discovery.

The authors pointed out,

Our analysis as shown in Table 3, confirms the initial hypothesis that the time required for a firm’s earnings or performance news to be fully incorporated into its stock price depends substantially on the prevailing volatility regime, as identified by the SETAR model. Under high-volatility conditions, investors process information more swiftly, leading to shorter impulse response functions (IRFs) being derived from the VAR model. Conversely, information diffuses slower in low-volatility markets, resulting in extended IRFs. While most stocks among the Dow 30 revert to baseline levels within 3–5 days, technology and financial firms often display more drawn-out fluctuations due to their heightened sensitivity to market sentiment and macroeconomic indicators.

…Higher-beta stocks, such as Boeing (BA) and Goldman Sachs (GS), exhibited more significant and persistent price swings, even in conditions where rapid market absorption would otherwise be expected. This phenomenon is especially evident during lower-volatility regimes, suggesting that firm-specific risk factors can prolong the market’s adjustment process despite a relatively subdued overall sentiment.

In short, the time required for earnings news to be fully reflected in stock prices depends on the prevailing volatility regime. In high-volatility markets, investors process information faster, while in low-volatility conditions, information diffuses more slowly. Higher-beta stocks show more persistent price swings, especially in low-volatility environments.

The authors also discuss the policy implications of their findings regarding the frequency of financial reporting,

…frequent reporting can enhance market transparency and amplify short-term volatility and sector-specific risks. Policymakers could exempt specific high-volatility sectors from mandatory frequent reporting to balance transparency with market stability.

Let us know what you think in the comments below or in the discussion forum.

References

[1] Ugras, Y. J., & Ritter, M. A. (2025). Market Reaction to Earnings Announcements Under Different Volatility Regimes. Journal of Risk and Financial Management, 18(1), 19.

Article Source Here: Market Reactions to Corporate Earnings Under Different Volatility Regimes

source https://harbourfronts.com/market-reactions-corporate-earnings-different-volatility-regimes/

Detecting Trends and Risks in Crypto Using the Hurst Exponent

The Hurst exponent is a statistical measure used to assess the long-term memory and persistence of a time series. It quantifies the tendency of a system to revert to the mean, follow a random walk, or exhibit a trending behavior. A Hurst exponent (H) value between 0 and 0.5 indicates mean-reverting behavior, H = 0.5 suggests a purely random process, and H between 0.5 and 1 signals persistent, trending behavior.

In finance, the Hurst exponent is widely used to analyze market efficiency, detect trends, and evaluate the predictability of asset prices. In this context, Reference [1] utilized the Detrended Fluctuation Analysis technique to study the Hurst exponent of the five major cryptocurrencies. Its main novelty is the calculation of a weekly time series of the Hurst exponent and its analysis.

The authors pointed out,

• Firstly, the Hurst exponent (H) can be utilized to monitor trend continuation or reversal. In our analysis, transitions between these regimes were observed in some cryptocurrencies, such as XRP, which displayed short-term persistence followed by long-term anti-persistence. These shifts could potentially serve as early indicators of trend changes. For instance, monitoring rolling-window DFA estimates over time could help identify when a cryptocurrency market transitions from trend-following (H>0.5) to mean-reverting behavior (H<0.5), aiding in dynamic strategy adjustments to enhance decision-making.
• Secondly, the study highlights distinct asset-specific behavioral characteristics across cryptocurrencies. The heterogeneous behaviors observed suggest that H-based analysis could inform tailored trading strategies for different assets…
• Lastly, the observed synchronization in H values across multiple cryptocurrencies during extreme market events offers potential for systemic risk monitoring. For instance, collective shifts to anti-persistent behavior (H<0.5) may signal heightened volatility or market instability, enabling traders to adjust portfolios or implement defensive measures such as diversification.

In short, the findings suggest opportunities for using Hurst exponents as tools to monitor trend continuation or reversal, develop asset-specific strategies, and detect systemic risks during extreme market conditions, offering valuable insights for traders and policymakers navigating the cryptocurrency market's inherent volatility.

This is a useful application of the Hurst exponent, and it is not limited to cryptocurrencies but can be applied to any market.

Let us know what you think in the comments below or in the discussion forum.

References

[1] Huy Quoc Bui, Christophe Schinckus and Hamdan Amer Ali Al-Jaifi, Long-Range Correlations in Cryptocurrency Markets: A Multi-Scale DFA Approach, Physica A: Statistical Mechanics and its Applications, (2025), j.physa.2025.130417

Post Source Here: Detecting Trends and Risks in Crypto Using the Hurst Exponent

source https://harbourfronts.com/detecting-trends-risks-crypto-using-hurst-exponent/

How End-of-Month Returns Predict the Next Month’s Performance

Calendar anomalies in the stock market refer to patterns where stock returns deviate from expected behavior based on the time of year, month, or even day. Well-documented anomalies include the "Sell in May and Go Away" strategy, which suggests that stocks underperform during the summer months, and the "Turn-of-the-Month Effect," where returns are typically higher in the last few days of one month and the first few days of the next.

Reference [1] introduced a novel calendar anomaly known as the end-of-month reversal effect. The study showed that end-of-month returns, i.e. returns from the fourth Friday to the last trading day of the month, are negatively correlated with returns in the following month. The author pointed out,

This paper documents a novel 1-month aggregate market reversal pattern. This pattern is driven by the previous end-of-the-moth market return. The empirical evidence is statistically significant both In- and Out-of-Sample. Importantly, I show that the reversal at the aggregate level has characteristics opposite to those established in the cross-sectional literature: it concentrates on high-priced and liquid stocks and is cyclical with the economy. Consequently, a simple rule of thumb and more sophisticated strategies deliver sizable economic gains.

In short, a simple trading strategy based on this effect, that is buying if the end-of-month return is negative and selling if it is positive, outperforms the buy-and-hold strategy over a 45-year period.

The author also provides an explanation for this anomaly, attributing it to pension funds’ liquidity trading, as they adjust their portfolios to meet pension payment obligations.

I rationalize the empirical findings via pension funds’ end-of-the-month liquidity trading. Leveraging on recent findings and on direct evidence from daily pension funds trading activity, I argue that the payment cycle potentially triggers a non-informational trading shock. Consistent with a payment cycle explanation, I show that the reversal pattern increases in absolute terms within the one month ahead, aligning with the time pension funds receive inflows.

Let us know what you think in the comments below or in the discussion forum.

References

[1] Graziani, Giuliano, Time Series Reversal: An End-of-the-Month Perspective, 2024, https://ift.tt/3Qoxzpe

Post Source Here: How End-of-Month Returns Predict the Next Month’s Performance

source https://harbourfronts.com/end-month-returns-predict-next-months-performance/