Investor Sentiment and Fund Market Anomalies: Evidence from Closed-end Fund, Exchange-traded Fund and Real Estate Investment Trust

Abstract

The investor sentiment hypothesis has become a promising avenue by way of a behavioural approach to complementing conventional explanations of financial market anomalies. In
response to the problems exhibited in the existing theories, the investor sentiment hypothesis has been widely tested and the results of which turn out to be able to successfully explain the market anomalies to a great extent. The thesis applies the investor sentiment theory to analysing the fund anomalies in both the UK and US markets. The test results and their interpretations may help promote a better understanding of the investor sentiment and its impacts including their geographical differences. We contribute to the literature by focusing on the sentiment measures, among others. Since the investor sentiment reflects the investors’
behaviour and psychology, it is hard to be properly captured. We have constructed the proxies for the sentiment factor in both direct and indirect forms.

The first fund anomaly we analysed is the “closed-end fund puzzle”. The puzzle is so-called because at IPO, the fund is issued at a premium to the net asset value (NAV); however, thispremium disappears in the next few months. The fund then trades at a discount. This discount is not fixed, varying substantially during the closure period. When the closed-end fund is either converted into an open-end fund or liquidated, the discount shrinks and the share price
will rise. We construct an out-of-sample test by using the two-factor and five-factor models. The results show that the investor sentiment can contribute to explaining closed-end fund discounts in the UK market and it is more prevalent in smaller size portfolios. We also find the evidence to support investor sentiment as an important factor to represent systematic risk in the return generating process.

Next, we examine the price deviations of Exchange Traded Funds (ETFs). Unlike closed end funds whose prices also deviate from the NAV, ETFs, through a mechanism known as
redemption in-kind, allow institutional investors to potentially earn a profit by arbitraging away these price deviations through creating and deleting outstanding shares of the ETF. Hence, we are motivated to identify the factors that may impact on the determination of these premiums and discounts to the NAV. We first construct a sentiment proxy from the derivative market variables such as the option put–call trading volume ratio and the open interest ratio. Then we develop a sentiment proxy based on the consumer confidence index, obtained from the mainstream consumer surveys and this proxy is taken to the individual fund level. The results provide evidence that this sentiment proxy has explanatory power for most individual ETF mispricing. We take the whole industry into account and find that the sentiment factor has incremental explanatory power and is positively related to the fund premium. The evidence also shows that more sentiment-sensitive ETFs are those that have smaller, younger and volatile stocks with low dividend yields.

Finally, the thesis considers the fund anomaly in the form of the REIT price momentum. In order to investigate the momentum profitability, we classify the formation period into two sentiment states, i.e. the optimistic and pessimistic periods. Evidence indicates that when
sentiment is high, the REIT momentum profitability is substantial and significant; however, when the sentiment is low, the profits from the REIT momentum are much lower and not significant. We also examine the interplay between REIT liquidity and momentum profitability. We find that high REIT liquidity portfolios generate higher momentum returns,
but this is only significant when the sentiment is optimistic. Furthermore, consistent with our previous findings, our evidence that momentum is generally larger for smaller companies confirms that the size effect is still available in the REIT industry. This is because the smaller
companies are often difficult to value, as they are more prone to subjective evaluations. The sentiment thus could be more significant in small size companies.