New Research from Cass Business School on Investor Sentiment Index

23 May

The franchise for reproducing this article is pending

The latest landmark report to emerge from Cass Business School’s consultancy division has proved it really does pay to ignore crowd mentality in investment decisions. Professors Nikos  and Dr Nomikos Papapostolou presented their report on dry-bulk shipping. They took historical data of boom-and-bust cycles, and found that investor sentiment acts as a contrarian indicator to returns.

Cass academics used regression analysis to determine the correlation between their investor sentiment index, and benchmark vessel prices dating back to 1996. The data set was from several different sources; they used a combination of different variables, and therefore more than one input. They used the supply-demand balance to determine asset valuation. From the same data they extracted proxy values and parameters. To set a value for market expectations, they quantified investment activity. And to measure liquidity they took the volume of trading, From the S&P they took key market parameters, including IPO returns and volatility, based on information regarding the frequency and size of trades, and the positions of speculators and hedgers.  Their deduction? That overwhelmingly, “sentiment is a contrarian predictor of asset prices.”

If the market seems to be moving up, it could be time to start preparing for a fall. Boom-and-bust cycles in the shipping industry shared a number of characteristics. The initial peak, (or trough), is located at the highest (or lowest) point within a five-month window either side of that point. A fall in investor sentiment either coincided with a decline in asset value, or predated it – as was the case in September 2000, April 2005 and July 2008. Each phase has a minimum duration of five months. Once you have exceeded this time period, it could be time to start betting against the market. Perhaps to buy or sell a ship.

Professor Nomikos drew the team’s findings to their logical conclusion, asserting a ‘buy signal’ was when the sentiment index crosses the zero line from above, and the ‘sell signal’ when the index crosses the zero line from below. This being said, they were careful to issue a disclaimer, insisting they were not offering trading advice, but a new, solid, economic indicator. They believe it superior to the pre-existing Moore-Stevens index, because it can be calculated in a monthly basis and does not rely on surveys, which are rarely accurate predictors of real decisions.

Their overall market sentiment indicator was set against to various sector-specific ones: for capesize, paramax, supramax and handy vessels. Regression analysis showed there was no strong correlation between the market and individual sector sentiment. Moreover, between the two points April 2008 and December 2009, “the capsize sentiment index moves in a more erratic way, reflecting the idiosyncratic features of each sector.” In contrast, “market sentiment is smooth, reflecting market side movements.” Generally, sector sentiment was only significant with the larger factors capesize and panamax.

This reflects the variable demand between capesize vessels, carriers typically above 150,000 dead weight tonnage, which are too large to pass through the Panama Canal, and the smaller more manoeuvrable handy vessels the Handysize and Handyman. Specific categories of ship have different qualities. The supramax is a bulk carrier designed for large quantities of materials which are dense, corrosive or abrasive: when demand or prices of cereals, coal, ore and cement rise, so do requirements for the supramax. Forming around a third of the world’s merchant vessels, it and the capesize are the only two sectors the researchers consider to have significant impact on overall sentiment.

Overall, their model had an 80-85% accuracy of capturing changes in the shipping sector. Partly this was attributed to sentiment contagion between sectors, to the extent that a separate synchronisation index was chalked at 80%, “consistent with strong market integration and herd-like behaviour.” Asked whether shipping was known for the high level of imitative behaviour within the industry, in terms of managerial decisions, Professor Nomikos replied: “There is a high degree of substitute ability – crossover between cargos they can carry and operational ability,” which necessitated inter-firm communication. Larger firms, he noted, were the first to notice a change in market trends as the most affected.

For comparison, they matched their Sentiment Index to two other predictive factors, the ‘Buy and Hold’ model which holds that all short-term market fluctuations of owned assets should be ignored as irrelevent, and the ‘Probability of Expansion’. The latter was estimated using “a specialist econometric model that calculates probabilities (rather than prices).” They calculated changing prices on a forward-looking basis, where the indices are calculated using the available information at each time interval. The Sentiment Index was found to be vastly superior, with a Sharpe ratio higher than the other two for every step.

Lest we put too much weight on this ephemeral concept, sentiment, it is worth clearly defining the term. Cass academics state that investor sentiment is “the propensity to trade on noise rather than info; it may also refer to investor optimism or pessimism.” So presumably if everyone was making properly informed, considered investment decisions, today’s findings would be null and void? Let’s hope for the sake of the few observant individuals that the rest of the sheep in the herd keep following their gut feeling.

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: