Every morning I receive an email from a South African financial institution that summarises the previous day’s news on the financial markets. I do not follow the financial markets closely, so I appreciate a quick summary of what had happened – and what the major news items for the day will be. Sometimes, when there is time, I would browse through the section on research investment ideas, where an analyst would discuss the fundamentals of a stock and its expected performance. But I’m often less interested in the stock itself – the consequence of a limited personal portfolio – and more interested in how accurate these analysts are in their advice. Are their forecasts actually correct more than 50% of the time?
This is even more true of things called ‘trade ideas’. Magazines like Finweek would often publish expert opinions on short-term stock price movements. This could be anything from a change in management that is likely to improve the value of a stock and suggests it is undervalued, to a looming trade war between China and the US that may affect the value of a specific stock negatively. Again, I would read these opinions and wonder whether they really do change investor’s decisions to buy or sell, and whether they actually do earn positive returns for those willing to follow their advice.
My gut feel is typically that this information is already priced into the market. And, of course, if you believe the efficient market hypothesis, there is no reason to expect that these reams of email alerts and magazine space matter at all. The efficient market hypothesis states that stocks always trade at their fair value, and that analysts and the information they provide cannot produce risk-adjusted excess returns (alpha) consistently. Yet analysists keep writing research reports and magazines keep publishing them. There clearly seems to be a demand for these services, suggesting that they must add some value. (Although, to be fair, this is not an entirely valid argument; horoscopes attract wide readership too.)
So do these financial analysts add value – or are they the equivalent of horoscopes for investors? That is the question four economists have now answered in a new NBER Working Paper. The authors chose to look at the effect of ‘trade ideas’ in particular. These are somewhat different from the normal stock recommendations which are based on fundamental research. Trade ideas have short time horizons – typically between a week and three months – whereas fundamental research recommendations have much longer time horizons – maybe a year or longer. That is because trade ideas are often issued in response to upcoming news or, less often, because of an over- or underreaction to past news that is expected to be corrected in the short run. It may thus be that an analyst makes a buy recommendation on a stock based on their fundamental research, but simultaneously recommend the short-term sale of the same stock – through a trade idea – based on a news item.
To identify whether these trade ideas yield positive returns, the authors construct a dataset of 4543 trade ideas from 688 analysts at 77 brokerage houses between 2000 and 2015. They then compare this to stock market data. Using this dataset, the first obvious question is whether these trade ideas result in additional price reactions at the time of the announcement. They do. Trading buys and sells have average benchmark-adjusted returns of 0.91% and -1.96% over the day of the announcement and the following day. What is more, the stock price impact continues to increase for three months after the trade idea is issued and exhibits no reversal, consistent with the fact that the information conveyed through trade ideas are permanent and stock prices do not fully incorporate all relevant trading call information at the time of the announcement.
And the magnitudes are not insignificant. The results show that ‘both trading buys and sells generate significant benchmark and risk-adjusted returns. In economic terms, the buy portfolio generates a daily characteristic-adjusted (7-factor risk-adjusted) alpha of 4.5 (3.9) basis points, which corresponds to about 90 (78) basis points on a monthly basis. Magnitudes are about twice as large for sells.’
The paper explores many other questions. Do trade ideas based on news items solicit bigger returns than trade ideas that are just based on mispricing? Yes. Do trade ideas exhibit stronger price reactions for calls that are in the opposite direction of outstanding stock recommendations? Yes. Do trade ideas from larger brokers and All-star analysts has a greater price impact? Yes.
One interesting question pertains to the investors. Are institutional investors more likely to benefit from these trade ideas? It seems so. Using a different dataset of institutional investments, the authors find that ‘consistent with institutional investors perceiving trade ideas to add value, the increased trading is in the direction of the trade idea, and we do not find evidence of increased institutional trading in the direction contradicting the call’. They also find no evidence of increased trading preceding the call. However, ‘when we focus only on trading activity by institutional clients of the broker generating the trading call, we find that commission-paying institutional clients exhibit statistically significant increases in trading activity as early as three days ahead of the announcement of trading call’. Being a client of a brokerage firm with an All-star analyst does have its benefits, it seems.
Finally, do the analysts with good trade ideas also generate better stock recommendations? They do. Stock upgrades are 0.63% higher for analysts producing trading research. This is about the same upgrade of an All-star analyst of 0.64%.
If this is true, why do not more analysts issue trade ideas? Trade ideas are obviously riskier than calls based on fundamental analysis, as it is much easier to assess whether the idea is correct or not, given the short time horizon. That is why it is generally analysts with more established reputations. But, as the authors show, issuing trade ideas call be a good career move: ‘We find that trading calls pay off for analysts who make them. Analysts who produce such calls are more likely to be subsequently included in the All-American Research Team roster.’
If I want to identify the next Warren Buffet, maybe I should pay more attention to those early morning emails in my inbox.