Hot Features
Premium
Level up your trading with our powerful tools and real-time insights all in one place.
Post-earnings announcement drift, PEAD, seems to be a feature of share markets, as shown in various rigorous academic studies. (How exploitable is another matter, given its high transaction costs).
There are variations on the theme, but the one I’m most familiar with is where you look at the trend in earnings per share over a few quarters or half years for each of the shares on the stock market. From the trend (a regression) you can calculate an expectation of what the next earnings per share figure for that company will be.
So, at any one time you have one figure: the anticipated eps. Then the actual eps figure is reported by the company.
For some companies the actual will be pretty close to the expected number. For others there will be small surprises – a minor difference between the actual and the expected, negatively or positively. For others there will be very large differences.
The main focus of research is on those companies than produce big surprises in earnings numbers.
When a big positive surprise (actual eps much larger than the previous trend would suggest) is announced the reaction of Mr Market is to raise the price of the share on the day of the announcement.
Conversely, a large negative surprise usually results in a falling share price on the day the new information is released.
But is Mr Market’s reaction to a surprise sufficient?
While the initial movements of collections of large-surprise shares tend to be in the right direction, many researchers have shown evidence that Mr Market is slow to react to the full extent of the news – prices go on rising for months afterwards.
Perhaps there is a tendency to think that a large positive deviation from trend will soon moderate, or even go into reverse and so investors do not raise the share price enough in the first day or so.
Perhaps such conservatism is over-done as investors do not accept sufficiently often that a company has broken out of its past trend and is now on a much faster growth track (or for negative large surprises is now doing much worse).
Gradually however, Mr Market, over a period of a few months catches up and assigns increasing prices to the large positive surprise firm (on average), and a gradually decreasing prices for negative large surprise firms.
Thus there is a potential market-beating strategy of buying large positive surprise firm shares shortly after the surprise in earnings is announced. Also, if possible, sell short the negative surprise (big surprise) firms.
The study I’ll look at today updates this research and combines evidence on the effect of a high/low proportion of institutional volume of trades with the PEAD effect. It is a 2013 paper written by Tao Shu (see recent newsletters for more details of the paper, in particular the institutional volume idea).
Method
For the period 1980-2005 the percentage of trades for a company accounted for by institutions is calculated. Each month shares are placed in one of three categories, Low, Medium or High proportion of trades by institutions rather than private investors.
Tao Shu also constructs an “earnings shock” measure for each month 1980 – 2005. This is the “unexpected earnings (UE)” for the most recent quarterly earnings announcement (only NYSE and AMEX shares in the sample).
This particular study uses analyst’s expectations of earnings (rather than a simple regression of past earnings per share numbers) as its benchmark to calculate the extent of the surprise:
“UE as the difference between reported earnings per share before extraordinary items and the consensus median analyst forecast, divided by the share price 20 trading days prior to the earnings announcement.”
Each month shares are each allocated to one of ten (decile) categories based on the extent of the surprise. Thus, the “Low” decile contains those shares with the most shocking negative surprises, decile 2 contains the next 10% in terms of negative surprise……..decile “High” contains the 10% of shares with the greatest positive surprise.
A one month gap after the announcement is left unobserved. Then the shares are “bought” and assumed to be held for six months. The results shown in the charts are the average monthly returns 1980-2005……………………….To read the rest of this article, and more like it, subscribe to my premium newsletter Deep Value Shares – click here http://newsletters.advfn.com/deepvalueshares/subscribe-1
