Stock Earnings and Preannouncements

"We find that firms are more likey to prenounce earning if the consensus of analysts' forecast is very difeerent from actual earnings, if the dispersion in theese forecasts is high, or the firm has negative news. Further, we document that there is a systematic pattern in the amount managers preanounce. On average, firms release only some of their positive news at the preanouncement date, thereby creating a predictable surprise at the earnings anouncement date. Given that preanouncements occur very close to the actual earnings anouncement date, it is likely that managers know the actual earnings amount with a high degree of certainity. Therefore, our results suggest that managers strategically withhold some of the positive earnings surprise. We show that earnings surprise is more persisitent for firms that preanounce positive news for firms that do not. "
Soffer, Leonard C., Thiagarajan, S. Ramu and Walther, Beverly R., "Earnings Preannouncements" (September 1997)

"We find a strong – economically and statistically – negative price reaction to aggregate earnings news. This finding suggests that unexpectedly high earnings are associated with higher discount rates, at least over the fairly short horizons we study. Aggregate earnings are strongly correlated with macroeconomic conditions, including measures of real activity and proxies for discount rates (Tbill rates, the term spread, and the default premium). However, the discount-rate proxies only partially explain the market’s negative reaction to earnings news. Thus, the results suggest that discount-rate shocks not captured by our proxies explain a significant fraction of returns."
Lewellen, Jonathan W., Kothari, S.P. and Warner, Jerold B., "Stock Returns, Aggregate Earnings Surprises, and Behavioral Finance" (February 2003)

"Further, upgraded stocks bought by herds initially outperform, then underperform their size, book-to-market, and momentum benchmarks, while downgraded stocks that are heavily sold exhibit the opposite pattern."
Brown, Nerissa C., Wei, Kelsey D. and Wermers, Russ R., "Analyst Recommendations, Mutual Fund Herding, and Overreaction in Stock Prices" (February 13, 2008).

“The one thing that everyone, from the most uninformed layperson to the most erudite professor knows about the stock market is that earnings determine returns. Chapter 1 of every finance text book, lecture 1 of every investment class and every news dispatch all agree:
1) Earnings and markets move up and down together
2) The greater the earnings increases, the higher the market return.
3) When earnings are up, it is time to buy stocks, and when earnings are down, it is time to sell stocks.
4) When the market’s price/earnings ratio is high, it is time to sell.
Analysis of each of these hypothesis’ would invalidate each and every one of them as false. Contradictorily, each is completely opposite to the actual empirical relations. Almost everything investors are taught about in relation to the earnings and stock markets returns, whether in business schools or on the stock market pages of the newspaper is wrong.Shrewd articles know that they will not get far in life by relying on the conventional explanation of market moves. But the great majority of investors are deceived on both the big and the small aspects of this relation.Firstly, it is necessary to examine the actual relations between earnings and the market returns. There are many different leads, lags and ratios to be tested. The best way to reveal these relationships is with a scatter diagram. If these data tend to life close to the line of best fit, then there is a strong relation. If the data tends to appear to lie In a shapeless mass around the line, then there is little or no relation. Each of the charts uses earnings data from S&P securities price index record, updated with current figures and returns data from London Business School’s comprehensive database of global markets.According to the induction, stocks are supposed to go down when earnings are down and go up when earnings are up. In fact, they don’t and the opposite relation holds good.1) If reported S&P 500 earnings rise in a year, the S&P 500 is likely to perform worse than average that year.2) If the reported earnings fall in a year, the S&P 500 is likely to perform better than average that year.The inverse relationship is shown in the above figure. Notice how the line of best fit slopes gently from the northwest quadrant to the south east. This is the hall mark of what statisticians call negative correlation. Regression analysis is used to calculate an equation that expresses this slightly inverse relation between earnings and next year’s market return as a line of best fit: S&P 500 return = 9.6 percent minus one fifth of the annual percentage change in the S&P earnings. The equation explains some 5% of the total variation in stock prices from normal. Another way of saying this is that if a person new nothing about earnings change in a year, the best forecast of S&P 500 returns in 2002 would be 9.6%. Knowledge of the earnings would increase the predictive ability of the analyst by only 5%”
Niderhoffer in Practical Speculation



Case Study: Bear Stearns
"Bear Stearns's balance sheet, liquidity and capital remain strong,'' Chief Executive Officer Alan Schwartz said in the company's statement. Alan ``Ace'' Greenberg, the former Bear Stearns chief executive officer and current board member, told CNBC that the liquidity rumors were "totally ridiculous.''
”Credit-default swaps protecting against a default by Bear Stearns for the next two years soared to 900 basis points, according to broker Phoenix Partners Group in New York. That's up from 514 basis points last week, CMA Datavision prices show.”
News headlines on 12th March.

Premarket BSC starts trading up 10% after the financing news came out on 14th March. See Chart:


What happened thereafter is well known.