If you're interested in the ongoing product recall issue, Jennifer Toney at Haas, who knows this material as well as anyone, recommends the following readings:
Widmer, Lori. “When Your Name Is at Risk" BNET Research Center. November 2000.
N. Craig Smith, Robert J. Thomas, and John A. Quelch. “A Strategic Approach to Managing Product Recalls” Harvard Business Review, 1996.
Chu, Teng-Heng, Che-Chun Lin, and Larry J. Prather. “An Extension of Security Price Reactions Around Product Recall Announcements.”Quarterly Journal of Business and Economics, Summer-Autumn, 2005.
There is a reasonably thorough literature around stock price reactions to product recalls, but it seems to me to mostly miss the point. The real damage of a recall is likely to be reputational, so any impact on the stock price is going to occur over time. I think it would be better to try and focus on the long-term valuation effects.
Johnson & Johnson's response to the Tylenol incident (good briefing here) is regarded as the classic blueprint for how to handle a safety problem. But not all companies do so successfully. Perrier's leadership in the premium bottled water segment once seemed unassailable, until small amounts of benzene were found in the product, prompting the recall of 160 mm bottles. But in both cases the ultimate impact on firm value wasn't apparent until long after the fact.
I don't know if recalls can be studied quantitatively - the big ones are rare enough that it might be better to use the case study method. One thing that seems apparent is that each situation has its own logic. At a minimum the analyst needs to consider:
The direct economic impact of the recall (usually small)
How well the recall was handled
The completeness of the recall (is there still potentially dangerous product out there)
The company's brand and reputation before the recall
The company's reputational exposure (consumer-facing companies might have more sales risk)
The initial impact of the recall on the brand
The potential for follow-on news (reports that the company hid problems, etc.)
The likely efficacy of the company's damage control measures
This is a probably a good example of something that can't be measured, but matters. I doubt we'll ever have an 'R' score that quantifies the impact of lost reputation of firm value. But analysts ignore reputational effects at their peril.
Last Sunday we had the pleasure of awarding the Moskowitz Prize to Alex Edmans of the Wharton School for his outstanding study of the 100 Best Companies to Work For. Alex's study covers the entire period that these ratings have been published in Fortune. His presentations at SRI in the Rockies were superb (see links below). If you're going to Wharton, get signed up for this guy's classes - he is a great lecturer, and he is rock solid on financial theory and his knowledge of the recent literature.
An audio recording of Alex's talk at SRI in the Rockies is here, and his slides are here
The winners of the Moskowitz Prize are taking on a different character, and I wanted to take note of it here. In the 'old days' (pre-2004), studies tended to focus on the broad concepts - social responsibility, sustainability, etc. In retrospect, Marc Orlitzky's study was the culmination of this line of thought. He demonstrated that the concept of social responsibility was not just conceptually valid, but could also be framed as a valid statistical construct. He then argued that social responsibility had been positively associated with financial outcomes (although the effects he found were much stronger for accounting-based than for market-based measures).
If Orlitzky was right that social policies have been financially beneficial (and there is still plenty of debate about that), the question becomes one of how the mechanism operates. It's probably true that some social policies are good for financial results while others are bad. But which policies contribute positively and which contribute negatively?
The three most recent winners have zoomed in on a single issue and tried to answer these questions. Nadja Guenster looked at the impact of Innovest environmental ratings on fundamentals and returns. Brad Barber examined the impact of the CalPERS corporate governance program on stock valuations. And Alex's study looks at how employee relations policies impact portfolio performance.
In each case, the analyst focused on a measurable and important subcategory, and demonstrated that there was a positive historical association with returns. Each study focused on a social variable that was well-specified, and used state-of-the-art risk models to assess performance.
Before we get too excited about these performance studies, however, it's important to remember last year's Honorable Mention paper by Harrison Hong and Marcin Kacperczyk, which showed that sin stocks have had exceptional returns over the years. Like Gunster, Barber, and Edmans, this study zooms in on an important social variable and looks at returns through the prism of a modern risk model.
This trend strikes me as a very healthy development for social investment research. Academics are moving away from general conceptions of social responsibility and doing detailed analysis of the individual stakeholder categories. The results have generally been happy, so far, but, as the case of sin stocks show, social investors should be ready for unpleasant surprises as well.
We know that, in aggregate, social screens haven't added value over the past 20 years. Now we know some have been positive and some have been negative. As we go further down this path, social investors will increasingly be challenged with hard data to re-consider some of their portfolio construction decisions. That will be healthy, but it will not be comfortable.
Just a couple more quick links on the toy safety situation:
According to some press reports (e.g., this one) Mattel has apologized...to China. This Washington Post article, however, says there are two sides to the story. (Also note the analysis in the article by Eric Johnson at Tuck, which sounds on-the-mark to me.)
According to the article, China makes about 80% of toys sold globally.
The toy business is highly seasonal - the September and December quarters accounted for 69% of Mattel's 2006 sales - so we won't have a strong sense of how much fundamental impact this is having until the current quarter is reported.
Despite this latest announcement, MAT stock has performed about in-line with the market over the past five trading days.
Both toy manufacturers that announced recalls this summer - MAT and RCRC - have been down significantly, both in absolute terms and vs. the market (green line is S&P 500, red line RCRC, blue line MAT):
If you are a business student looking for an interesting paper topic, this situation has plenty of material to work with.