Share price for strategy analysis

It is often be useful to study the historical share price of a company. The historical development can tell us what strategic actions and initiative that are liked and disliked by the investors. We should not assume the investors are always right, in fact they are often wrong, but it is nevertheless important to understand the investor perspective. We can also learn about the mindset of the key executives by looking whether their experience is one of growth or decline. 

For an analysis to be meaningful, the share price cannot be analysed directly. In particular it is important to remove the effect of the general stock market. In finance, the resulting time series is called the company's abnormal return. This post will outline how to describe a company's abnormal returns and what kind of question you can ask yourself when analysing the abnormal returns.

Calculation

It is a quite simple process to create an abnormal return chart:

1. Download time series data of daily closing pricing for the stock (e.g Hasbro) and the general market index (e.g. S&P500). 
2. Normalise both time series so that they start at 1 on the first day. This is accomplished by dividing each time series with its value on the first day. 
3. Divide the normalised stock time series with the normalised index time series. The result is the abnormal return time series for the stock. 
4. It is preferable to plot the time series on a log scale.

Figure 1 shows the the share price for Hasbro (orange) and its abnormal return (blue) since 1980. When the blue line is rising, the company is performing better than the general market. When the blue curve is falling, the company is performing worse than the general market. If you use the default arithmetic scale in Excel it looks like the the abnormal returns are very small. However, this is largely an artefact of the y-axis scale. 

Figure  1. Hasbro's share price since 1980 (orange) and abnormal return (blue). The abnormal returns are calculated by dividing the normalised share price with the S&P500 index. The scale is arithmetic, which is a very bad idea. 

Figure 2 shows the same chart using a logarithmic scale. This is a better way to present information because the y-axis is now measuring percent change. Notice in particular how the performance between 1980 and 1986 is totally lost in figure 1. I have used Excel 2013, which can properly handle logarithmic charts. (Microsoft you sat on your ass for 15+ years on that one.)

Figure 2. Same figure as above but with a logarithmic scale. 

It is possible to perform more sophisticated adjustments. It might be relevant to control for the company's beta value. Further adjustments can be made when the company is trading ex-dividend (but then you should also compare with an index with reinvested dividends). 

Instead of comparing with the general market index, it might also make more sense to compare against other companies in the same industry. You can perform a separate analysis for each competitor or you can use (or create) an industry index instead of the using the general market index. Comparing with an industry index might be especially insightful. Sometimes one industry does well depending on the business cycle. This is different than attributing all performance to the company.

Irrespective of the amount of adjustments (and there are many that could be made), the exercise will create a basic understanding of the company's past. In my experience, there is no need to do too many detailed adjustments if you only are interested in the broad-brush history of the company. 

A practical point if you are using CapitalIQ: Note that the automatic function is faulty and you get garbage. Instead download the share price and index manually and do the calculations yourself in Excel.

Analysis

How many years of past history covered depends on the purpose of the analysis. The broad brush perspective in figure 2, shows a few interesting facts:

• The company had spectacular performance between 1980 and 1986. Anyone who got promoted internally to the top management team in 1986 would have been very confident about his own performance. After such stellar performance year after year, there is a clear chance that management will overreach.
• Ever since 1986 the performance has been flat, but with ups and downs. Note that flat performance means that the company performed in line with the general market. That is generally acceptable.
• Between 1986 and 1991, the company clearly had a crisis and made some major mistakes.
• Between 1991 and 1993, the company must have rationalised and improved their operations. Maybe they cut cost and reduced advertising and R&D. That is a typical response to a crisis.
• Between 1993 and 1999 the company was drifting slowly down (blue). This is normally evidence of management not adjusting to the changing environment. Note how the share price moves up during this time period (orange), possibly lulling management into thinking that everything is okay. However the abnormal returns reveal the truth.
• In 1999 some kind of slow-motion crash happened. The crash was not instant and the it took almost two years for the share price to bottom. Maybe the management stayed with the same losing course of action and the board was too week to intervene. Maybe there is another reason.
• Between 2000 and 2014 the company gradually did better and better. 

The above information is a useful starting point and it is possible to go into more detail to understand the nature of the different periods. Are the crises driven by new CEOs? Is the company changing strategy? Are there big acquisitions that failed? It all depends on how much you want to learn about the company's history.

However, generally, it is more useful to look at the more recent history of the company. Figure 3 provides the same information starting in 2000. Here are a few observations:

• Between 2001 and 2010, the company is performing well. An upward sloping abnormal return (blue) is indication that the company is doing better than the the general market. 
• In 2010 and 2011, the company is lagging behind. I would look for changes in management and changes in strategy as possible explanations. The corporate accounts will provide useful information. Did they reduce advertising or R&D around 2007, and experience a delayed effect? Did they do a big acquisition around 2009 that created unexpected problems?
• Starting in 2012, the company is performing in-line with the market. It is impossible to know whether the upward trend since 2001 is broken, but the development during the last three years is a bit worrying. (Or it could just be a repeat of 2004 and 2005).

Figure 3. Share price of Hasbro and abnormal return since 2000.

It is of course possible to zoom in to an even more recent time period. Figure 4, shows the abnormal returns since 2010. The shorter time horizon is more important for investment decisions, but can also be valuable for a strategic analysis. Some observations:

• Bad things happen in December 2010, July 2011, and December 2011. Each time the company's abnormal return drops around 15%. What is going on? Did management respond by short term measures only? Is management's explanation credible?
• There is actually a slightly positive abnormal return since July 2012. Can we trace this to management taking certain actions? Do these actions install confidence about the future?
• There is a strong increase between January and April 2013. What explains this very consistent and strong performance?
• There is a sharp increase in October 2013. Why? If it is due to laying off personnel we can expect a one-off benefit only. If it is due to analysts moving from hold to buy recommendations, then the effect could be more long term. Assuming that the analysts are able to justify their upgrade based on company fundamentals.

Figure 4. Hasbro abnormal return since 2010.

The site Finwiz has good information about recent events of companies. However, Thomson and Standard & Poor have proprietary databases with more information. However, do not expect the proprietary databases to serve the analysis on a plate. You have to download the data and do manual calculations. If you want to look at news pieces you have to do that manually too. The proprietary data bases provide some extra help (e.g. short interest and average analyst recommendation), but even this information must be interpreted by you.