Retail Sales and Financial Markets

According to data from the National Retail Federation, the US Black Friday weekend was a record breaker. A record 247 million shoppers visited retail stores over the Black Friday weekend, a 9.3% increase over last year’s 226 million. The average holiday shopper spent $423 which was higher than the $398 spent last year. Total spending is estimated at a whopping $59.1 billion. 

Very good numbers to be sure, and possibly strong enough overall sales to help some struggling retailers like Best Buy (BBY). Retail sales are not, however, very useful for predicting the future. The important point to remember is that retail sales are not a leading economic indicator but a coincident indicator. Retail sales tend to peak and trough in line with other economic indicators like GDP or industrial production. Coincident indicators are useful for telling us something about where the economy is, not where it is going. The chart below shows how retail sales have closely matched the S&P 500.There are no obvious examples of retail sales leading the S&P 500 and, therefore, retail sales may not be very useful in helping to forecast financial markets.

Business Investment as a Leading Indicator of Employment

For a slightly more optimistic perspective on the US economy and financial markets, here is a chart showing how US business investment and US employment have performed over the past ten years.

Notice how business investment tends to peak and trough before employment. This chart indicates that business investment may be a useful leading indicator of employment. For a closer analysis, I downloaded the data and constructed a relatively simple autoregressive distributed lag (ARDL) model with 2 lags on each of investment and employment. For modelling purposes I took the natural logarithm of each variable. The resulting regression output looks like this:

Tests on the residuals and squared residuals indicate that serial correlation is not a problem. Since the variables are measured in natural logarithms, the coefficient estimates can be interpreted as elasticities. The short-run elasticity of employment with respect to business investment is 0.053% while the long-run elasticity is 0.160%. In the short-run, a 1% increase in business investment increases employment by 0.053%, while in the long-run, a 1% increase in business investment increases employment by 0.160%.


Here is a plot of the fitted values and residuals. Notice how closely the fitted values track the actual values.

Since business investment is currently trending upwards, this indicates higher employment (and hopefully lower unemployment) which is good for the economy and financial markets.

HP Needs a New Way

I have a particular fondness for HP because I remember reading "The HP Way" when it first came out in 1995. The book is written by David Packard and tells the story of HP from the beginning. In the book Packard explains that he is not a big fan of the MBA degree but is more interested in the MBWA (Management By Walking Around). HP's success was built on excellent products, technological innovation, knowing the customers, and adapting to change.After reading the book, I though, "wow! what an amazing company". HP was admired by many people and many business cases were written about it. HP is after all, credited as the founder of Silicon Valley.
 
In the 2000s HP pursued an aggressive expansion strategy that involved buying big companies (the merger with Compaq in 2002, the acquisition of EDS in 2008, the acquisition of 3Com in 2009 and the buyout of Palm in 2010). With this many big deals in such a relatively short period of time, critics wondered where the synergies were coming from. Along the way, HP had a number of product issues and accounting issues.
 
Now the company is in a death spiral. The stock is now trading at a 10 year low after recently announcing a $8.8 billion write-down and poor quarterly results (see here). HP is reeling from its $11 billion purchase of Autonomy. This was a move designed to help HP move farther into software and services (something that IBM has successfully done).

Over the past 2 years HP is down almost 70%. DELL has also had a difficult past 2 years, while  MSFT is holding steady.

HP can be compared with its competitors to see how efficient it is. To calculate technical efficiency I use data envelope analysis (DEA).  DEA is a non-parametric approach to the estimation of production functions. I use three inputs (employees, total assets, total operating costs) and one output (total revenues). Data are averages over the years 2004 - 2011. For those interested in the technical details, I use the 2 stage input approach with variable returns to scale (VRS).

The DEA results are presented in the above table. Total technical efficiency (CRS_TE) can be broken down into pure technical efficiency (VRS_TE) and a scale effect. The total technical efficiency measures indicate that Apple, Google, Dell, Microsoft and Accenture are efficient since their CRS_TE measures are equal to one. Cisco and HP are inefficient. HP. for example, can reduce its inputs by 1% and still produce the same output. In the case of HP, the pure technical efficiency measure of 1 indicates that allocation of inputs to output is efficient and the inefficiencies are coming from the scale effect which measures the size of the company. HP exhibits decreasing returns to scale (RTS) which means that given its production possibility frontier it is producing too much output. Value creation under decreasing returns to scale is difficult because decreasing returns to scale means that HP needs to get smaller.

An Update on Canada's Broadband Performance

Here is an update to my previous post on broadband performance. Average broadband download speed in Canada has improved 50% from last year. This data is current through October 2012.

How Efficient is Reseach in Motion?

Watching Research in Motion's (RIMM) stock price fall over the past few years has been difficult. While RIMM gets singled out for its poor stock price performance and Apple becomes the most valuable company in the world by stock market capitalization, it is important to point out that some of RIMM's competitors have not been doing so well. Over the past year, Apple has been the clear winner, but Ericson, Nokia and Research in Motion have each underperformed.

One thing to remember about companies based on cutting edge technologies is that no company can be the leader forever. The product landscape changes. What is new today becomes common place in the future. Think about the cutting edge technology companies of 15 years ago (Microsoft, Cisco, Dell). These companies are still in business but now they are no longer considered young dynamic upstarts but rather established mature companies. In my view, handset makers are entering a more mature phase in which the explosive growth in hand sets is going  to slow considerably. A new United Nations report reveals that there are now 6 billion cellphone users in the world. That means that 86 out of every 100 people now have a cellphone.

A different way to compare RIMM with its competitors is to calculate how efficient it is. To calculate technical efficiency I use data envelope analysis (DEA).  DEA is a non-parametric approach to the estimation of production functions. I use three inputs (employees, total assets, total operating costs) and one output (total revenues). Data are for the year 2011. Samsung is omitted from the comparison because it is a huge multi-product conglomerate. For those interested in the technical details, I use the 2 stage input approach with variable returns to scale (VRS).

The DEA results are presented in the above table. Total technical efficiency (CRS_TE) can be broken down into pure technical efficiency (VRS_TE) and a  scale effect. The total technical efficiency measures indicate that Apple, Google and Research in Motion are efficient since their CRS_TE measures are equal to one. Ericsson, Motorola are Nokia are inefficient. Nokia. for example, can reduce its inputs by 29% and still produce the same output. In the case of Motorola, the pure technical efficiency measure of 1 indicates that management is efficient and the inefficiencies are coming from the scale effect which measures the size of the company. So, while Research in Motion's stock price has suffered over the past year, it is, at least by these calculations, an efficient company.

Prediciting US Presidential Elections

President Obama was re-elected as President of the United States and for some market watchers, this was  not a surprise given that the stock market rose  in the two months prior to the election. According to InvestTech Research the stock market is the most reliable indicator of who will win the presidency. InvestTech Research tested the hypothesis over the past 100 years that if the stock market gains in the two months leading up to the presidential election, the incumbent party wins. If the market falls, the incumbent party loses.

"In the 16 elections when the stock market climbed before Election Day, the incumbent party was re-elected 15 of 16 times. And, in the 12 election years when the stock market suffered losses, the incumbent party lost 10 of 12 elections."

The stock market accurately predicted the next president 25 out of 28 times. This corresponds to a 89.2% probability of success. The three years that the stock market failed to accurately predict the outcome of the election were 1956, 1968, and 2004.

With last nights election over, the numbers can be updated. To date the stock market has accurately predicted the next president 26 out of 29 times. 

Hurricane Sandy and Cat Bonds

In BSUS 6600 I spend some time talking about the role of catastrophic bonds in managing severe weather disturbances. Cat bonds are potentially an important way to diversify weather related risk. Unlike weather derivatives, that are traded on exchanges, cat bonds are non-traded insurance linked securities. Cat bonds are structured in such a way that insurance payouts are rare. Here is an update of what effect Hurricane Sandy may have on cat bonds.

For those interested in following the cat bond market, Artemis is a useful blog.