Jeremy Johnson's  Instablog

Walgreen Co's (WAG) desire to increase market share and Rite Aid Corp's (RAD) poor capital structure have led some to speculate an acquisition may occur at some point in the future. One of the major considerations in evaluating a takeover in this space is store overlap. The following is an analysis of the two store's footprints across the U.S. with a focus on population per store by state for each company and a hypothetical combined company. This analysis is not meant to be exhaustive or definitive, but rather to discover where potential problem areas may be. In order to determine how many stores is really too many, a market-by-market and store-by-store analysis would need to be done which would include all the local competitors not just Walgreen and Rite Aid.

The following table shows the store footprints, and state population per store for each entity and for the combined companies:

(click to enlarge)

The table has been sorted by combined company population per store with states below the national average highlighted in red. Those highlighted in red are not automatically red flags by any means; rather it is a place where we can start the analysis since those not highlighted are likely of lesser concern.

Within this group, we can see there are three states with no Rite Aid stores which is useful because it gives a benchmark for an acceptable concentration. Illinois and Florida on average have about 21,800 individuals per store. If we look at states where the store concentration is 10% greater than this level, we are left with all the states above Kentucky in the table. A few of these states are likely of little concern because there is not great overlap; for example, Maine, West Virginia, Vermont and Pennsylvania. If we exclude these states we are left with 1,647 potentially problematic (either from a regulatory or business case point of view) stores or about 36% of the total. Of this amount, it is difficult to say how many stores are in actuality problematic - perhaps it is 20% in some states and 40% in others.

Even in California, a state with no great overlap, there are Rite Aid stores within walking distance of Walgreen stores in certain areas from my experience. If we use 35% as a guidepost for Rite Aid's with a Walgreen in close proximity in the most problematic states then there are potentially 575 stores that would need to be considered. There is likely a second set of states where the percentage is in the 10-15% range, or 200-300 stores, and then the states with no overlap where there are no problems. In total this may mean something on the order of 800-900 stores would be of concern or a little under 20% of the total.

It is also useful to look at problematic states specifically. Of the larger states, New York, New Jersey, Michigan and Massachusetts all appear to have significant amount of overlap and relatively high concentrations. These four states and four other smaller states with large overlap have 1,564 stores. New York is especially interesting because of Walgreen's recent purchase of a New York focused chain in that state which shifted it from a likely non-problematic state, to a potentially problematic one.

In order to extend the analysis, the below table shows population per store restricted to individuals 65 and older, on the theory they are heavier users of drugstores:

(click to enlarge)

One fact that can gleaned from this table is that while Florida looks relatively concentrated on a total population basis, its concentration 65 and up is much less. This is somewhat problematic because it challenges the above use of Florida as a benchmark for concentration. Illinois and Wisconsin are already circumspect because they are Walgreen's oldest markets and it is highly likely the company has greater market share there than in other states. Rather than focus on precise figures here, we can use this a qualitative factor auguring for some additional caution compared to the total population analysis since it is difficult to determine the precise relevance of 65+ for drug store revenue.

In order to help readers form their own conclusions, the spreadsheet with the data can be found here. There are a number of methods that can be used to come up with estimates of overlap and some may give more optimistic or pessimistic conclusions. Ultimately, Walgreen is in the best position to judge the long-term financial impact of an acquisition, because it is not just about how many stores would be problematic. For example, there is a possibility that a large set of stores could be sold with little value diminution. There is also the issue of the value in building scale to better negotiate with pharmacy benefit managers who are themselves consolidating.

Disclosure: I am long WAG.

Valuation is a challenging discipline because of the inherent difficulty of integrating large amounts of information in way that is useful for an investor.  For a model to be helpful, it should take relevant financial information and turn it into a set of relationships that are understandable and comparable across companies and industries.  Economic profit models are especially helpful in this regard because they bring to the forefront a company’s ability to generate returns on its investments thus probing deeper into a company than simply analyzing the cash it generates.  Such models can thereby help investors make judgments about a company’s ability to sustain current levels of cash generation and provide insight into how changes within a company and in the external environment can drive changes in corporate value.

Economic Profit

The following equation provides the foundation of an economic profit model:

FV =  (R - d) * NCI / d + NCI

Where:

FV = Firm value

NCI = Net capital invested

R = Rate of return

d = Discount rate (cost of capital)

The model defines firm value in terms of future economic profits discounted to perpetuity plus the value of a firm’s assets.  Economic profits are those above the required rate of return (the discount rate), highlighting one key differentiating feature of the model which is that it explicitly takes into account capital costs which is important in analyzing the value of growth.  There is no explicit growth component in the model; however, by adding some math to above equation, the value of growth can be explicitly integrated by calculating projected growth in net capital invested assuming a given rate of return, discounted back to the present.  One implication of the economic profit model is that growth only has value if the firm is earning in excess of the cost of capital.  Otherwise, there is only enough value in the growth to pay capital providers the minimum required return and no economic profits are generated.  For this reason, often times the value of growth is minimal.

Calculating the return on invested capital is not made easy by the current accounting paradigm which is heavily tilted toward loading the income statement with non-cash charges in an attempt to match the consumption of capital with the generation of profits.  The purpose of this paradigm is to have one number, namely net income that should convey all there is to know about a company’s profitability.  The problem is that calculating returns on investment becomes more difficult with this approach especially when compounded with other biases introduced by current accounting standards such as the accelerated nature of accounting depreciation.

One approach which attempts to solve this problem is to calculate an internal rate of return (IRR) for the entire company using gross assets as the initial investment, gross income as the regular periodic cash flow and asset life as the number of cash flows until asset exhaustion.  Using an IRR based on gross assets as opposed to a simple return based on net assets helps eliminate the problem of accounting distortions which tend to minimize net capital invested and makes for a more accurate view of a company’s ability to make investments at a given rate of return.  Calculating returns based on an IRR model also focuses investor attention on useful economic lives allowing them to make more intuitive adjustments for replacement costs.

Turning back the question of growth, the IRR can embed some growth assumptions into the valuation since the periodic income is assumed to be reinvested at the IRR rate.  The reinvestment of all income at the IRR can be an unrealistic assumption especially for companies with high returns on capital and limited ability to reinvest profits.  For this reason, it is important to adjust the reinvestment rate, producing what it termed a Modified IRR (MIRR).  For companies with returns around the cost of capital, the adjustment leads to little change in value, but scales higher as returns on capital increase.  In general, it is best to make growth assumptions as explicit as possible instead of letting the IRR calculation carry a large share of the value of growth.

Expense capitalization also embeds assumptions about growth.  If investments such as R&D increase, new profit must emerge in order for returns on invested capital to remain the same.  Initially, the impact of additional expenditures is limited but over time becomes greater as more years pass at the higher expenditure level.  If profit does not increase in the wake of greater expenditures, returns on capital will decrease warranting a lower value.

The ultimate implication of the model is that companies generating higher returns on invested capital should trade at higher multiples of invested capital.  In fact, there is a direct one-for-one relationship between the ratio of returns on capital versus cost of capital and the market value of invested capital so that if a company earns twice the cost of capital, its assets should be worth twice the invested value.

Accounting Adjustments

An advantage of any systematic valuation process is the consideration of a defined set of variables across all potential investments.  By relying on a set model to contextualize financial information, it forces an investor to consider the same set of variables across companies and reminds them not to leave any part of the analysis undone.  This is one aspect of making adjustments to financial statements – making sure all relevant information is found and considered no matter the precise technique or model used to integrate that information.

Once information has been identified it must be adjusted to better reflect economic reality rather than accounting rules.  All the adjustments made to financial statements have only a few key goals: First, to identify all capital providers and quantify their exposure; second, to identify all the capital employed in a company; and third to determine the age and remaining useful life of that capital.  Along the way certain other adjustments are necessary for the sake of consistency.  For example, if gross assets are the basis of calculating the return (which is done to better reflect all the capital used in the business), then income must be adjusted to a gross basis to remain comparable.  In this way, adjustments to income are done to match the balance sheet presentation of assets or liabilities and are not done for their own sake.

Typical specific adjustments made to financial statements include capitalizing expenses such as R&D, advertising and rental expenses, inflation adjusting existing assets, adjusting depreciation schedules to better reflect replacement costs and economic value, and the consolidation of equity-accounted investments to better capture the amount of capital employed in these ventures. Regarding the identification of liabilities everything from stock options, to pension plans to other financial obligations are identified in order to determine the full current trading value of the firm.  Although simplistic in theory, there is often a lack of disclosure in company’s financial statements which can make for a significant amount of guess work and financial reconstruction.

Interpreting the Results

In practice, valuation can only give investors a guide post as to the future level of a company’s total firm value.  An undervalued company can quickly become an overvalued company if financial results deteriorate rapidly.  In 2007, building products companies capitalized with debt at three times gross cash flows were threatened with bankruptcy two years later due to a collapse in demand for their products.  In the public markets many historical examples exist of companies that are seemingly cheap -- even taking account of all the economic profit adjustments -- that turn out to be fairly valued or overvalued as financial results underperform historical trends. 

For this reason, a valuation cannot replace a strategic assessment of a company’s position; however, a valuation can be a useful tool to help make a strategic assessment of a company.  The most important aspect of this assessment is looking at returns on invested capital in the context of a company’s life cycle.  Companies with high returns on capital are more apt to experience competitive threats since other companies will be incentivized to enter their businesses.  Companies with returns on capital around the cost of capital are more likely to sustain their performance absent large changes in the macro environment.  Finally, companies earning below the cost of capital should be expected over time to improve their performance if the industry they operate in is not threatened with extinction and after a potentially painful restructuring. 

Investors can also use a model to determine where a disconnect in value lies.  If a company is undervalued based on run-rate profits, then there is little need to discuss what growth prospects are embedded in the market value because the answer is none.  Instead of focusing on whether the company can grow, an investor can turn their attention to whether the company can maintain its position.  The takeaway is that each company will have an individual strategic situation that must be analyzed, but an economic profit model can help investors with that task.

The Economy

Even a perfect valuation cannot protect investors from a recession which will impact the valuations of the vast majority of companies regardless of whether a company was overvalued or undervalued before a recession or whether a company’s profit is marginally or seriously impacted by the effects of a recession.  A valuation can alert investors to companies whose valuations make them more susceptible to larger falls in value.  Economic profit models cannot ring fence investors from the impact of a recession, and those owning equities should expect that recessions will lower valuations on all companies regardless of valuation. 

Further Reading

For those wanting to know more about economic profit models, I suggest taking a look at the following books:

Costantini, P. (2006). Cash Return on Capital Invested. Burlington: Elsevier.

Fabozzi, F. J., & Grant, J. L. (2000). Value-Based Metrics: Foundations and Practice. New Hope: Frank J. Fabozzi Associates.

Madden, B. (1999). CFROI Valuation: A Total System Approach to Valuing the Firm. Woburn: Butterworth-Heinemann.

Viebig, J., Poddig, T., & Varmaz, A. (2008). Equity Valuation: Model from Leading Investment Banks. Chicester: John Wiley & Sons.

Young, S. D., & O'Byrne, S. F. (2001). EVA and Value Based Management: A Practical Guide to Implementation. New York: McGraw Hill.

I became interested in gold around the year 2002 because of the long stagnation in price and the aggressive monetary policy being implemented at the time.  I was a teaching assistant in an accounting class at university and we had a stock market contest.  I recommended that one of my student's select a gold stock and since the teaching assistant's were allowed to participate as well, I selected one.  Both our stocks were up 40% in 3 months.  Having caught the gold bug, I began doing further research on the metal.

After graduating I read an investment piece by Barton Biggs that highlighted the case for gold.  The work wasn't his own, he was in fact passing along some of the analysis of a former colleage whom gained some notoriety by convincing the firm that put option prices were extremely cheap in late 1987.  The volatility smile was born shortly thereafter and this individual secured himself a nice opportunity to set out on his own for his contribution to finance.  Unfortunately, I can't remember this individual's name, but he founded a small firm that dedicated itself to gold market research.  One of the key insights this individual had was that gold performed well in times of negative real interest rates.  Seems obvious in 2010, but this relatively simple analysis with some other supporting facts turned me into a gold bull in 2003.  As a recent college grad with plenty of debt, I had limited means to express a view on gold, but in 2004 I purchased about 2.5 ounces of gold at the equivalent of $390 an ounce.  My opinion on why gold has performed well over the past 10 years is far more nuanced than it was at the time which is to say I don't think I saw all the risks but this is the genesis of my interest in gold and should establish some background for my current view.

About two years ago, I turned my attention back to the gold market because based on intuition it seemed it was becoming overpriced.  Gold wasn't popular in 2002, but by 2008 there were regular ads on TV for gold and it was a common conversation piece.  My car mechanic recommended buying gold and storing it in Australia.  Having entered the financial markets during the tech wreck, it's impossible for me to not question anything where the positive zeitgeist has become so palpable.

You often hear people say the dollar as a cash instrument is worthless -- not literally worthless, but that it provides a very poor, even negative return.  This is one of the core arguments today for gold, essentially that it is the anti-dollar.  In keeping with this thinking, it occurred to me compare gold to cash over a long period of time.  I also added a couple other comparisons which cash investors often look at.  To set the stage, here is the chart:



Let me explain the above graph.  It is a comparison between five time-series all re-indexed to a 100 start point.  I have estimated year-end 2010 levels from those available mid-year 2010 in the case of CPI and personal income per capita, the current price of gold and an estimate of T-Bill interest through year-end which is effectively zero.  

First, the cumulative return on 3-month T-Bills is charted in blue with an ending index value of $722.51.  Second, the CPI is charted in red over the same period with an ending index value of 494.84.  The CPI is an oft-maligned measure but my own view is that it measures what it was designed to measure quite well.  What it does not measure is the additional costs added each year to a person's expenditures brought about by quality improvements and new products and services added to the economy.  Without getting into the debate further, let me propose an alternative measure by suggesting that gold could be a hedge not just against the price level of goods available today, but also of goods available tomorrow.  To get an approximate grasp of this I have taken personal income per capita which has risen to an index level of 768.44 and charted it in green.  It is interesting that this level roughly approximates that of the returns on short-term interest.  

Charting the price of gold for this exercise isn't as simple as it seems.  The single biggest issue to address is the start date.  If the starting price is low, it will look like gold is currently expensive.  If the starting price is high the opposite will occur.  Therefore, changing the start year can have a very large impact on the relative returns of gold.  The price of gold was set at $35 until 1968, and under a two-tier system until 1971 which means the gold price isn't meaningful until after these dates.  By 1973 gold was $106 and by 1980 it was $595.

To determine the start year I have looked at differences in the return on cash, the price level and gold from 1913 to the early 1970s.  Here are some facts from that analysis.  CPI increased 224% between 1933, the year gold was set at $35, and 1973.  Gold increased 204% over the same period.  If you go back four years earlier to 1929 before the deflation of the Great Depression when gold was $20.67 and the price level had been stable for almost 10 years, the gold return would have been 415% (1929 - 1973) while the increase in CPI would have been only 159%.  Just looking at this data, it seems gold in 1973 was fairly priced if not overvalued based on the relative price levels prevailing at the time.  I have also included a time-series which assumes the average price of 1973-1975 as the starting point, which at $143.  This is a more generous starting point for gold.  Let's be clear here before moving on.  If an investor would have bought gold in 1929 and been able to keep it until 1973, they would have outperformed inflation by 256% based on a $106 gold price and 432% based on a $143 gold price.

Having set the stage, let me address the implications of the graph.  By any measure on the above graph, even the baseline CPI figure, gold was undervalued in 2001.  This I think goes a long way toward explaining its performance over the past 10 years.  In 2010, it seems the story has reversed.  Whereas holding T-Bills was the better course of action on average from 1973 to 2006, now gold firmly has the upper hand.  Furthermore, gold has vastly outperformed since 2001. 

We should be cautious in drawing too many conclusions from such a simple analysis.  For example, the cost inputs in the production of gold are not aligned well with the CPI.  Gold mining is energy intensive and it could be that gold is following the price of oil more than the general price level.  While the economy has become less energy intensive, the gold price may not have.  In fact, it may have become more energy intensive.  Gold may have shifted from being a hedge against generalized inflation, to that of energy inflation.  This analysis also does not address the future: gold can be overvalued today, but based on future inflation and wage growth it may not.

To conclude, I think looking at an analysis such as this is just one part of the picture.  The best lesson I can draw from it for my own trading is that today it is likely that gold is more of a symmetrical bet on the future course of inflation which wages and interest will follow, rather than a value investment in an unloved fundamentally undervalued asset.  As an investment, gold has changed -- it is not the same investment it was 10 years ago.  That does not in itself make it a bad investment.


Disclosure: Long a de minimus amount of physical gold.