As a caveat to this article, I approached Cisco's (NASDAQ:CSCO) stock from a purely quantitative perspective. I analyzed trends in its financials, built a conservative valuation model, and calculated a fair value of $26.13. The following illustrates the methodology I used in valuing the investment, which I believe has an immediate upside of 17.57% from its current price.
Calculating Free Cash Flow to Equity
As many are aware, the driving force of any investment is the free cash flow it generates. The formula I used for calculating FCFE (Free Cash Flow to Equity) is:
- FCFE = (Net Income + (Depreciation - Capital Expenditures)*(1-DR))
The formula is taken from Aswath Damodaran's definition of FCFE, where DR is the company's Debt Ratio (Total Debt/Total Assets). A simplification I made was to exclude the calculation of changes in net working capital, which I argue would balance out throughout the years.
Calculating an appropriate measure of risk (Cost of Equity)
In calculating Cisco's cost of equity, I will be utilizing the Capital Asset Pricing Model, which was introduced by William Sharpe, John Lintner, and Jan Mossin. The creators were awarded the Nobel Memorial Prize in Economics for this tremendous contribution to the financial world.
The model bases its theory on a well diversified portfolio and stipulates that any particular stock's cost of equity is dependent on its level of exposure to market movements. To anyone who follows the stock market, it's quickly apparent that stocks tend to move together, or correlate each other. This asset pricing model extrapolates that a stock's level of risk depends how heavily it correlates to the market. Essentially, if a stock moves twice the magnitude of the market (in an upwards, or downwards direction), then the investor should be compensated for that increased volatility.
Overall, though, investment risk isn't solely dependent on market risk (otherwise known as systematic risk). There's also tremendous business (idiosyncratic) risk associated to investing in a company. Cisco, for example, is exposed to competition which has been eating into its margins; however, in a well diversified portfolio, an investor would also own its competitors, which have done well during the past year. Overall, the ups and downs of a particular stock would be nullified by ups and downs in other stocks. Theoretically, idiosyncratic risk would be eliminated in a well diversified portfolio, leaving only systematic risk, or market risk, to worry about. That's where the Capital Asset Pricing Model comes into play. The formula is as follows:
- E(R) is the expected return, or the cost of equity.
- Rf is the risk free rate
- β represents Beta
- E(Rm) represents the expected return of the market
The cost of equity essentially represents the risk an investor takes on when investing in an asset, which has to be at a premium to a risk-free investment, like investing in U.S. treasuries. It's arguable that U.S. treasuries aren't risk-free anymore, like they've historically been, but for the sake of this model, we will assume they are still the most risk-free benchmark.
Now, to simply explain what Beta is. In essence, Beta calculates a stock's magnitude of correlation with the stock market. So, if a stock's Beta is 1.0 and the stock market moves up/down 10%, then the stock's expected upward/downward movement is also 10%. If a stock's Beta is 2.0 and the stock market moves up/down 10%, then the stock's expected upward/downward movement will be 20%. Beta helps quantify the level of volatility a stock has with respect to the stock market, which helps quantify the expected return for a stock dependent on its market risk.
Quantifying the Cost of Equity
Assumptions will have to be made when building the model. We'll look at historical and current data to do so. For the sake of conservatism, we'll assume a 30-year holding period.
Taking the 30-year holding period into account, I needed to define our market risk premium, which is a function of the expected market return and current 30-year risk free rate. To emphasize why I chose the 30-year risk free rate, it's because I wanted to match our holding period with my duration. Furthermore, I'll be utilizing the 30-year arithmetic return of the Russell 3000, which has historically been 11.40%. Also, the Russell 3000 resembles the market the closest, assuming its 3000 stock portfolio more definitively correlates the market than the other indices. The current yield on a 30-year treasury sits at 3.87% as of January 9, 2014.
Using the current 30-year risk free rate of 3.87%, as represented by the 30-year U.S. Treasury yield, Beta of 1.25, Market Risk Premium of 6.60%, and expected market return of 11.40%, as represented by the Russell 3000's 30-year arithmetic average, we can derive a cost of equity of 12.12%.
So, how does this cost of equity really play into the whole Cisco valuation thing? Well, the 12.12% is the rate used to discount Cisco's future free cash flows, which is used to calculate Cisco's present value. So, for example, let us assume that Cisco will generate $10 billion in free cash flow in the next year. The present value of that free cash flow will be ($10 billion)/(1.1212), which comes to $8.92 billion. If Cisco generates (for example) $15 billion in 10 years, then the present value becomes ($15 billion)/((1.1212)^10) which comes to $4.78 billion.
The key takeaway here is that the cost of equity is used to discount Cisco's future free cash flows to present values, which tremendously influences the stock's fair value. Essentially, in the long run, an investor who owns shares of Cisco is expected to generate 12.12% returns annually (assuming Cisco's current price is fairly valued). From my analysis, I believe it is undervalued and will generate excess returns.
Building The Valuation Model
Building a valuation model is mechanically simple. Anyone can do it, but not anybody can derive the correct assumptions which drive the ultimate valuation. In building my assumptions, I analyzed trends in Cisco's financial statements to project its financials into the future, while remaining as objective as possible.
Historical Cost and Growth Analysis
This Google Doc illustrates Cisco's historical growth and cost analysis from 2004-2013, with data taken from Morningstar. The blue column illustrates the growth in revenues/costs year-over-year, while the yellow column illustrates the costs as a percentage of total revenues; however, there is a modification. Rather than calculating the "Cost of Sales: Product" and "Cost of Sales: Service" as a percentage of total revenues, I calculated them as a percentage of its corresponding revenues. Thus, for 2004, Cost of Sales: Products was 31.08% of Product Revenues. Also, for 2004, Cost of Sales: Services was 32.99% of Service Revenues. I believe this approach better illustrates Cisco's cost structure throughout time. Furthermore, "Provisions for income taxes" is shown as a percentage of "Income before taxes". A quick observation illustrates that growth for product and service revenues has been declining, while cost of sales have been increasing.
Product Revenue Growth and Cost of Sales
As can be observed throughout the years, product revenue growth has begun slowing at a relatively stable pace from 2010-2013. Furthermore, cost of sales have also consistently increased from 2005-2013. The numbers illustrate Cisco's business environment: one of increased competition, which has been eating at its growth and margins.
The following graph illustrates the product revenue growth for Cisco from 2005-2014*. In calculating the 2014 growth rate, I used Cisco's most recent 10-Q report, which is the first quarter in the company's fiscal year. I made the assumption, correct or not, that the growth in this quarter can be used to annualize Cisco's 2014 growth.
I felt that creating a graph illustrating the growth from 2010-2014 would be more helpful since the decline in 2009 was so atypical. In the following graph you will notice a trendline I created in Excel, with an R-squared of 91.98%. I used that trend line to forecast Cisco's product revenue growth, which we'll get into later.
The following graph illustrates the Product Cost of Sales throughout the same time period:
Service Revenue Growth and Cost of Sales
Although Cisco's service revenue growth has also been declining, I want to note the stability of the service revenues. In 2009, Cisco's product revenue growth declined by 2433 basis points, while Cisco's service revenue growth "only" 951 basis points. This is because Cisco locks in multi-year servicing contracts, which create stability in the company's revenues.
From 2005-2014*, Cisco's service revenue growth was the following:
From 2011-2014*, the service revenue growth illustrates a clearer picture:
From 2004-2014*, Cisco's service cost of sales were:
Overall, we're seeing the same picture between the Product and Services side: declining growth and increasing costs.
Forecasting Revenue Growth and Costs
In forecasting Cisco's product revenue growth from 2015-2018, I analyzed growth trends from 2010-2014* and utilized them to make projections with Excel's trendline function. The projected product revenue growth is as follows:
I utilized the same approach in projecting the product cost of sales:
In forecasting the service revenue growth, I utilized growth trends from 2011-2014, which I believed were more helpful in creating an accurate forecast:
I utilized the same approach in forecasting the service cost of sales:
It may seem inconsistent that for some growth/cost projections I used the 2004-2014* timeframe while, for others, I used the 2010/2011-2014* timeframe. It really depending on which timeframe helped create the most accurate projections, which were done with trendlines.
Projecting Operating Expenses
Overall, I used the same approach in forecasting the operating expenses; however, I had to make adjustments to normalize some of the data. For instance, from 2011-2013, I adjusted for Restructuring, merger and acquisition expenses, which were atypical. Furthermore, I made a slight downward adjustment to normalize for the abnormal 20.5% increase in R&D. The normalized operating expenses from 2004-2014* were:
Given the volatility in operating expenses, I utilized the trends from 2010-2014* to forecast Cisco's operating expenses from 2015-2018:
Overall, I'm projecting a decline in operating expenses as a percentage of revenues.
Projecting the "Net Spread"
Although "Net Spread" isn't a term typically used, I used it to represent the sum of Cisco's Interest Income, Interest Expense, and Other Income. I utilized cost trends from 2008-2014* to forecast the net spread:
Notice how the spread turns negative in 2015, which overall illustrates that interest expense will rise faster than interest income.
I utilized cost trends from 2005-2014* to help forecast Depreciation's percentage of revenues:
Admittedly, this part was tricky to forecast. Capital expenditures are cash outflows which are used to reinvest in the business through investments in Property, Plants, and Equipments (PPE), but also through business acquisitions. Although Cisco's investments in PPE have been historically stable, its investments in acquisitions are more volatile and unpredictable. To forecast future Capital Expenditures (PPE + Acquisitions), I took an approach I felt is very conservative going forward. I looked at Cisco's historical Capital Expenditures over the past ten years. I then broke them up into five different groups (2004 + 2005 were in the first group, then 2006 + 2007, etc.), took an average of the Capital Expenditures as a percentage of revenues and applied those figures as Cisco's forecasted Capital Expenditures. The historical analysis is as follows:
The forecasted Capital Expenditures as a percentage of revenues are as follows:
Again, I would like to reiterate that this approach is very conservative, especially since revenues have grown so much throughout the past ten years. Consequently, this approach will likely overstate the amount of capital outlays done by Cisco.
Forecasted Income Statement
Analyzing the trends in revenues and expenses and using trend lines to forecast through 2018, I forecasted Cisco's revenues and costs (as a percentage of revenues) as follows:
To reiterate, Product Cost of Sales are as a percentage of product revenues and Service Cost of Sales are as a percentage of Service Revenues. Operating Expense, "Net Spread", Depreciation, and Capex + Acquisitions are as a percentage of Total Revenues.
Utilizing this data, I forecasted Cisco's income statement as follows:
Calculating Free Cash Flows
Also utilizing the previous data, I forecasted Cisco's Depreciation and Capital Expenditures, netted them, then adjusted for Cisco's Debt Ratio:
Now we have everything necessary to calculate Cisco's Free Cash Flows to Equity, which is calculated as:
FCFE = (Net Income + (Depreciation - Capital Expenditures)*(1-DR))
The following shows the projected FCFE:
To value Cisco's stock, we'll discount all its future cash flows by the appropriate discount rate. We have already projected its estimated Free Cash Flows through 2018 and have calculated an appropriate discount rate using the CAPM model. At the end of 2018, I calculated Cisco's terminal value as well and will elaborate on the assumptions used.
The following illustrates the discounted free cash flows + terminal value of Cisco's stock:
To calculate Cisco's Terminal Value, I made the following assumptions:
- Long Term Depreciation is 5.49% (the same as Depreciation in 2018)
- Long Term Capital Expenditures is 7.80% (the average of the projected 2014-2018 forecasted Capex)
- The Long Term Discount Rate is 10.47% (applying a Beta of 1.0 as we assume that, in the long term, Cisco's stock will be identically volatile as the market)
- The Long Term growth rate of Cisco's free cash flow will be 2%
Calculating Intrinsic Value
When a shareholder owns a stock, they have a stake in the company's current assets and a stake in the future free cash flows the company generates. Thus, there are two sources of value when investing in a company.
The first source of value comes from the company's balance sheet, which is its Book Value/Share (or Shareholders' Equity). This is calculated by taking all of Cisco's Assets and subtracting its Liabilities. What's left over is equity for the shareholders. Cisco's Book Value/Share (from the 2013 10-K report) is currently $10.97/share. If we look at their 2013 balance sheet, though, they have currently capitalized $21.9 billion in Goodwill, which is an intangible asset. If we deduct Goodwill from its asset base in the calculation, we'll arrive at a Tangible Book Value/Share of $6.90. There is a noticeably large gap between Cisco's Book Value and Tangible Book Value; however, in my calculations I used total Book Value, since I believe Goodwill still represents value for the firm.
The next source of value for shareholders comes from the present value of all future free cash flows. This is where the Income Statement and Statement of Cash Flows came in hand with our forecasting. The present value of Cisco's forecasted Free Cash Flows have been calculated to be $15.15/share.
Now, taking the sum of both sources of value (Book Value and Present Value of FCFE), we arrive at an intrinsic value of $26.13/share.
I am 100% sure that Cisco's actual future Free Cash Flows won't be exactly what I calculated; however, I do feel confident that my approach was conservative and accurate. To expand on my analysis, I want to see the impact on Cisco's shares if different scenarios were to apply. Using data tables, I was able to create over 400,000 scenarios which impacted Cisco's stock valuation. Since stock valuation is most heavily impacted on a company's growth and discount rate, I calculated Cisco's stock price if:
- Total Revenues were to be +- 5% from current expected figures throughout 2014-2018
- The discount were to be +- 3% from the current rate (9.12%-15.12%)
Out of the 401,468 possible valuations (given the bounds in Total Revenue growth and the discount rate), the median expected value was $26.79, representing a 20.57% premium to the current price. Furthermore, given the distribution of valuations and Cisco's current stock price of $22.22, there is a 99.8% chance that the stock is undervalued. The distribution of values is shown in the following distribution table:
Finally, assuming a 5-year holding period and a convergence of Cisco's stock to fair value by the 5th year, the stock should generate average annual returns of 16.39%. Not only will investors be compensated fairly for the level of systematic risk taken on with this investment, they should also realize a 4.27% premium annually in excess of the risk they took on. This 4.27% premium was calculated by taking the expected returns (16.39%) and subtracting the Cost of Equity (12.12%). Then, because the 4.27% premium compounds over 5 years, it generates total excess returns (in excess of being compensated for risk) of 23.27%. Total returns over the next 5 years is expected to be 113.62%.
At this point, some viewers may view that my revenue/cost projections are too conservative. Management provided a 3-5 year revenue growth guidance of 3-6%. I decided to apply this guidance to see the impact on valuation (6% growth in 2014, 5% growth in 2015, 4% growth in 2016, 3% growth in 2017 and 2018). I also assumed the same cost, depreciation, and capex projections. I utilized the same long-term discount rate of 10.47%, but assumed a long-term growth rate of 3%. With management's revenue guidance, I calculated a fair value of $29.96/share.
Ultimately, I took a conservative approach and calculated an expected fair value of $26.13/share, but a median value (through my sensitivity analysis) of $26.79/share. Assuming the median value is more correct, the stock should generate annual returns of 16.39% and generate total excess returns (in excess of the risk taken on) of 23.27% over a 5-year holding period. Given the conservative approach I took, from purely a quantitative perspective, I feel comfortable stating that Cisco shares are undervalued.