I've re-read the book cover to cover, and the most important takeaway I seem to be getting is that the super-investors profiled in the book (beyond the obvious Graham and Buffett) actually practice looking to actively buy $1's for $.50. This is the goal of the concept of margin of safety. It is the act of striving to buy $1's for $.50 over and over again, while at the same time avoiding cigar butts. For the average investor (myself included), it is probably one of the most difficult concepts to master, but the more I read (and the more I invest), the more I think we investors make things more difficult than they need to be. During any 52 week period, there will be no shortage of cheap stocks which can be down anywhere from 20% to 80%. The key question is, why are these companies so cheap. Usually, a stock down 80% is down 80% for good reason. But sometimes, there is no good reason, and therein lies an opportunity.
Sometimes investor sentiment towards a certain type of business causes an entire sector to get cheap for no good reason other than selling perpetuating more selling. Case in point, currently (i.e., as of August 2015), legacy media-content or broadcasting companies (think Viacom, Sripps, Discovery, CBS, Fox) have cratered in the wake of a prevailing paradigm that new media-content companies like Netflix are the indisputable and irrefutable wave of the future. And along the way, a funny thing happens. Short-term, performance oriented group-think bins legacy media in favour of Netflix irregardless of valuation, despite the fact that Netflix is essentially insolvent. If you don't believe me, have a look at Netflix's most recent quarterly balance sheet - the ratio of content liabilities to liquid assets is something like 10:1, but Wall Street ignores the leverage in favour of metrics like subscriber growth. Which brings me to a discussion of Greenwald.
Greenwald describes three different methods of valuing companies:
1) Balance Sheet Valuation
2) Earnings Power Value
3) Value of Growth
I'm going to discuss Balance Sheet Valuation and Earnings Power Value, leaving the Value of Growth for a later date. I find Greenwald's model appealing because, being contradictory by nature, I fully believe that the Value of Growth is actually a joke. Wall Street evaluates investment candidacy principally on the presumption (and extrapolation) of growth, therefore, if Wall Street believes it, I must shun it. Greenwald emphasizes that growth only has value if it is growth within a franchise, and seeing how rare sustainable franchises actually are, growth for the most part has no value (in general). Growth seems to actually destroy value, because if it requires additional ongoing investment to support that growth, and if the ROI is less than the cost of growth, the value of growth is destructive. Sounds counter-intuitive, I know, but Greenwald explains it pretty thoroughly.
Balance Sheet Valuation
According to Greenwald, Balance Sheet Valuation provides the most comfort in terms of valuing a company, as any analysis performed is based on the company's most recent Balance Sheet only.
Greenwald demonstrates how to work through a sample Balance Sheet line by line, by ascribing an estimated valuation bump to each Balance Sheet item under two hypothetical states: 1) Liquidation and 2) Going Concern.
Under either state, an investor can potentially uncover value in overlooked or misunderstood situations (think spin-offs, restructurings, conglomerates &/or companies with holdings in other companies) however, each state requires a modicum of expertise in uncovering value. Liquidation analysis requires expertise in understanding valuation of companies going through bankruptcy/restructuring, with eventual investment geared towards buying distressed debt.
Going Concern analysis requires utilization of a sum of the parts approach to determine whether the market has overlooked the value of certain items on a company's Balance Sheet carried at historical cost (i.e. equity method investments, land carried at historical cost, etc.), or for that matter,
underpriced the entire value of the company at less than the fair value of hidden assets on the Balance Sheet.
Greenwald's Balance Sheet valuation approach ignores growth, and seems to be the valuation method that most closely approximates Graham's net-net approach.
Earnings Power Value (EPV)
According to Greenwald, EPV provides somewhat less comfort than Balance Sheet Valuation, but can still be used to identify companies which can be excellent candidates for further research. The steps involved are first, computing an EPV (which is basically, sustainable capitalized earnings under a no growth scenario), and second, comparing the computed EPV to Balance Sheet reproduction cost (i.e., how much it would take a new entrant to the incumbent's industry to reproduce the incumbent's balance sheet).
In the case of EPV < Reproduction Cost, there is no sustainable franchise and no value of growth. This scenario could be typical of mature industries, or industries in decline.
In the case of EPV = Reproduction Cost, there is no sustainable franchise and no value of growth, and all competitors are making equivalent economc profits. This scenario could by typical of regulated utilities.
In the case of EPV > Reproduction Cost, this is the only scenario where there appears to be a sustainable franchise and growth has value. This scenario is probably typical of companies like Microsoft or Coca Cola.
Greenwald defines Earnings Power Value as:
Adjusted sustainable earnings / wacc
Greenwald uses a range of possible wacc's in order to determine a valuation matrix for capitalized EPV's. The beauty is, you don't actually need to come up with a precise calculation of wacc. A sensitivity range is sufficient. Say 5-6% for mature, established companies as the lower bound, 7-8% as the middle bound, and 10-12% as the upper bound. If the company being evaluated is speculative, you could even use an upper-upper bound of 15% (typical of VC type required discount rates).
In order to derive adjusted sustainable earnings, a certain mount of expertise is required, which involves more than just a cursory glance at a company's financial statements.
Greenwald works through two examples.: WD-40 and Intel. WD-40 is the simpler of the two comparables and only requires small adjustments to its reported results.
I've tried to summarize my understanding of the determination of adjusted sustainable earnings and Balance Sheet Reproduction by cycling through the following steps:
- Analyze EBIT margin over last 10 years. Take an average of the last 10 years EBIT margin to factor in cyclical swings
- Analyze special charges (restructuring, impairment, layoffs, etc.). Take an average over last 5 years and deduct from EBIT
- Depending on industry, may want to analyze R&D and SG&A as a % of sales over 10 years and add back this %ge to arrive at adjusted no growth EBIT
- Add back 25% of R&D and SG&A to determine adjusted EBIT
- Determine after tax adjusted EBIT
- Add Dep'n, deduct Capx, gives adjusted AT EBIT
- For Balance Sheet Reproduction, work line by line through the Balance Sheet and ascribe a recoverability %ge to each line. For example, cash is most likely worth 100% on the $1. Inventory on the other hand, may only be worth 50% in the case of a business with high rates of obsolescence.
- For Balance Sheet Reproduction, capitalize as a rule of thumb, 3 x (R&D + SG&A) in order to approximate what a competitor would need to spend in order to compete with the incumbent company.
- For Balance Sheet Reproduction, deduct non-interest bearing liabilities and excess cash
Working through an example, using Procter & Gamble, I came up with the following:
|Add special charges||0||0||0||0||0||0||0||1,576||308||0||973|
|EBIT before special charges||10,469||13,249||15,450||16,637||15,374||16,021||15,495||14,868||14,638||15,288||15,828||4.22%|
|Special charges as % of sales||0.00%||0.00%||0.00%||0.00%||0.00%||0.00%||0.00%||1.88%||0.37%||0.00%||1.20%||0.31%|
|Deduct avg special charges||-286||-286||-286||-286||-286||-286||-286||-286||-286||-286||-286|
|EBIT after avg special charges||10,183||12,963||15,164||16,351||15,088||15,735||15,209||14,582||14,352||15,002||15,542||4.32%|
|SG&A as % of sales||32.43%||32.02%||31.83%||31.29%||29.51%||31.67%||31.75%||31.57%||32.15%||30.48%||30.57%||31.39%|
|R&D as % of sales||0.00%||0.00%||0.00%||0.00%||0.00%||0.00%||0.00%||0.00%||0.00%||0.00%||0.00%||0.00%|
|Add back 25% of SG&A||4,600||5,462||6,085||6,394||5,658||6,250||6,438||6,605||6,638||6,329||6,210|
|Add back 25% of R&D||0||0||0||0||0||0||0||0||0||0||0|
|Adj EBIT AT||10,790||13,779||16,122||17,179||15,403||16,530||15,767||15,131||14,717||15,291||15,761||3.86%|
|NI per F/S||6,923||8,684||10,340||12,075||13,436||12,736||11,927||10,904||11,402||11,785||9,226||2.91%|
|Cost of capital rates:||EPV||Adj for Debt & Excess Cash||Adj EPV||O/S shares||EPV / share||BVPS / share||Excess EPV||Current price||Price / EPV||Prem/Discount vs EPV|
|Upper bound - VC||15%||105,071.75||-21722||83,349.75||2883||28.91||24.4||1.18||75.49||2.61||161.11%|
|Arbitrary 2 <10% - 12 %>||10.00%||157,607.62||-21722||135,885.62||2883||47.13||24.4||1.93||75.49||1.60||60.16%|
|LT equity return; 2014 - 1928||8.16%||193,049.48||-21722||171,327.48||2883||59.43||24.4||2.44||75.49||1.27||27.03%|
|Combined NI & Div growth||4.67%||337,699.31||-21722||315,977.31||2883||109.60||24.4||4.49||75.49||0.69||-31.12%|
|Adj Beta||Rf||Risk Premium||%D||%E||AT Int cost|
So, in general, I determined very little margin of safety at current prices. At all possible wacc's between 10% and 7.5%, my EPV model suggested a premium to EPV of between 15.5% and 60%.
Using a 7.5% wacc, in order to provide me with a 40% margin of safety (i.e., buying $1 for $.60), P&G would either have to fall to $40 from here, or AT adjusted EBIT would have to increase to $125 per share, all else being equal.
Now obviously, 10% is too high a wacc to use given the stability and predictability of the business, so maybe, I should include my CAPM derived wacc in my model results, which indicates a possible discount to EPV of about 23%. Duly noted, but then again, CAPM is derived based on beta, and beta is regularly accepted convention on Wall Street, therefore, I give little credit to it.
Next, Balance Sheet Reproduction:
You can see from below, that based on my initial evaluation, P&G does not appear to have EPV per share under my calculated wacc ranges in excess of Reproduction Cost per share. This sort of makes intuitive sense. Sales growth has stalled, whether due to competitive pressures across product lines or due to foreign exchange impact on sales (or both). This doesn't mean that P&G is not a good candidate for consideration. To me, it just means that at the current price, there is little margin of safety when evaluated in the context of endeavouring to buy $1 for $.50.
|Assets||As reported||Adjustment||Reproduction cost|
|Deferred tax assets||820||0%||0|
|Assets held for sale||3,632||0%||0|
|3 yrs SG&A and R&D||74,514||100%||74,514|
|Equity method investments||0||100%||0|
|Non-interest bearing liab's||-15,691|
|Equals total net reproduction cost||172,711|
|O/S shares||2,883||Greenwald Three States|
|Reproduction cost/sh||60||State One:|
|EPV/sh (avg)||57||AV > EPV||AV = EPV||AV < EPV|
|EPV less Reproduction cost/sh||-3||No franchise||Equilibrium||Franchise|
|A/P and accruals||15,691||100%||15,691|