Forbes recently published an interesting article examining the stock picking approaches of 7 legendary investors; highly recommended as food for thought. The article goaded me to quickly write down my own process and attempt to determine what these legends were doing that I was missing. Investment process determines a significant proportion of investment returns. A defined process continuously documented, practiced, measured and improved-over a sufficient length of time and a sufficient number of examples-reduces mistakes and improves returns. A defined process also allows benchmarking with other investors, leading to accelerated improvement for everyone as hypotheses are tested against market opportunities and success or failure is determined. Where better to start this effort than the stock picking approaches of these professional investors?
Briefly, then, my process is outlined below in the hope other investors will describe their approach and allow all of us playing the home game to compete more successfully against the matrix. Many professional investors have described their process in some detail: Benjamin Graham, Warren Buffett, Philip Fisher, Christopher Brown, William Bernstein, et. al. But, I find some of these long on theoretical constructs and short on practical advice. I hear, and agree, when Mr. Buffett intones "it's far better to buy a wonderful company at a fair price than it is to buy a fair company at a wonderful price." However, when my finger is hovering above the buy button and my family's future house payment is hinged on divining what "wonderful" actually means in this particular case with my hard earned dollars on the line then I crave the transcendental less and the procedural more. Mr. Buffett seems to enjoy enormous reserves of self-confidence in his investment decisions; I am closer to Heisenberg-statistically right, but precisely wrong. Mr. Buffett, perhaps quoting Keynes or Fermi, answers this unasked question when he argues "it's better to be approximately right than precisely wrong." Thank you, but now do I put my chips on red or black?
My process is simple: 1) screening, 2) valuation, 3) investigation, 4) purchasing, and 5) selling. Although my goal should be to buy wonderful companies at a fair price, my personality draws me to companies which are potentially valuable over an indefinitely long time horizon, preferably paying a good dividend while I wait, but currently experiencing hard times and therefore cheap by almost any valuation standard. I do love a wonderful price. This is not completely irrational, Fama and French provided academic validation for this approach, as did Joseph Piotroski and "dogs of the Dow" was once seriously considered a portfolio strategy. Still, I repeat Mr. Buffett's "wonderful company, fair price" mantra 50 times each morning while I brush my teeth and then hit my desk and buy Gazprom (OTC:OGZPY) again at a new low. Arrggh.
I screen on market cap > $500 million, stock price > $5, price-to-book < 1, dividend yield > 3, dividend payout ratio < 50%, and P/E < 16. This gives me a list of about 15 stocks, 2 of which I'm already in, 6 I've looked at and don't like, with a remaining group I have no interest in researching. And this list does not change rapidly. Many of the companies I bought in 2010 are still on this list of losers or future winners. So then I read what the various value-oriented funds are buying, scan the recommended dividend picks on Seeking Alpha and look for companies hitting 52-week lows or experiencing big headline risks. Many of my favorite bargain basement picks over the past few years, Sanofi (SNY), Bank of America (BAC), Citigroup (C), AXA (AXAHY.PK) increased share prices so much they no longer fit in my shopping basket. It's the right problem to have, but the dividends keep piling up and must (?) be invested somewhere. Of course, BAC and C don't fit the dividend requirement, but I remain hopeful on that point and as recompense for patience they've skyrocketed from their lows. Still and all, these screening criteria are agonizingly random compared to the legends described in the Forbes article. Only the P/E < 16 and the low price-to-book ratio have much in the way of expert testimony and academic research behind them. There are also dozens of other screening triggers, which could or should be used, but this is an area needing serious research and back-testing. Screening is very quick, I glance once a month and stare seriously once a year in the 4th quarter. The Forbes screen using Peter Lynch's methodology picked three of my favorite picks - CEMIG (CIG), Telecom Argentina (TEO) and Lukoil (OTC:LUKOY). LUKOY also hit on the Benjamin Graham screen run by Forbes. But, companies like CEMIG are really tough to love. You have to believe electricity usage will rise steadily as Brazilian GDP rises, that CEMIG will retain its licenses in the face of strong government opposition, that CEMIG's "aggregator" strategy makes sense, and on and on. Assuming these contingencies fall in the right direction CEMIG will be money, but it's not for those looking for a quick turn. TEO was rumored to be a target for nationalization and sunk like a stone. LUKOY is hamstrung by Gazprom's pipeline monopoly and Rosneft's penchant for snapping up competitors at pennies on the dollar. These all promise high dividends, but are they wonderful companies just experiencing a spot of trouble? We'll wait and see.
Valuation is a standard DCF (discounted cash flow) analysis using 10 years of historical EPS and FCF data projected forward for another 10 years and double-checked against the analyst consensus. Naturally, the single biggest issue is these "temporarily" broken companies I buy don't have predictable earnings so the DCF analysis is precisely right and approximately wrong all the time. To offset the garbage in/garbage out syndrome I always use 10% as the discount rate, never set growth above 10% and cap terminal growth at 3%. Once this bit of math is complete I take an interest in companies priced 50% below fair value. The DCF analysis is updated weekly just because it's fun and it's automated-new EPS and FCF figures are incorporated after every quarterly earnings release. Any extra dollars are dropped into stocks with prices below fair value, any stock more than 50% above fair value is sold-unless I really think it's a wonderful company or it's above fair value because a one-time event dropped earnings precipitously-like AT&T's (T) failure to consummate the purchase of T-Mobile. Valuation takes at most an hour per company. But I'm only dealing with 5 to 10 companies at this step; my portfolio turnover in an average year is usually below 5%. The total portfolio has 30 companies, but 10-15 of these will be sold once the Eurozone is back on track. These were companies just too cheap to pass up back when I thought the global economic meltdown would only last 1 year or 2 at most. Who guessed so many central bankers would think austerity gins up employment? I hate the fact I don't model companies as part of the valuation step. The analyst consensus is a poor substitute for actually throwing the company into Excel and testing some assumptions, but this remains work in slow progress.
Investigation is next and is the most time consuming step in the entire process. I greatly admire Peter Lynch's recommendation to buy what you know-although I can only buy Frito Lay so many times just because I eat Doritos. And I am openly envious of Philip Fisher's description of badgering suppliers, customers, clients and experts for the inside scoop on wonderful companies, but I just don't have enough on the line to justify peeking up and down the supply chain. I do force myself to ask strangers at parties which companies they admire the most then grill them on specifics-but this conduit is shrinking a bit as I am invited to fewer and fewer parties. In lieu of Fisher's prescription I read 10 years of SEC filings for each company I'm seriously considering, listen to a sampling of their conference calls-particularly the Q&A and review their financials and strategy against 3 to 5 key competitors. There are a number of financial screens, like excessive debt or low ROA, which can cause a company to be dropped at any point in this process. The amount of time spent here can be indefinitely large as the research is entertaining. A company might be on the potential buy list for a year or more before a final decision is made. Many companies are a solid buy based on strategy, market strength and financials, but go into the alert queue awaiting the right price. I would love to own Church and Dwight (CHD), but not at 23 times earnings. I felt the same way about Caterpillar (CAT) and Cummins (CMI) for a long time. During the global recession I was finally able to buy both, patience is not only a virtue, in investing it is essential. To me the year-over-year financials are less important than the strategy and the product value proposition. I bought Sanofi and AstraZeneca (AZN) not because their financials looked great-they looked terrible as both of these organizations tumbled over the patent cliff. I bought them because the global economy is likely to consume more drugs as incomes rise around the world and drug companies make unbelievable margins. Drug companies, oil companies, and software technology companies are basically licenses to print money once they hit upon a successful value proposition. Any of these companies can be destroyed by events, but usually it's a slow motion crash and much money can be made and an exit found long before the writing on the wall is fully visible to everyone. Too often my investigation is short circuited by emotion. Once a set of numbers is too beautiful to bear, the urge to buy is impossible to resist. Bank of America at $5? Mr. Buffett awakened to BAC in his bathtub. I bought BAC at $14, continued buying all the way down to $5 and bought on the way back up until $9. AXA at $10? Luckily I have limited funds or I would have violated all the diversification rules Mr. Bernstein tried to teach me. I'm not unhappy with my valuation process as it's outlined, I just need to ensure I stop breaking my own rules and take the time to let the math counter-balance my desire for the new shiny object. On average, maybe 20-40 hours are invested in each purchase decision-not counting the many hours spent on the pile that are ultimately rejected. Still, there is huge room for improvement here and I look forward to digging both deeper and more effectively. I now suspect understanding the economic structure of the value proposition is more important than understanding the financials as they currently stand. Many of the most successful companies have gone through periods where the financials looked terrible, but if they had a compelling value proposition and a focused strategy they could find a way to dig back up from any ephemeral exigencies. Against this backdrop I look at a company such as Gazprom and say, perhaps the management is unenlightened, perhaps they do treat shareholders like yesterday's garbage, maybe they negotiate with the Eurozone like a carjacker negotiating for your keys, but they do have a boat load of reserves and at some point corruption may fall out of favor. Although I worry Rosneft may consume them through a political process, failing this catastrophe, Gazprom may sell that boat load of reserves for a boat load of money in the far distant future-they may even start filing their financials contemporaneous with the year in question, if not the quarter. Worse comes to worse, if Gazprom implodes then my British Petroleum (BP) stock will just take a larger market share and grow faster. And, of course, I bought BP when everyone hated it and I hope they'll still be in business even if the legal settlements are double the current figures.
Once I have found cheap companies (I wish they were all wonderful, but some are just cheap) I buy cheap. Assuming the discipline provided by the DCF fair value analysis and the valuation assessment are followed the only issue with purchasing is to continue to purchase as long as the company's strategy and value proposition adhere to the original understanding. Many of these cheap companies are deeply hated, most will experience severe reductions in their purchase price right after you buy them. If Banco Santander (SAN) was a great buy at $13, then it is also a great buy at $5. It doesn't seem like a great buy as it sinks and sinks and sinks, but it is (or of course, it isn't-the game is played with real money after all). Whether the market ever recognizes or rewards your valuation is an unknown. You can be right all the way to zero and never recover. But, if you've done all this work and get it wrong-you're in good company. All you have to do is get it right a percentage of the time; way less than 70% and you'll still do fine. If this level of uncertainty is nerve-wracking, then index ETFs are extremely cheap and a large portfolio can be managed for a few hours of time a year and sleep is sweet.
The table below provides a bit more numeric detail about some of the troubling positions among my favorite picks. There are many which have done so well, relative to their original purchase price, they're no longer top of mind: BAC, C, CMI, AXAHY.PK, SNY, COP, GSH, MFG, and TEO. The positions in the table, however, range from problematic to scary so they may be interesting to review in more detail.
|Company||Fair Value||Stock Price||Margin of Safety||Dividend Yield|
*Gazprom does not have a fixed dividend. 5.6% was their forecast for 2013. Gazprom pays 25% of net income under Russian accounting standards, which is approximately 12% under IFRS. CEMIG does not have a fixed dividend. They pay 50% of net profits from the prior year and every two years may pay a special dividend.
I hold Intel (INTC) because it hasn't hit my sell price, but I also wouldn't put a lot of money back in until Intel demonstrates more earnings growth. Meanwhile the dividend is welcome. There are no hard and fast cutoffs as the DCF analysis can easily jump 10-20% through small changes in assumptions. My investment thesis is that the world will always need chips and Intel is able to effectively allocate capital building chip foundries, which produce slight competitive advantages for a limited period of time. I don't see anyone in this space establishing a clear leadership position; it will be a continual seesaw battle. Unlike the airline business there seems to be room for aggressive competition and adequate margins. Until this changes Intel is a dividend paying wait and see.
The Microsoft (MSFT) story is well known. They're inexpensive because they're going out of business sometime in the next 10 weeks or 10 years. PCs are dead, smartphones are taking over and Microsoft doesn't understand the consumer hardware business so there is no hope. On the other hand, Microsoft does have a recognized brand, an enormous installed base, and adequate financial resources to engage in serious transformation initiatives. It's not clear Microsoft is a wonderful company right now, but they seem to have all the ingredients to become a wonderful company.
AstraZeneca was purchased because it pays a reasonable dividend and was priced for patent cliff Armageddon. My investment thesis for AZN is that smart people will figure out how to bridge the patent cliff and the entire pharma industry will expand as global wealth expands over the next decade. One can spend a lot of time figuring out which is the best pharma to buy, but I don't have that expertise so I bought a few that were suffering and one or two of them will pay off big, others will just pay big dividends.
Caterpillar is one of those companies I just had to own. My uncle was a Caterpillar repair technician and gave us kids giant ball bearings from renovated machines. To me Caterpillar is a technology company like Microsoft and Intel. Their basic task is innovation, turning ideas into useful machines, and Caterpillar seems to do this as well as anyone. At some point global growth will return and these giant technology companies will once again look like bargains.
British Petroleum is more interesting because fair value is so far above the current share price. The rationale for this discount is obvious, BP faces somewhere between $10 to $40 billion in additional legal damages for the Gulf oil spill and their alleged prevarication around same. My investment thesis is simple here. Any foreseeable level of damages will not bankrupt the company and BP will come out of this leaner with a more enlightened management team. Although oil is a commodity it is relatively expensive to find and refine and the largest players maintain monopoly pricing for the entire industry. BP is a case where the home player has a decided advantage over the professional investor. Most professionals want to buy under-valued stocks with a near-term catalyst to goose their yearly performance. As a home player I don't care whether I get my return in 1 year or 10 years. Since my goal is to buy long-term dividends on the cheap I'm getting 5% to wait and could be handed a 50-100% capital appreciation at some future point. A large legal judgment could cut investment and dividends for a few years, but in the long run this won't matter.
With Lukoil we enter serious return and serious risk territory. Fair value is 150% above the current share price. Lukoil faces strategic threats from Rosneft and Gazprom, but they also have an experienced management with a track record of overcoming these threats. Currently paying a 4.88% dividend Lukoil appears to be heavily discounted from any reasonable valuation.
Gazprom is heavily discounted and probably deserves some of that discount. Fair value is way above the current share price. Gazprom reports results months after the quarter, explains corporate policy with either extreme double-talk or incredibly inept translations and generally runs the company without regard to shareholder communication. Despite all this Gazprom enjoys a relationship with China, controls gas supplies to much of Europe and sits on some of the largest reserves in Russia. This is not a wonderful company, this is a cheap stock, but Russia will evolve and Gazprom will evolve and they will eventually make enough money to be completely corrupt or not and still reward shareholders adequately.
CEMIG and Electrobras basically suffer from the same malady-government intervention. The newly elected President of Brazil, rightly or wrongly can be debated, determined electricity prices were hindering economic development. The government decided to tie future licenses for generation sites to price reductions. CEMIG took the fight to the courts and the stock price dropped precipitously. Electrobras, a government-controlled entity, decided to agree to the new terms and their stock price dropped even more. My investment thesis is that electricity rates short term don't really matter. Brazil's per capita income is a fraction of the U.S. and will continue to rise. As GDP rises so will electricity usage and rates will sort themselves out to ensure sufficient capital investment to maintain growth. CEMIG and Electrobras will grow around 5%+/- as GDP grows and rates will normalize as the various forces compete. If CEMIG wins in court the stock price will jump. If CEMIG loses then Electrobras and CEMIG will both figure out how to run at reduced expense levels. There may be a few years of disruption, but this is de minimus when the long-term upside is considered. Meanwhile we are able to establish a low-cost entry point in two utilities that will provide an attractive dividend stream once the short-term disruption is sorted.
Selling is the final step and one seldom used unless a better opportunity presents itself, the company diverts off-strategy or money is being forced into your hands. Keep in mind the portfolio strategy here is to buy dividends cheaply. We're looking for solid, temporarily hated or scorned or risky companies, paying reasonable dividends. When they're being dragged through the mud their dividends are either sky-high, like Banco Santander , or are severely cut while they recover, like Citi. Once these companies spring back to life, the rest of the world is buying them again at fair value and making a normal 2-3% dividend, we're raking in 10-15% dividends because we bought them dirt cheap, waited patiently while they cleaned themselves up, and kept them in our portfolio until we reaped the rewards. Assuming the average return earned in the market is 8%, then we should be happy shepherding our flock of fallen angels toward a steady dividend with a normalized 3-4% underlying appreciation as icing. These stocks often continue to bounce around a lot. They will gyrate up and down 10%, 20%, or 30% as the market loves them or ignores them. If the market is willing to pay you 50% more than fair value for something you bought at 50% below fair value then selling becomes pretty compelling. However, it is also rewarding to just ignore the pricing waves and sell only when the value proposition shows signs of deterioration.
Briefly then, these are my 5 steps. It is an approach almost guaranteed to not provide predictable or quick returns-other than the dividends and the dividends are often cut as these wounded entities struggle back to their feet. In many cases large losses will quickly follow large investments and it may take years to recover-if there is any recovery. If you like certainty, do not try this at home. If you do not believe a valuable company becomes more valuable as its stock price drops then this is not a good strategy for you. On the other hand, this approach may at least be a starting point for codifying your own investment strategy and enhancing your investment process and it does have its monetary rewards when the chips fall your way.
The investing legends mentioned in the Forbes article achieved extraordinary returns. Some portion of this return resulted from the disciplined, focused approach they brought to their investment regimen. Mr. Buffett, in particular, often argues the value investing approach can be taught to anyone and the patient application of its principles will bring appropriate rewards to the studious practitioner. Each of us may benefit from documenting our own approach and comparing those results against the legends. It's a much tougher standard than the traditional index benchmarks, but not a bad target at which to aim.